Produced by Giovanni Fini, Marilynda Fraser-Cunliffe and
the Online Distributed Proofreading Team at
http://www.pgdp.net (This file was produced from images
generously made available by The Internet Archive)







                         TRANSCRIBER’S NOTES:

—Obvious print and punctuation errors were corrected.

—Superscripts have been rendered as a^b or a^{bc}.

—Subscripts have been rendered as a_{b}.

—Bold text has been rendered as =bold text=.




                        MENTAL EVOLUTION IN MAN




                          BY THE SAME AUTHOR.


                           _Crown 8vo. 5s._

               =JELLY-FISH, STAR-FISH, AND SEA-URCHINS.=

            Being a Research on Primitive Nervous Systems.

                  [_International Scientific Series._


                           _Crown 8vo. 5s._

                        =ANIMAL INTELLIGENCE.=

                            Fourth Edition.

                  [_International Scientific Series._


                           _Demy 8vo. 12s._

                    =MENTAL EVOLUTION IN ANIMALS.=

                           Second Thousand.

     With a Posthumous Essay on Instinct by CHARLES DARWIN, F.R.S.

                   LONDON: KEGAN PAUL, TRENCH & CO.

[Illustration]


                        MENTAL EVOLUTION IN MAN

                       _ORIGIN OF HUMAN FACULTY_

                                  BY
               GEORGE JOHN ROMANES, M.A., LL.D., F.R.S.

[Illustration: LOGO]

                                LONDON
            KEGAN PAUL, TRENCH & CO., 1, PATERNOSTER SQUARE
                                 1888




    (_The rights of translation and of reproduction are reserved._)




PREFACE.


In now carrying my study of mental evolution into the province of human
psychology, it is desirable that I should say a few words to indicate
the scope and intention of this the major portion of my work. For it is
evident that “Mental Evolution in Man” is a subject comprehending so
enormous a field that, unless some lines of limitation are drawn within
which its discussion is to be confined, no one writer could presume to
deal with it.

The lines, then, which I have laid down for my own guidance are these.
My object is to seek for the principles and causes of mental evolution
in man, first as regards the origin of human faculty, and next as
regards the several main branches into which faculties distinctively
human afterwards ramified and developed. In order as far as possible
to gain this object, it has appeared to me desirable to take large
or general views, both of the main trunk itself, and also of its
sundry branches. Therefore I have throughout avoided the temptation
of following any of the branches into their smaller ramifications, or
of going into the details of progressive development. These, I have
felt, are matters to be dealt with by others who are severally better
qualified for the task, whether their special studies have reference to
language, archæology, technicology, science, literature, art, politics,
morals, or religion. But, in so far as I shall subsequently have to
deal with these subjects, I will do so with the purpose of arriving
at general principles bearing upon mental evolution, rather than with
that of collecting facts or opinions for the sake of their intrinsic
interest from a purely historical point of view.

Finding that the labour required for the investigation, even as thus
limited, is much greater than I originally anticipated, it appears
to me undesirable to delay publication until the whole shall have
been completed. I have therefore decided to publish the treatise in
successive instalments, of which the present constitutes the first. As
indicated by the title, it is concerned exclusively with the Origin
of Human Faculty. Future instalments will deal with the Intellect,
Emotions, Volition, Morals, and Religion. It will, however, be several
years before I shall be in a position to publish these succeeding
instalments, notwithstanding that some of them are already far advanced.

Touching the present instalment, it is only needful to remark that from
a controversial point of view it is, perhaps, the most important. If
once the genesis of conceptual thought from non-conceptual antecedents
be rendered apparent, the great majority of competent readers at
the present time would be prepared to allow that the psychological
barrier between the brute and the man is shown to have been overcome.
Consequently, I have allotted what might otherwise appear to be a
disproportionate amount of space to my consideration of this the
_origin_ of human faculty—disproportionate, I mean, as compared
with what has afterwards to be said touching the _development_ of
human faculty in its several branches already named. Moreover, in the
present treatise I shall be concerned chiefly with the psychology of
my subject—reserving for my next instalment a full consideration of
the light which has been shed on the mental and social condition of
early man by the study of his own remains on the one hand, and of
existing savages on the other. Even as thus restricted, however, the
subject-matter of the present treatise will be found more extensive
than most persons would have been prepared to expect. For it does not
appear to me that this subject-matter has hitherto received at the
hands of psychologists any approach to the amount of analysis of
which it is susceptible, and to which—in view of the general theory
of evolution—it is unquestionably entitled. But I have everywhere
endeavoured to avoid undue prolixity, trusting that the intelligence
of any one who is likely to read the book will be able to appreciate
the significance of important points, without the need of expatiation
on the part of the writer. The only places, therefore, where I feel
that I may be fairly open to the charge of unnecessary reiteration, are
those in which I am endeavouring to render fully intelligible the newer
features of my analysis. But even here I do not anticipate that readers
of any class will complain of the efforts which are thus made to assist
their understanding of a somewhat complicated matter.

As no one has previously gone into this matter, I have found myself
obliged to coin a certain number of new terms, for the purpose at
once of avoiding continuous circumlocution, and of rendering aid to
the analytic inquiry. For my own part I regret this necessity, and
therefore have not resorted to it save where I have found the force of
circumstances imperative. In the result, I do not think that adverse
criticism is likely to fasten upon any of these new terms as needless
for the purposes of my inquiry. Every worker is free to choose his own
instruments; and when none are ready-made to suit his requirements, he
has no alternative but to fashion those which may.

To any one who already accepts the general theory of evolution as
applied to the human mind, it may well appear that the present
instalment of my work is needlessly elaborate. Now, I can quite
sympathize with any evolutionist who may thus feel that I have brought
steam-engines to break butterflies; but I must ask such a man to
remember two things. First, that plain and obvious as the truth may
seem to him, it is nevertheless a truth that is very far from having
received general recognition, even among more intelligent members of
the community: seeing, therefore, of how much importance it is to
establish this truth as an integral part of the doctrine of descent,
I cannot think that either time or energy is wasted in a serious
endeavour to do so, even though to minds already persuaded it may
seem unnecessary to have slain our opponents in a manner quite so
mercilessly minute. Secondly, I must ask these friendly critics to take
note that, although the discussion has everywhere been thrown into the
form of an answer to objections, it really has a much wider scope: it
aims not only at an overthrow of adversaries, but also, and even more,
at an exposition of the principles which have probably been concerned
in the “Origin of Human Faculty.”

The Diagram which is reproduced from my previous work on “Mental
Evolution in Animals,” and which serves to represent the leading
features of psychogenesis throughout the animal kingdom, will reappear
also in succeeding instalments of the work, when it will be continued
so as to represent the principal stages of “Mental Evolution in Man.”

  18, CORNWALL TERRACE, REGENT’S PARK,
  _July, 1888_.




CONTENTS.


  CHAPTER                                          PAGE

     I. MAN AND BRUTE                                 1

    II. IDEAS                                        20

   III. LOGIC OF RECEPTS                             40

    IV. LOGIC OF CONCEPTS                            70

     V. LANGUAGE                                     85

    VI. TONE AND GESTURE                            104

   VII. ARTICULATION                                121

  VIII. RELATION OF TONE AND GESTURE TO WORDS       145

    IX. SPEECH                                      163

     X. SELF-CONSCIOUSNESS                          194

    XI. THE TRANSITION IN THE INDIVIDUAL            213

   XII. COMPARATIVE PHILOLOGY                       238

  XIII. ROOTS OF LANGUAGE                           264

   XIV. THE WITNESS OF PHILOLOGY                    294

    XV. THE WITNESS OF PHILOLOGY—_continued_       326

   XVI. THE TRANSITION IN THE RACE                  360

  XVII. GENERAL SUMMARY AND CONCLUDING REMARKS      390




MENTAL EVOLUTION IN MAN.




CHAPTER I.

MAN AND BRUTE.


Taking up the problems of psychogenesis where these were left in my
previous work, I have in the present treatise to consider the whole
scope of mental evolution in man. Clearly the topic thus presented
is so large, that in one or other of its branches it might be taken
to include the whole history of our species, together with our
pre-historic development from lower forms of life, as already indicated
in the Preface. However, it is not my intention to write a history
of civilization, still less to develop any elaborate hypothesis of
anthropogeny. My object is merely to carry into an investigation of
human psychology a continuation of the principles which I have already
applied to the attempted elucidation of animal psychology. I desire to
show that in the one province, as in the other, the light which has
been shed by the doctrine of evolution is of a magnitude which we are
now only beginning to appreciate; and that by adopting the theory of
continuous development from the one order of mind to the other, we are
able scientifically to explain the whole mental constitution of man,
even in those parts of it which, to former generations, have appeared
inexplicable.

In order to accomplish this purpose, it is not needful that I should
seek to enter upon matters of detail in the application of those
principles to the facts of history. On the contrary, I think that
any such endeavour—even were I qualified to make it—would tend only
to obscure my exposition of those principles themselves. It is enough
that I should trace the operation of such principles, as it were, in
outline, and leave to the professed historian the task of applying them
in special cases.

The present work being thus a treatise on human psychology in relation
of the theory of descent, the first question which it must seek to
attack is clearly that as to the evidence of the mind of man having
been derived from mind as we meet with it in the lower animals. And
here, I think, it is not too much to say that we approach a problem
which is not merely the most interesting of those that have fallen
within the scope of my own works; but perhaps the most interesting
that has ever been submitted to the contemplation of our race. If it
is true that “the proper study of mankind is man,” assuredly the study
of nature has never before reached a territory of thought so important
in all its aspects as that which in our own generation it is for the
first time approaching. After centuries of intellectual conquest in
all regions of the phenomenal universe, man has at last begun to
find that he may apply in a new and most unexpected manner the adage
of antiquity—_Know thyself_. For he has begun to perceive a strong
probability, if not an actual certainty, that his own living nature is
identical in kind with the nature of all other life, and that even the
most amazing side of this his own nature—nay, the most amazing of all
things within the reach of his knowledge—the human mind itself, is but
the topmost inflorescence of one mighty growth, whose roots and stem
and many branches are sunk in the abyss of planetary time. Therefore,
with Professor Huxley we may say:—“The importance of such an inquiry
is indeed intuitively manifest. Brought face to face with these blurred
copies of himself, the least thoughtful of men is conscious of a
certain shock, due perhaps not so much to disgust at the aspect of
what looks like an insulting caricature, as to the awaking of a sudden
and profound mistrust of time-honoured theories and strongly rooted
prejudices regarding his own position in nature, and his relations
to the wider world of life; while that which remains a dim suspicion
for the unthinking, becomes a vast argument, fraught with the deepest
consequences, for all who are acquainted with the recent progress of
anatomical and physiological sciences.”[1]

The problem, then, which in this generation has for the first time been
presented to human thought, is the problem of how this thought itself
has come to be. A question of the deepest importance to every system
of philosophy has been raised by the study of biology; and it is the
question whether the mind of man is essentially the same as the mind of
the lower animals, or, having had, either wholly or in part, some other
mode of origin, is essentially distinct—differing not only in degree
but in kind from all other types of psychical being. And forasmuch as
upon this great and deeply interesting question opinions are still much
divided—even among those most eminent in the walks of science who
agree in accepting the principles of evolution as applied to explain
the mental constitution of the lower animals,—it is evident that the
question is neither a superficial nor an easy one. I shall, however,
endeavour to examine it with as little obscurity as possible, and also,
I need hardly say, with all the impartiality of which I am capable,[2]

It will be remembered that in the introductory chapter of my previous
work I have already briefly sketched the manner in which I propose to
treat this question. Here, therefore, it is sufficient to remark that
I began by assuming the truth of the general theory of descent so far
as the animal kingdom is concerned, both with respect to bodily and to
mental organization; but in doing this I expressly excluded the mental
organization of man, as being a department of comparative psychology
with reference to which I did not feel entitled to assume the
principles of evolution. The reason why I made this special exception,
I sufficiently explained; and I shall therefore now proceed, without
further introduction, to a full consideration of the problem that is
before us.

       *       *       *       *       *

First, let us consider the question on purely _a priori_ grounds. In
accordance with our original hypothesis—upon which all naturalists of
any standing are nowadays agreed—the process of organic and of mental
evolution has been continuous throughout the whole region of life and
of mind, with the one exception of the mind of man. On grounds of
analogy, therefore, we should deem it antecedently improbable that the
process of evolution, elsewhere so uniform and ubiquitous, should have
been interrupted at its terminal phase. And looking to the very large
extent of this analogy, the antecedent presumption which it raises is
so considerable, that in my opinion it could only be counterbalanced by
some very cogent and unmistakable facts, showing a difference between
animal and human psychology so distinctive as to render it in the
nature of the case virtually impossible that the one could ever have
graduated into the other. This I posit as the first consideration.

Next, still restricting ourselves to an _a priori_ view, it is
unquestionable that human psychology, in the case of every individual
human being, presents to actual observation a process of gradual
development, or evolution, extending from infancy to manhood; and that
in this process, which begins at a zero level of mental life and may
culminate in genius, there is nowhere and never observable a sudden
leap of progress, such as the passage from one order of psychical
being to another might reasonably be expected to show. Therefore, it
is a matter of observable fact that, whether or not human intelligence
differs from animal in kind, it certainly does admit of gradual
development from a zero level. This I posit as the second consideration.

Again, so long as it is passing through the lower phases of its
development, the human mind assuredly ascends through a scale of
mental faculties which are parallel with those that are permanently
presented by the psychological species of the animal kingdom. A glance
at the Diagram which I have placed at the beginning of my previous
work will serve to show in how strikingly quantitative, as well as
qualitative, a manner the development of an individual human mind
follows the order of mental evolution in the animal kingdom. And when
we remember that, at all events up to the level where this parallel
ends, the diagram in question is not an expression of any psychological
theory, but of well-observed and undeniable psychological fact, I think
every reasonable man must allow that, whatever the explanation of
this remarkable coincidence may be, it certainly must admit of _some_
explanation—_i.e._ cannot be ascribed to mere chance. But, if so, the
only explanation available is that which is furnished by the theory of
descent. These facts, which I present as a third consideration, tend
still further—and, I think, most strongly—to increase the force of
antecedent presumption against any hypothesis which supposes that the
process of evolution can have been discontinuous in the region of mind.

Lastly, it is likewise a matter of observation, as I shall fully
show in the next instalment of this work, that in the history of our
race—as recorded in documents, traditions, antiquarian remains, and
flint implements—the intelligence of the race has been subject to a
steady process of gradual development. The force of this consideration
lies in its proving, that if the process of mental evolution was
suspended between the anthropoid apes and primitive man, it was again
resumed with primitive man, and has since continued as uninterruptedly
in the human species as it previously did in the animal species.
Now, upon the face of these facts, or from a merely antecedent point
of view, such appears to me, to say the least, a highly improbable
supposition. At all events, it certainly is not the kind of supposition
which men of science are disposed to regard with favour elsewhere; for
a long and arduous experience has taught us that the most paying kind
of supposition which we can bring with us into our study of nature, is
that which recognizes in nature the principle of _continuity_.

Taking, then, these several _a priori_ considerations together, they
must, in my opinion, be fairly held to make out a very strong _primâ
facie_ case in favour of the view that there has been no interruption
of the developmental process in the course of psychological history;
but that the mind of man, like the mind of animals—and, indeed, like
everything else in the domain of living nature—has been evolved. For
these considerations show, not only that on analogical grounds any such
interruption must be held as in itself improbable; but also that there
is nothing in the constitution of the human mind incompatible with the
supposition of its having been slowly evolved, seeing that not only in
the case of every individual life, but also during the whole history
of our species, the human mind actually _does_ undergo, and _has_
undergone, the process in question.

In order to overturn so immense a presumption as is thus erected on
_a priori_ grounds, the psychologist must fairly be called upon to
supply some very powerful considerations of an _a posteriori_ kind,
tending to show that there is something in the constitution of the
human mind which renders it virtually impossible—or at all events
exceedingly difficult to imagine—that it can have proceeded by way of
genetic descent from mind of lower orders. I shall therefore proceed to
consider, as carefully and as impartially as I can, the arguments which
have been adduced in support of this thesis.

       *       *       *       *       *

In the introductory chapter of my previous work I observed, that the
question whether or not human intelligence has been evolved from animal
intelligence can only be dealt with scientifically by comparing the one
with the other, in order to ascertain the points wherein they agree
and the points wherein they differ. I shall, therefore, here begin
by briefly stating the points of agreement, and then proceed more
carefully to consider all the more important views which have hitherto
been propounded concerning the points of difference.

If we have regard to Emotions as these occur in the brute, we cannot
fail to be struck by the broad fact that the area of psychology which
they cover is so nearly co-extensive with that which is covered by the
emotional faculties of man. In my previous works I have given what I
consider unquestionable evidence of all the following emotions, which
I here name in the order of their appearance through the psychological
scale,—fear, surprise, affection, pugnacity, curiosity, jealousy,
anger, play, sympathy, emulation, pride, resentment, emotion of the
beautiful, grief, hate, cruelty, benevolence, revenge, rage, shame,
regret, deceitfulness, emotion of the ludicrous.[3]

Now, this list exhausts all the human emotions, with the exception
of those which refer to religion, moral sense, and perception of the
sublime. Therefore I think we are fully entitled to conclude that, so
far as emotions are concerned, it cannot be said that the facts of
animal psychology raise any difficulties against the theory of descent.
On the contrary, the emotional life of animals is so strikingly similar
to the emotional life of man—and especially of young children—that I
think the similarity ought fairly to be taken as direct evidence of a
genetic continuity between them.

And so it is with regard to Instinct. Understanding this term in
the sense previously defined,[4] it is unquestionably true that in
man—especially during the periods of infancy and youth—sundry
well-marked instincts are presented, which have reference chiefly to
nutrition, self-preservation, reproduction, and the rearing of progeny.
No one has ventured to dispute that all these instincts are identical
with those which we observe in the lower animals; nor, on the other
hand, has any one ventured to suggest that there is any instinct which
can be said to be peculiar to man, unless the moral and religious
sentiments are taken to be of the nature of instincts. And although it
is true that instinct plays a larger part in the psychology of many
animals than it does in the psychology of man, this fact is plainly of
no importance in the present connection, where we are concerned only
with identity of principle. If any one were childish enough to argue
that the mind of a man differs in kind from that of a brute because
it does not display any particular instinct—such, for example, as
the spinning of webs, the building of nests, or the incubation of
eggs,—the answer of course would be that, by parity of reasoning, the
mind of a spider must be held to differ in kind from that of a bird. So
far, then, as instincts and emotions are concerned, the parallel before
us is much too close to admit of any argument on the opposite side.

With regard to Volition more will be said in a future instalment of
this work. Here, therefore, it is enough to say, in general terms,
that no one has seriously questioned the identity of kind between the
animal and the human will, up to the point at which so-called freedom
is supposed by some dissentients to supervene and characterize the
latter. Now, of course, if the human will differs from the animal will
in any important feature or attribute such as this, the fact must be
duly taken into account during the course of our subsequent analysis.
At present, however, we are only engaged upon a preliminary sketch of
the points of resemblance between animal and human psychology. So far,
therefore, as we are now concerned with the will, we have only to note
that up to the point where the volitions of a man begin to surpass
those of a brute in respect of complexity, refinement, and foresight,
no one disputes identity of kind.

Lastly, the same remark applies to the faculties of Intellect.[5]
Enormous as the difference undoubtedly is between these faculties
in the two cases, the difference is conceded not to be one of kind
_ab initio_. On the contrary, it is conceded that up to a certain
point—namely, as far as the highest degree of intelligence to which
an animal attains—there is not merely a similarity of kind, but an
identity of correspondence. In other words, the parallel between
animal and human intelligence which is presented in my Diagram, and to
which allusion has already been made, is not disputed. The question,
therefore, only arises with reference to those superadded faculties
which are represented above the level marked 28, where the upward
growth of animal intelligence ends, and the growth of distinctively
human intelligence begins. But even at level 28 the human mind is
already in possession of many of its most useful faculties, and these
it does not afterwards shed, but carries them upwards with it in the
course of its further development—as we well know by observing the
psychogenesis of every child. Now, it belongs to the very essence of
evolution, considered as a process, that when one order of existence
passes on to higher grades of excellence, it does so upon the
foundation already laid by the previous course of its progress; so that
when compared with any allied order of existence which has not been
carried so far in this upward course, a more or less close parallel
admits of being traced between the two, up to the point at which
the one begins to distance the other, where all further comparison
admittedly ends. Therefore, upon the face of them, the facts of
comparative psychology now before us are, to say the least, strongly
suggestive of the superadded powers of the human intellect having been
due to a process of evolution.

Lest it should be thought that in this preliminary sketch of
the resemblances between human and brute psychology I have been
endeavouring to draw the lines with a biased hand, I will here quote
a short passage to show that I have not misrepresented the extent
to which agreement prevails among adherents of otherwise opposite
opinions. And for this purpose I select as spokesman a distinguished
naturalist, who is also an able psychologist, and to whom, therefore,
I shall afterwards have occasion frequently to refer, as on both these
accounts the most competent as well as the most representative of my
opponents. In his Presidential Address before the Biological Section of
the British Association in 1879, Mr. Mivart is reported to have said:—

“I have no wish to ignore the marvellous powers of animals, or the
resemblance of their actions to those of man. No one can reasonably
deny that many of them have feelings, emotions, and sense-perceptions
similar to our own; that they exercise voluntary motion, and perform
actions grouped in complex ways for definite ends; that they to a
certain extent learn by experience, and combine perceptions and
reminiscences so as to draw practical inferences, directly apprehending
objects standing in different relations one to another, so that, in
a sense, they may be said to apprehend relations. They will show
hesitation, ending apparently, after a conflict of desires, with what
looks like choice or volition; and such animals as the dog will not
only exhibit the most marvellous fidelity and affection, but will
also manifest evident signs of shame, which may seem the outcome of
incipient moral perceptions. It is no great wonder, then, that so many
persons, little given to patient and careful introspection, should fail
to perceive any radical distinction between a nature thus gifted and
the intellectual nature of man.”

       *       *       *       *       *

We may now turn to consider the points wherein human and brute
psychology have been by various writers alleged to differ.

The theory that brutes are non-sentient machines need not detain us,
as no one at the present day is likely to defend it.[6] Again, the
distinction between human and brute psychology that has always been
taken more or less for granted—namely, that the one is rational and
the other irrational—may likewise be passed over after what has been
said in the chapter on Reason in my previous work. For it is there
shown that if we use the term Reason in its true, as distinguished
from its traditional sense, there is no fact in animal psychosis more
patent than that this psychosis is capable in no small degree of
_ratiocination_. The source of the very prevalent doctrine that animals
have no germ of reason is, I think, to be found in the fact that reason
attains a much higher level of development in man than in animals,
while instinct attains a higher development in animals than in man:
popular phraseology, therefore, disregarding the points of similarity
while exaggerating the more conspicuous points of difference,
designates all the mental faculties of the animal instinctive, in
contradistinction to those of man, which are termed rational. But
unless we commit ourselves to an obvious reasoning in a circle, we must
avoid assuming that all actions of animals are instinctive, and then
arguing that, because they are instinctive, therefore they differ in
kind from those actions of man which are rational. The question really
lies in what is here assumed, and can only be answered by examining
in what essential respect instinct differs from reason. This I have
endeavoured to do in my previous work with as much precision as the
nature of the subject permits; and I think I have made it evident, in
the first place, that there is no such immense distinction between
instinct and reason as is generally assumed—the former often being
blended with the latter, and the latter as often becoming transmuted
into the former,—and, in the next place, that all the higher animals
manifest in various degrees the faculty of inferring. Now, _this is the
faculty of reason, properly so called_; and although it is true that in
no case does it attain in animal psychology to more than a rudimentary
phase of development as contrasted with its prodigious growth in man,
this is clearly quite another matter where the question before us is
one concerning difference of kind.[7]

Again, the theological distinction between men and animals may be
passed over, because it rests on a dogma with which the science of
psychology has no legitimate point of contact. Whether or not the
conscious part of man differs from the conscious part of animals in
being immortal, and whether or not the “spirit” of man differs from
the “soul” of animals in other particulars of kind, dogma itself would
maintain that science has no voice in either affirming or denying.
For, from the nature of the case, any information of a positive kind
relating to these matters can only be expected to come by way of a
Revelation; and, therefore, however widely dogma and science may differ
on other points, they are at least agreed upon this one—namely, if the
conscious life of man differs thus from the conscious life of brutes,
Christianity and Philosophy alike proclaim that only by a Gospel could
its endowment of immortality have been brought to light.[8]

Another distinction between the man and the brute which we often find
asserted is, that the latter shows no signs of mental progress in
successive generations. On this alleged distinction I may remark,
first of all, that it begs the whole question of mental evolution in
animals, and, therefore, is directly opposed to the whole body of facts
presented in my work upon this subject. In the next place, I may remark
that the alleged distinction comes with an ill grace from opponents of
evolution, seeing that it depends upon a recognition of the principles
of evolution in the history of mankind. But, leaving aside these
considerations, I meet the alleged distinction with a plain denial of
both the statements of fact on which it rests. That is to say, I deny
on the one hand that mental progress from generation to generation is
an invariable peculiarity of human intelligence; and, on the other
hand, I deny that such progress is never found to occur in the case of
animal intelligence.

Taking these two points separately, I hold it to be a statement
opposed to fact to say, or to imply, that all existing savages, when
not brought into contact with civilized man, undergo intellectual
development from generation to generation. On the contrary, one of the
most generally applicable statements we can make with reference to the
psychology of uncivilized man is that it shows, in a remarkable degree,
what we may term a _vis inertiæ_ as regards upward movement. Even
so highly developed a type of mind as that of the Negro—submitted,
too, as it has been in millions of individual cases to close contact
with minds of the most progressive type, and enjoying as it has in
many thousands of individual cases all the advantages of liberal
education—has never, so far as I can ascertain, executed one single
stroke of original work in any single department of intellectual
activity.

Again, if we look to the whole history of man upon this planet as
recorded by his remains, the feature which to my mind stands out
in most marked prominence is the almost incredible slowness of his
intellectual advance, during all the earlier millenniums of his
existence. Allowing full weight to the consideration that “the
Palæolithic age, referring as the phrase does to a stage of culture,
and not to any chronological period, is something which has come and
gone at very different dates in different parts of the world;”[9]
and that the same remark may be taken, in perhaps a smaller measure,
to apply to the Neolithic age; still, when we remember what enormous
lapses of time these ages may be roughly taken to represent, I think
it is a most remarkable fact that, during the many thousands of years
occupied by the former, the human mind should have practically made no
advance upon its primitive methods of chipping flints; or that during
the time occupied by the latter, this same mind should have been so
slow in arriving, for example, at even so simple an invention as that
of substituting horns for flints in the manufacture of weapons. In my
next volume, where I shall have to deal especially with the evidence
of intellectual evolution, I shall have to give many instances, all
tending to show its extraordinarily slow progress during these æons of
pre-historic time. Indeed, it was not until the great step had been
made of substituting metals for both stones and horns, that mental
evolution began to proceed at anything like a measurable rate. Yet
this was, as it were, but a matter of yesterday. So that, upon the
whole, if we have regard to the human species generally—whether over
the surface of the earth at the present time, or in the records of
geological history,—we can no longer maintain that a tendency to
improvement in successive generations is here a leading characteristic.
On the contrary, any improvement of so rapid and continuous a kind as
that which is really contemplated, is characteristic only of a small
division of the human race during the last few hours, as it were, of
its existence.

On the other hand, as I have said, it is not true that animal species
never display any traces of intellectual improvement from generation
to generation. Were this the case, as already remarked, mental
evolution could never have taken place in the brute creation, and
so the phenomena of mind would have been wholly restricted to man:
all animals would have required to present but a vegetative form of
life. But, apart from this general consideration, we meet with many
particular instances of mental improvement in successive generations of
animals, taking place even within the limited periods over which human
observations can extend. In my previous work numerous cases will be
found (especially in the chapters on the plasticity and blended origin
of instincts), showing that it is quite a usual thing for birds and
mammals to change even the most strongly inherited of their instinctive
habits, in order to improve the conditions of their life in relation
to some change which has taken place in their environments. And if it
should be said that in such a case “the animal still does not rise
above the level of birdhood or of beasthood,” the answer, of course,
is, that neither does a Shakespeare or a Newton rise above the level of
manhood.

On the whole, then, I cannot see that there is any valid distinction to
be drawn between human and brute psychology with respect to improvement
from generation to generation. Indeed, I should deem it almost more
philosophical in any opponent of the theory of evolution, who happened
to be acquainted with the facts bearing upon the subject, if he were to
adopt the converse position, and argue that for the purposes of this
theory there is _not a sufficient_ distinction between human and brute
psychology in this respect. For when we remember the great advance
which, according to the theory of evolution, the mind of palæolithic
man must already have made upon that of the higher apes, and when we
remember that all races of existing men have the immense advantage of
some form of language whereby to transmit to progeny the results of
individual experience,—when we remember these things, the difficulty
appears to me to lie on the side of explaining why, with such a start
and with such advantages, the human species, both when it first appears
upon the pages of geological history, and as it now appears in the
great majority of its constituent races, should so far resemble animal
species in the prolonged stagnation of its intellectual life.

I shall now pass on to consider the views of Mr. Wallace and Mr.
Mivart on the distinction between the mental endowments of man and of
brute. Both these authors are skilled naturalists, and also professed
evolutionists so far as the animal world is concerned: moreover, they
further agree in maintaining that the principles of evolution cannot
be held to apply to man. But it is curious that, so far as psychology
is concerned, they base their arguments in support of their common
conclusion on precisely opposite premisses. For while Mr. Mivart
argues that human intelligence cannot be the same in kind as animal
intelligence, because the mind of the lowest savage is incomparably
superior to that of the highest ape; Mr. Wallace argues for the same
conclusion on the ground that the intelligence of savages is so little
removed from that of the higher apes, that the fact of their brains
being proportionately larger must be held to point prospectively
towards the needs of civilized life. “A brain,” he says, “slightly
larger than that of the gorilla would, according to the evidence before
us, fully have sufficed for the limited mental development of the
savage; and we must therefore admit that the large brain he actually
possesses could never have been developed solely by any of the laws of
evolution.”[10]

Now, I have presented these two opinions side by side because I deem it
an interesting, if not a suggestive circumstance, that the two leading
dissenters in this country from the general school of evolutionists,
although both holding the doctrine that man ought to be separated
from the rest of the animal kingdom on psychological grounds, are
nevertheless led to their common doctrine by directly opposite reasons.

The eminent French naturalist, Professor Quatrefages, also adopts
the opinion that man should be separated from the rest of the animal
kingdom as a being who, on psychological grounds, must be held to
have had some different mode of origin. But he differs from both
the English evolutionists in drawing his distinction somewhat more
finely. For while Mivart and Wallace found their arguments upon the
mind of man considered as a whole, Quatrefages expressly limits his
ground to the faculties of conscience and religion. In other words, he
allows—nay insists—that no valid distinction between man and brute
can be drawn in respect of rationality or intellect. For instance, to
take only one passage from his writings, he remarks:—“In the name of
philosophy and psychology, I shall be accused of confounding certain
intellectual attributes of the human reason with the exclusively
sensitive faculties of animals. I shall presently endeavour to answer
this criticism from the standpoint which should never be quitted by
the naturalist, that, namely, of experiment and observation. I shall
here confine myself to saying that, in my opinion, the animal is
intelligent, and, although an (intellectually) rudimentary being, that
its intelligence is nevertheless of the same nature as that of man.”
Later on he says:—“Psychologists attribute religion and morality to
the reason, and make the latter an attribute of man (to the exclusion
of animals). But with the reason they connect the highest phenomena of
the intelligence. In my opinion, in so doing they confound, and refer
to a common origin, facts entirely different. Thus, since they are
unable to recognize either morality or religion in animals, which in
reality do not possess these two faculties, they are forced to refuse
them intelligence also, although the same animals, in my opinion, give
decisive proof of their possession of this faculty every moment.”[11]

Touching these views I have only two things to observe. In the first
place, they differ _toto cælo_ from those both of Mr. Wallace and Mr.
Mivart; and thus we now find that the _three_ principal authorities
who still stand out for a distinction of kind between man and brute
on grounds of psychology, far from being in agreement, are really in
fundamental opposition, seeing that they base their common conclusion
on premisses which are all mutually exclusive of one another. In the
next place, even if we were fully to agree with the opinion of the
French anthropologist, or hold that a distinction of kind has to be
drawn only at religion and morality, we should still be obliged to
allow—although this is a point which he does not himself appear
to have perceived—that the superiority of human intelligence is a
necessary _condition_ to both these attributes of the human mind. In
other words, whether or not Quatrefages is right in his view that
religion and morality betoken a difference of kind in the only animal
species which presents them, at least it is certain that neither of
these faculties could have occurred in that species, had it not also
been gifted with a greatly superior order of intelligence. For even the
most elementary forms of religion and morality depend upon ideas of a
much more abstract, or intellectual, nature than are to be met with
in any brute. Obviously, therefore, the first distinction that falls
to be considered is the intellectual distinction. If analysis should
show that the school represented by Quatrefages is right in regarding
this distinction as one of degree—and, therefore, that the school
represented by Mivart is wrong in regarding it as one of kind,—the
time will then have arrived to consider, in the same connection, these
special faculties of morality and religion. Such, therefore, is the
method that I intend to adopt. The whole of the present volume will
be devoted to a consideration of “the origin of human faculty” in
the larger sense of this term, or in accordance with the view that
distinctively human faculty begins with distinctively human ideation.
When this matter has been thoroughly discussed, the ground will have
been prepared for considering in subsequent volumes the more special
faculties of Morality and Religion.[12]




CHAPTER II.

IDEAS.[13]


I now pass on to consider the only distinction which in my opinion
can be properly drawn between human and brute psychology. This is the
great distinction which furnishes a full psychological explanation of
all the many and immense differences that unquestionably do obtain
between the mind of the highest ape and the mind of the lowest savage.
It is, moreover, the distinction which is now universally recognized
by psychologists of every school, from the Romanist to the agnostic in
Religion, and from the idealist to the materialist in Philosophy.

The distinction has been clearly enunciated by many writers, from
Aristotle downwards, but I may best render it in the words of Locke:—

“If it may be doubted, whether beasts compound and enlarge their ideas
that way to any degree; this I think I may be positive in, that the
power of abstracting is not at all in them; and that the having of
general ideas is that which puts a perfect distinction betwixt man and
brutes, and is an excellency which the faculties of brutes do by no
means attain to. For it is evident we observe no footsteps in them of
making use of general signs for universal ideas; from which we have
reason to imagine, that they have not the faculty of abstracting, or
making general ideas, since they have no use of words, or any other
general signs.

“Nor can it be imputed to their want of fit organs to frame articulate
sounds that they have no use or knowledge of general words; since
many of them, we find, can fashion such sounds, and pronounce words
distinctly enough, but never with any such application; and, on the
other side, men, who through some defect in the organs want words,
yet fail not to express their universal ideas by signs, which serve
them instead of general words; a faculty which we see beasts come
short in. And therefore I think we may suppose, that it is in this
that the species of brutes are discriminated from men; and it is that
proper difference wherein they are wholly separated, and which at
last widens to so vast a distance; for if they have any ideas at all,
and are not bare machines (as some would have them), we cannot deny
them to have some reason. It seems evident to me, that they do some
of them in certain instances reason, as that they have sense; but it
is only in particular ideas, just as they received them from their
senses. They are the best of them tied up within those narrow bounds,
and have not (as I think) the faculty to enlarge them by any kind of
abstraction.”[14]

Here, then, we have stated, with all the common-sense lucidity of this
great writer, what we may term the initial or basal distinction of
which we are in search: it is that “proper difference” which, narrow
at first as the space included between two lines of rails at their
point of divergence, “at last widens to so vast a distance” as to end
almost at the opposite poles of mind. For, by a continuous advance
along the same line of development, the human mind is enabled to think
about abstractions of its own making, which are more and more remote
from the sensuous perception of concrete objects; it can unite these
abstractions into an endless variety of ideal combinations; these, in
turn, may become elaborated into ideal constructions of a more and
more complex character; and so on until we arrive at the full powers
of introspective thought with which we are each one of us directly
cognisant.

       *       *       *       *       *

We now approach what is at once a matter of refined analysis, and a
set of questions which are of fundamental importance to the whole
superstructure of the present work. I mean the nature of abstraction,
and the classification of ideas. No small amount of ambiguity still
hangs about these important subjects, and in treating of them it is
impossible to employ terms the meanings of which are agreed upon by all
psychologists. But I will carefully define the meanings which I attach
to these terms myself, and which I think are the meanings that they
ought to bear. Moreover, I will end by adopting a classification which
is to some extent novel, and by fully giving my reasons for so doing.

Psychologists are agreed that what they call particular ideas, or
ideas of particular objects, are of the nature of mental images,
or memories of such objects—as when the sound of a friend’s voice
brings before my mind the idea of that particular man. Psychologists
are further agreed that what they term general ideas arise out of an
assemblage of particular ideas, as when from my repeated observation
of numerous individual men I form the idea of Man, or of an abstract
being who comprises the resemblances between all these individual men,
without regard to their individual differences. Hence, particular
ideas answer to percepts, while general ideas answer to concepts: an
individual preception (or its repetition) gives rise to its mnemonic
equivalent as a particular idea; while a group of similar, though not
altogether similar perceptions, gives rise to its mnemonic equivalent
as a conception, which, therefore, is but another name for a general
idea, thus _generated_ by an assemblage of particular ideas. Just as
Mr. Galton’s method of superimposing on the same sensitive plate a
number of individual images gives rise to a blended photograph, wherein
each of the individual constituents is partially and proportionally
represented; so in the sensitive tablet of memory, numerous images of
previous perceptions are fused together into a single conception, which
then stands as a composite picture, or class-representation, of these
its constituent images. Moreover, in the case of a sensitive plate it
is only those particular images which present more or less numerous
points of resemblance that admit of being thus blended into a distinct
photograph; and so in the case of the mind, it is only those particular
ideas which admit of being run together in a class that can go to
constitute a clear concept.[15]

So much, then, for ideas as particular and general. Next, the term
abstract has been used by different psychologists in different
senses. For my own part, I will adhere to the usage of Locke in the
passage above quoted, which is the usage adopted by the majority of
modern writers upon these subjects. According to this usage, the term
“abstract idea” is practically synonymous with the term “general
idea.” For the process of abstraction consists in mentally analysing
the complex which is presented by any given object of perception, and
ideally extracting those features or qualities upon which the attention
is for the time being directed. Even the most individual of objects
cannot fail to present an assemblage of qualities, and although it is
true that such an object could not be divided into all its constituent
qualities actually, it does admit of being so divided ideally. The
individual man whom I know as John Smith could not be disintegrated
into so much heat, flesh, bone, blood, colour, &c., without ceasing to
be a man at all; but this does not hinder that I may ideally abstract
his heat (by thinking of him as a corpse), his flesh, bones, and blood
(by thinking of him as a dissected “subject”), his white colour of
skin, his black colour of hair, and so forth. Now, it is evident that
in the last resort our power of forming general ideas, or concepts,
is dependent on this power of abstraction, or the power of ideally
separating one or more of the qualities presented by percepts, _i.e._
by objects of particular ideas. My general idea of heat has only been
rendered possible on account of my having ideally abstracted the
quality of heat from sundry heated bodies, in most of which it has
co-existed with numberless different associations of other qualities.
But this does not hinder that, wherever I meet with that one quality, I
recognize it as the same; and hence I arrive at a general or abstract
idea of heat, apart from any other quality with which in particular
cases it may happen to be associated.[16]

This faculty of ideal abstraction furnishes the _conditio sine quâ
non_ to all grades in the development of thought; for by it alone
can we compare idea with idea, and thus reach ever onwards to higher
and higher levels, as well as to more and more complex structures
of ideation. As to the history of this development we shall have
more to say presently. Meanwhile I desire only to remark two things
in connection with it. The first is that throughout this history
the development is a _development_: the faculty of abstraction is
everywhere the same in _kind_. And the next thing is that this
development is everywhere dependent on the faculty of _language_. A
great deal will require to be said on both these points in subsequent
chapters; but it is needful to state the facts thus early—and they
are facts which psychologists of all schools now accept,—in order
to render intelligible the next step which I am about to make in my
classification of ideas. This step is to distinguish between the
faculty of abstraction where it is not dependent upon language,
and where it is so dependent. I have just said that the faculty of
abstraction is _everywhere_ the same in kind; but, as I immediately
proceeded to affirm that the _development_ of abstraction is dependent
upon language, I have thus far left the question open whether or not
there can be any rudimentary abstraction without language. It is to
this question, therefore, that we must next address ourselves.

On the one hand it may be argued that by restricting the term abstract
to ideas which can only be formed by the aid of language, we are
drawing an arbitrary line—fixing upon one degree in the continuous
scale of a faculty which is throughout the same in kind. For, say some
psychologists, it is evident that in our own case most of our more
simple abstract or general ideas are not dependent for their existence
upon words. Or, if this be disputed, these psychologists are able to
point to infants, and even to the lower animals, in proof of their
assertion. For an infant undoubtedly exhibits the possession of simple
general ideas prior to the possession of any articulate language;
and after it begins to use such language it does so by spontaneously
widening the generality of signification attaching to its original
words. In proof of both these statements numberless observations might
be quoted, and further on will be quoted; but here I need only wait
to give one in proof of each. As regards the first, Professor Preyer
tells us that at eight months old,[17] and therefore long before it
was able to speak, his child was able to classify all glass bottles
as resembling—or belonging to the order of—a feeding-bottle.[18]
As regards the second, M. Taine tells us of a little girl eighteen
months old, who was amused by her mother hiding in play behind a piece
of furniture, and saying “Coucou.” Again, when her food was too hot,
when she went too near the fire or candle, and when the sun was warm,
she was told “Ça brûle.” One day, on seeing the sun disappear behind
a hill, she exclaimed, “‘A b’ûle coucou,” thereby showing both the
formation and combination of general ideas, “not only expressed by
words which we do not employ (and, therefore, not by any other words
that she can have previously employed), but also corresponding to
ideas, _consequently to classes of objects and general characters_
which in our cases have disappeared. The hot soup, the fire on the
hearth, the flame of the candle, the noonday heat in the garden, and
last of all, the sun, make up one of these classes. The figure of the
nurse or mother disappearing behind a hill, form the other class.”[19]

Coming next to the case of brutes, and to begin with the simplest
kind of illustrations, all the higher animals have general ideas of
“Good-for-eating,” and “Not-good-for-eating,” quite apart from any
particular objects of which either of these qualities happens to be
characteristic. For, if we give any of the higher animals a morsel of
food of a kind which it has never before met with, the animal does
not immediately snap it up, nor does it immediately reject our offer;
but it subjects the morsel to a careful examination before consigning
it to the mouth. This proves, if anything can, that such an animal
has a general or abstract idea of sweet, bitter, hot, or, in general,
Good-for-eating and Not-good-for-eating—the motives of the examination
clearly being to ascertain which of these two general ideas of kind
is appropriate to the particular object examined. When we ourselves
select something which we suppose will prove good to eat, we do not
require to call to our aid any of that higher class of abstract ideas
for which we are indebted to our powers of language: it is enough to
determine our decision if the particular appearance, smell, or taste of
the food makes us feel that it probably conforms to our general idea
of Good-for-eating. And, therefore, when we see animals determining
between similar alternatives by precisely similar methods, we cannot
reasonably doubt that the psychological processes are similar; for,
as we know that these processes in ourselves do not involve any of
the higher powers of our minds, there is no reason to doubt that the
processes, which in their manifestations appear so similar, really
are what they appear to be—the same. Again, if I see a fox prowling
about a farm-yard, I infer that he has been led by hunger to go where
he has a general idea that there are a good many eatable things to be
fallen in with—just as I myself am led by a similar impulse to visit
a restaurant. Similarly, if I say to my dog the word “Cat,” I arouse
in his mind an idea, not of any cat in particular—for he sees so many
cats,—but of a Cat in general. Or when this same dog accidentally
crosses the track of a strange dog, the scent of this strange dog makes
him stiffen his tail and erect the hair on his back in preparation for
a fight; yet the scent of an unknown dog must arouse in his mind, not
the idea of any dog in particular, but an idea of the animal Dog in
general.

Thus far, it will be remembered, I have been presenting evidence in
favour of the view that both infants and animals show themselves
capable of forming general ideas of a simple order, and, therefore,
that to the formation of such ideas the use of language is not
essential. I will next consider what has to be said on the other
side of the question; for, as previously remarked, many—I may say
most—psychologists repudiate this kind of evidence _in toto_, as not
germain to the subject of debate. First, therefore, I will consider
their objections to this kind of evidence; next I will sum up the whole
question; and, lastly, I will suggest a classification of ideas which
in my opinion ought to be accepted by both sides as constituting a
common ground of reconciliation.

To begin with another quotation from Locke, “How far brutes partake
in this faculty [_i.e._ that of comparing ideas] is not easy to
determine; I imagine they have it not in any great degree: for though
they probably have several ideas distinct enough, yet it seems to me
to be the prerogative of human understanding, when it has sufficiently
distinguished any ideas, so as to perceive them to be perfectly
different, and so consequently two, to cast about and consider in what
circumstances they are capable to be compared: and therefore I think
beasts compare not their ideas further than some sensible circumstances
annexed to the objects themselves. The other power of comparing, which
may be observed in men, belonging to general ideas, and useful only to
abstract reasonings, we may probably conjecture beasts have not.

“The next operation we may observe in the mind about its ideas, is
composition; whereby it puts together several of those simple ones it
has received from sensation and reflection, and combines them into
complex ones. Under this head of composition may be reckoned also that
of enlarging; wherein, though the composition does not so much appear
as in more complex ones, yet it is nevertheless a putting several
ideas together, though of the same kind. Thus, by adding several units
together, we make the idea of a dozen; and by putting together the
repeated ideas of several perches, we frame that of a furlong.

“In this, also, I suppose, brutes come far short of men; for though
they take in, and retain together several combinations of simple ideas,
as possibly the shape, smell, and voice of his master make up the
complex idea a dog has of him, or rather are so many distinct marks
whereby he knows him; yet I do not think they do of themselves ever
compound them, and make complex ideas. And perhaps even where we think
they have complex ideas, it is only one simple one that directs them
in the knowledge of several things, which possibly they distinguish
less by sight than we imagine; for I have been credibly informed that
a bitch will nurse, play with, and be fond of young foxes, as much as,
and in place of, her puppies; if you can but get them once to suck her
so long, that her milk may go through them. And those animals, which
have a numerous brood of young ones at once, appear not to have any
knowledge of their number: for though they are mightily concerned for
any of their young that are taken from them whilst they are in sight
or hearing; yet if one or two be stolen from them in their absence, or
without noise, they appear not to miss them, or have any sense that
their number is lessened.”[20]

Now, from the whole of this passage, it is apparent that the
“comparing,” “compounding,” and “enlarging” of ideas which Locke has
in view, is the _conscious_ or _intentional_ comparing, compounding,
and enlarging that belongs only to the province of reflection, or
thought. He in no way concerns himself with such powers of “comparing
and compounding of ideas” as he allows that animals present, unless
it can be shown that animals are able to “cast about and consider
in what circumstances they are capable to be compared.” And then he
adds, “Therefore, I think, beasts compare not their ideas _further
than some sensible circumstances annexed to the objects themselves_.
The _other_ power of comparing, which may be observed in men,
_belonging to general ideas, and useful only to abstract reasonings_,
we may probably conjecture beasts have not.” So far, then, it seems
perfectly obvious that Locke believed animals to present the power
of “comparing and compounding” “simple ideas,” up to the point where
such comparison and composition begins to be assisted by the power of
reflective thought. Therefore, when he immediately afterwards proceeds
to explain abstraction thus: “The same colour being observed to-day
in chalk or snow, which the mind yesterday received from milk, it
considers that appearance alone, makes it a representative of all of
that kind; and having given it the name whiteness, it by that sound
signifies the same quality, wheresoever it be imagined or met with;
and thus universals, whether ideas or terms, are made”—when he thus
proceeds to explain abstraction, we can have no doubt that what he
means by abstraction is the power of _ideally contemplating qualities
as separated from objects_, or, as he expresses it, “_considering_
appearances alone.” Therefore I conclude, without further discussion,
that in the terminology of Locke the word abstraction is applied only
to those higher developments of the faculty which are rendered possible
by reflection.

Now, on what does this power of reflection depend? As we shall see
more fully later on, it depends on Language, or on the power of
affixing names to abstract and general ideas. So far as I am aware,
psychologists of all existing schools are in agreement upon this point,
or in holding that the power of affixing names to abstractions is at
once the condition to reflective thought, and the explanation of the
difference between man and brute in respect of ideation.

It seems needless to dwell upon a matter where all are agreed,
and concerning which a great deal more will require to be said in
subsequent chapters. At present I am only endeavouring to ascertain
the ground of difference between those psychologists who attribute,
and those who deny to animals the faculty of abstraction. And I think
I am now in a position to render this point perfectly clear. As we
have already seen, and we shall frequently see again, it is allowed
on all hands that animals in their ideation are not shut up to the
special imaging (or remembering) of particular perceptions; but that
they do present the power, as Locke phrases it, of “taking in and
retaining together several combinations of simple ideas.”[21] The
only question, then, really is whether or not this power is the power
of abstraction. In the opinion of some psychologists it is: in the
opinion of other psychologists it is not. Now, on what does an answer
to this question depend? Clearly it depends on whether we hold it
essential to an abstract or general idea that it should be incarnate
as a word. Under one point of view, to “take in and retain together
several combinations of simple ideas,” is to form a general concept of
so many percepts. But, under another point of view, such a combination
of simple ideas is only then entitled to be regarded as a concept, when
it has been conceived by the mind _as_ a concept, or when, in virtue
of having been bodied forth in a name, it stands before the mind as
a distinct and organic offspring of mind—so becoming an object as
well as a product of ideation. For then only can the abstract idea be
known _as_ abstract, and then only can it be available as a definite
creation of thought, capable of being built into any further and more
elaborate structure of ideation. Or, to quote M. Taine, who advocates
this view with great lucidity, “Of our numerous experiences [_i.e._
individual perceptions of a show of araucarias] there remain on the
following day four or five more or less distinct recollections, which
obliterated themselves, leave behind in us a simple, colourless,
vague representation, into which enter as components various reviving
sensations, in an utterly feeble, incomplete, and abortive state. But
this representation is not the general or abstract idea. It is but
its accompaniment, and, if I may say so, the one from which it is
extracted. For the representation, though badly sketched, is a sketch,
the sensible sketch of a distinct individual; in fact, if I make it
persist and dwell upon it, it repeats some special visual sensation;
I see mentally some outline which corresponds only to some particular
araucaria, and, therefore, cannot correspond to the whole class: now,
my abstract idea corresponds to the whole class; it differs, then, from
the representation of an individual. Moreover, my abstract idea is
perfectly clear and determinate; now that I possess it, I never fail
to recognize an araucaria among the various plants I may be shown; it
differs, then, from the confused and floating representation I have of
some particular araucaria. What is there, then, within me so clear and
determinate, corresponding to the abstract character, corresponding
to all araucarias, and corresponding to it alone? A class-name, the
name araucaria.... Thus we conceive the abstract characters of things
by means of abstract names which _are_ our abstract ideas, and the
formation of our abstract ideas is nothing more than the formation of
names.”[22]

The real issue, then, is as to what we are to understand by this term
abstraction, or its equivalents. If we are to limit the term to the
faculty of “taking in and retaining together several combinations of
simple ideas,” _plus_ the faculty of giving a name to the resulting
compound, then undoubtedly animals differ from men in not presenting
the faculty of abstraction; for this is no more than to say that
animals have not the faculty of speech. But if the term in question be
not thus limited—if it be taken to mean the first of the above-named
processes irrespective of the second,—then, no less undoubtedly,
animals resemble men in presenting the faculty of abstraction. In
accordance with the former definition, it necessarily follows that
“we conceive the abstract characters of things _by means of abstract
names which_ ARE _our abstract ideas_;” and, therefore, that “the
formation of our abstract ideas is nothing more than the formation of
names.” But, in accordance with the latter view, great as may be the
importance of affixing a name to a compound of simple ideas for the
purpose of giving that compound greater clearness and stability, the
essence of abstraction consists in the act of compounding, or in the
blending together of particular ideas into a general idea of the class
to which the individual things belong. The act of bestowing upon this
compound idea a class-name is quite a distinct act, and one which is
necessarily subsequent to the previous act of compounding: why then, it
may be asked, should we deny that such a compound idea is a general or
abstract idea, only because it is not followed up by the artifice of
giving it a name?

In my opinion so much has to be said in favour of both of these views
that I am not going to pronounce against either. What I have hitherto
been endeavouring to do is to reveal clearly that the question whether
or not there is any difference between the brute and the man in respect
of abstraction, is nothing more than a question of terminology. The
real question will arise only when we come to treat of the faculty
of language: the question before us now is merely a question of
psychological classification, or of the nomenclature of ideas. Now, it
appears to me that this question admits of being definitely settled,
and a great deal of needless misunderstanding removed, by a slight
re-adjustment and a closer definition of terms. For it must be on
all hands admitted that, whether or not we choose to denominate
by the word abstraction the faculty of compounding simple ideas
without the faculty of naming the compounds, at the place where this
additional faculty of naming supervenes, so immense an accession to
the previous faculty is furnished, that any system of psychological
nomenclature must be highly imperfect if it be destitute of terms
whereby to recognize the difference. For even if it were conceded by
psychologists of the opposite school that the essence of abstraction
consists in the compounding of simple ideas, and not at all in the
subsequent process of naming the compounds; still the effect of this
subsequent process—or additional faculty—is so prodigious, that
the higher degrees of abstraction which by it are rendered possible,
certainly require to be marked off, or to be distinguished from, the
lower degrees. Without, therefore, in any way prejudicing the question
as to whether we have here a difference of degree or a difference of
kind, I will submit a classification of ideas which, while not open to
objection from either side of this question, will greatly help us in
our subsequent treatment of the question itself.

The word “Idea” I will use in the sense defined in my previous
work—namely, as a generic term to signify indifferently any product of
imagination, from the mere memory of a sensuous impression up to the
result of the most abstruse generalization.[23]

By “Simple Idea,” “Particular Idea,” or “Concrete Idea,” I understand
the mere memory of a particular sensuous perception.

By “Compound Idea,” “Complex Idea,” or “Mixed Idea,” I understand the
combination of simple, particular, or concrete ideas into that kind of
composite idea which is possible without the aid of language.

Lastly, by “General Idea,” “Abstract Idea,” “Concept,” or “Notion,” I
understand that kind of composite idea which is rendered possible only
by the aid of language, or by the process of naming abstractions as
abstractions.

Now in this classification, notwithstanding that it is needful to
quote at least ten distinct terms which are either now in use among
psychologists or have been used by classical English writers upon
these topics, we may observe that there are really but three separate
classes to be distinguished. Moreover, it will be noticed that, for
the sake of definition, I restrict the first three terms to denote
memories of particular sensuous perceptions—refusing, therefore, to
apply them to those blended memories of many sensuous perceptions
which enable animals and infants (as well as ourselves) to form
compound ideas of kind or class without the aid of language. Again,
the first division of this threefold classification has to do only
with what are termed percepts, while the last has to do only with what
are termed concepts. Now there does not exist any equivalent word to
meet the middle division. And this fact in itself shows most forcibly
the state of ambiguous confusion into which the classification of
ideas has been wrought. Psychologists of both the schools that we are
considering—namely, those who maintain and those who deny that there
is any difference of kind between the ideation of men and animals—are
equally forced to allow that there is a great difference between what
I have called a simple idea and what I have called a compound idea. In
other words, it is a matter of obvious fact that the only distinction
between ideas is _not_ that between the memory of a particular percept
and the formation of a named concept; for between these two classes
of ideas there obviously lies another class, in virtue of which even
animals and infants are able to distinguish individual objects as
belonging to a sort or kind. Yet this large and important territory of
ideation, lying between the other two, is, so to speak, unnamed ground.
Even the words “compound idea,” “complex idea,” and “mixed idea,” are
by me restricted to it without the sanction of previous usage; for, as
above remarked, so completely has the existence of this intermediate
land been ignored, that we have no word at all which is applicable to
it in the same way that Percept and Concept are applicable to the
lands on either side of it. The consequence is that psychologists of
the one school invade this intermediate province of ideation with terms
that are applicable only to the lower province, while psychologists
of the other school invade it with terms which are applicable only to
the higher: the one matter upon which they all appear to agree being
that of ignoring the wide area which this intermediate territory
covers—and, consequently, also ignoring the great distance by which
the territories on either side of it are separated.

In addition, then, to the terms Percept and Concept, I coin the word
_Recept_. This is a term which seems exactly to meet the requirements
of the case. For as perception literally means a _taking wholly_, and
conception a _taking together_, reception means a _taking again_.
Consequently, a recept is that which is taken again, or a _recognition_
of things previously _cognized_. Now, it belongs to the essence of
what I have defined as compound ideas (recepts), that they arise in
the mind out of a repetition of more or less similar percepts. Having
seen a number of araucarias, the mind _receives_ from the whole mass
of individuals which it _perceives_ a composite idea of Araucaria,
or of a class comprising all individuals of that kind—an idea which
differs from a general or abstract idea only in not being consciously
fixed and signed as an idea by means of an abstract name. Compound
ideas, therefore, can only arise out of a _repetition_ of more or less
similar percepts; and hence the appropriateness of designating them
recepts. Moreover, the associations which we have with the cognate
words, Receive, Reception, &c., are all of the _passive_ kind, as the
associations which we have with the words Conceive, Conception, &c.,
are of the _active_ kind. Now, here again, the use of the word recept
is seen to be appropriate to the class of ideas in question, because
in receiving such ideas the mind is passive, as in conceiving abstract
ideas the mind is active. In order to form a concept, the mind must
intentionally bring together its percepts (or the memories of them),
for the purpose of binding them up as a bundle of similars, and
labelling the bundle with a name. But in order to form a recept, the
mind need perform no such intentional actions: the similarities among
the percepts with which alone this order of ideation is concerned, are
so marked, so conspicuous, and so frequently _repeated_ in observation,
that in the very moment of perception they sort themselves, and,
as it were, fall into their appropriate classes spontaneously, or
without any conscious effort on the part of the percipient. We do
not require to name stones to distinguish them from loaves, nor fish
to distinguish them from scorpions. Class distinctions of this kind
are conveyed in the very act of perception—_e.g._ the case of the
infant with the glass bottles,—and, as we shall subsequently see, in
the case of the higher animals admit of being carried to a wonderful
pitch of discriminative perfection. Recepts, then, are _spontaneous
associations, formed unintentionally_ as what may be termed
_unperceived abstractions_.[24]

One further remark remains to be added before our nomenclature of
ideas can be regarded as complete. It will have been noticed that
the term “general idea” is equally appropriate to ideas of class or
kind, whether or not such ideas are named. The ideas Good-for-eating
and Not-good-for-eating are as general to an animal as they are to a
man, and have in each case been formed in the same way—namely, by
an accumulation of particular experiences spontaneously assorted in
consciousness. General ideas of this kind, however, have not been
contemplated by previous writers while dealing with the psychology of
generalization: hence the term “general,” like the term “abstract,”
has by usage become restricted to those higher products of ideation
which depend on the faculty of language. And the only words that I can
find to have been used by any previous writers to designate the ideas
concerned in that lower kind of generalization which does not depend
on language, are the words above given—namely, Complex, Compound,
and Mixed. Now, none of these words are so good as the word General,
because none of them express the notion of _genus_ or _class_; and the
great distinction between the idea which an animal or an infant has,
say of an individual man and of men in general, is not that the one
idea is simple, and the other complex, compound, or mixed; but that the
one idea is _particular_ and the other _general_. Therefore consistency
would dictate that the term “general” should be applied to _all_
ideas of class or kind, as distinguished from ideas of particulars
or individuals—irrespective of the _degree_ of generality, and
irrespective, therefore, of the accident whether or not, _quâ_ general,
such ideas are dependent on language. Nevertheless, as the term has
been through previous usage restricted to ideas of the higher order of
generality, I will not introduce confusion by extending its use to the
lower order, or by speaking of an animal as capable of generalizing. A
parallel term, however, is needed; and, therefore, I will speak of the
general or class ideas which are formed without the aid of language as
_generic_. This word has the double advantage of retaining a verbal
as well as a substantial analogy with the allied term _general_. It
also serves to indicate that generic ideas, or recepts, are not only
ideas of class or kind, but have been _generated_ from the intermixture
of individual ideas—_i.e._ from the blended memories of particular
percepts.

My nomenclature of ideas, therefore, may be presented in a tabular form
thus:—

        { General, Abstract, or Notional = Concepts.
  IDEAS { Complex, Compound, or Mixed = Recepts, or Generic Ideas.
        { Simple, Particular, or Concrete = Memories of Percepts.[25]




CHAPTER III.

LOGIC OF RECEPTS.


We have seen that the great border-land, or _terra media_, lying
between particular ideas and general ideas has been strangely neglected
by psychologists, and we may now be prepared to find that a careful
exploration of this border-land is a matter of the highest importance
for the purposes of our inquiry. I will, therefore, devote the present
chapter to a full consideration of what I have termed generic ideas, or
recepts.

It has already been remarked that, in order to form any of these
generic ideas, the mind does not require to combine _intentionally_
the particular ideas which go to construct it: a recept differs from
a concept in that it is _received_, not _conceived_. The percepts
out of which a recept is composed are of so comparatively _simple_ a
character, are so frequently _repeated_ in observation, and present
among themselves resemblances or analogies so _obvious_, that the
mental images of them run together, as it were, spontaneously, or
in accordance with the primary laws of merely sensuous association,
without requiring any conscious act of comparison. This is a truth
which has been noticed by several previous writers. For instance, I
have in this connection already quoted a passage from M. Taine, and,
if necessary, could quote another, wherein he very aptly likens what
I have called recepts to the unelaborated ore out of which the metal
of a concept is afterwards smelted. And still more to the purpose
is the following passage, which I take from Mr. Sully:—“The more
_concrete_ concepts, or _generic_ images, are formed to a large extent
by a _passive_ process of _assimilation_. The likeness among dogs, for
example, is so great and striking that when a child, already familiar
with one of these animals, sees a second, he recognizes it as identical
with the first in certain obvious respects. The representation of the
first combines with the representation of the second, bringing into
distinct relief the common dog features, more particularly the canine
form. In this way the images of different dogs come to overlap, so to
speak, giving rise to a typical image of dog. Here there is very little
of _active_ direction of the mind from one thing to another in order
to discover where the resemblance lies: _the resemblance forces itself
upon the mind_. When, however, the resemblance is less striking, as in
the case of more abstract concepts, a _distinct operation of active
comparison is involved_.”[26]

Similarly, M. Perez remarks, “the necessity which children are under
of seeing in a detached and scrappy manner in order to see well, makes
them continually practise that kind of abstraction by which we separate
qualities from objects. From those objects which the child has already
distinguished as individual, there come to him at different moments
particularly vivid impressions.... Dominant sensations of this kind, by
their energy or frequency, tend to efface the idea of the objects from
which they proceed, _to separate or abstract themselves_.... The flame
of a candle is not always equally bright or flickering; tactile, sapid,
olfactory, and auditive impressions do not always strike the child’s
sensorium with the same intensity, nor during the same length of time.
This is why the recollections of individual forms, although strongly
graven on their intelligence, lose by degrees their first precision,
so that the idea of a tree, for instance, furnished by direct and
perfectly distinct memories, comes back to the mind in a vague and
indistinct form, which might be taken for a general idea.”[27]

Again, in the opinion of John Stuart Mill, “It is the doctrine of
one of the most fertile thinkers of modern times, Auguste Comte,
that besides the logic of signs, there is a logic of images, and a
logic of feelings. In many of the familiar processes of thought, and
especially in uncultured minds, a visual image serves instead of a
word. Our visual sensations, perhaps only because they are almost
always present along with the impressions of our other senses, have a
facility of becoming associated with them. Hence, the characteristic
visual appearance of an object easily gathers round it, by association,
the ideas of all other peculiarities which have, in frequent
experience, co-existed with that appearance; and, summoning up these
with a strength and certainty far surpassing that of merely casual
associations which it may also raise, it concentrates the attention
on them. This is an image serving for a sign—the logic of images.
The same function may be fulfilled by a feeling. Any strong and
highly interesting feeling, connected with one attribute of a group,
spontaneously classifies all objects according as they possess, or
do not possess, that attribute. We may be tolerably certain that the
things capable of satisfying hunger form a perfectly distinct class in
the mind of any of the more intelligent animals; quite as much as if
they were able to use or understand the word food. We here see in a
strong light the important truth that hardly anything universal can be
affirmed in psychology except the laws of association.”[28]

Furthermore, Mansel tersely conveys the truth which I am endeavouring
to present, thus:—“The mind recognizes the impression which a tree
makes on the retina of the eye: this is presentative consciousness. It
then depicts it. From many such pictures it forms a general notion, and
to that notion it at last appropriates a name.”[29] Almost in identical
language the same distinction is conveyed by Noiré thus:—“All trees
hitherto seen by me may leave in my imagination a mixed image, a kind
of ideal representation of trees. Quite different from this is the
concept, which is never an image.”[30]

And, not to overburden the argument with quotations, I will furnish
but one more, which serves if possible with still greater clearness
to convey exactly what it is that I mean by a recept. Professor
Huxley writes:—“An anatomist who occupies himself intently with the
examination of several specimens of some new kind of animal, in course
of time acquires so vivid a conception of its form and structure, that
the idea may take visible shape and become a sort of waking dream.”[31]

Although the use of the word “conception” here is unfortunate in one
way, I regard it as fortunate in another: it shows how desperate is the
need for the word which I have coined.

The above quotations, then, may be held sufficient to show that the
distinction which I have drawn has not been devised merely to suit my
own purposes. All that I have endeavoured so far to do is to bring
this distinction into greater clearness, by assigning to each of its
parts a separate name. And in doing this I have not assumed that the
two orders of generalization comprised under recepts and concepts are
the same in kind. So far I have left the question open as to whether
a mind which can only attain to recepts differs in degree or in kind
from the intellect which is able to go on to the formation of concepts.
Had I said, with Sully, “When the resemblance is less striking, as in
the case of more abstract concepts, a distinct operation of active
comparison is involved,” I should have been assuming that there is only
a difference of degree between a recept and a concept: designating
both by the same term, and therefore implying that they differ only in
their level of abstraction, I should have assumed that what he calls
the “passive process of assimilation,” whereby an infant or an animal
recognizes an individual man as belonging to a class, is really the
same kind of psychological process as that which is involved “in the
case of more abstract concepts,” where the individual man is designated
by a proper name, while the class to which he belongs is designated
by a common name. Similarly, if I had said, with Thomas Brown, that
in the process of generalization there is, “in the first place, the
perception of two or more objects [percept]; in the second place, the
feeling of their resemblance [recept]; and, lastly, the expression of
this common relative feeling by a name, afterwards used as a general
name [concept];”—if I had spoken thus, I should have virtually begged
the question as to the universal continuity of ideation, both in brutes
and men. Of course this is the conclusion towards which I am working;
but my endeavour in doing so is to proceed in the proof step by step,
without anywhere pre-judging my case. These passages, therefore, I
have quoted merely because they recognize more clearly than others
which I have happened to meet with what I conceive to be the true
psychological classification of ideas; and although, with the exception
of that quoted from Mill, no one of the passages shows that its writer
had before his mind the case of animal intelligence—or perceived the
immense importance of his statements in relation to the question which
we have to consider,—this only renders of more value their independent
testimony to the soundness of my classification.[32]

The question, then, which we have to consider is whether there is a
difference of kind, or only a difference of degree, between a recept
and a concept. This is really the question with which the whole of the
present volume will be concerned, and as its adequate treatment will
necessitate somewhat laborious inquiries in several directions, I will
endeavour to keep the various issues distinct by fully working out each
branch of the subject before entering upon the next.

First of all I will show, by means of illustrations, the highest levels
of ideation that are attained within the domain of recepts; and, in
order to do this, I will adduce my evidence from animals alone, seeing
that here there can be no suspicion—as there might be in the case of
infants—that the logic of recepts is assisted by any nascent growth
of concepts. But, before proceeding to state this evidence, it seems
desirable to say a few words on what I mean by the term just used,
namely, Logic of Recepts.

As argued in my previous work, all mental processes of an adaptive kind
are, in their last resort, processes of classification: they consist in
discriminating between differences and resemblances. An act of simple
perception is an act of noticing resemblances and differences between
the objects of such perception; and, similarly, an act of conception is
the taking together—or the intentional _putting_ together—of ideas
which are recognized as analogous. Hence abstraction has to do with
the abstracting of analogous qualities; reason is ratiocination, or
the comparison of ratios; and thus the highest operations of thought,
like the simplest acts of perception, are concerned with the grouping
or co-ordination of resemblances, previously distinguished from
differences.[33] Consequently, the middle ground of ideation, or the
territory occupied by recepts, is concerned with this same process on
a plane higher than that which is occupied by percepts, though lower
than that which is occupied by concepts. In short, the object or use,
and therefore the method or _logic_, of all ideation is the same. It
is, indeed, customary to restrict the latter term to the higher plane
of ideation, or to that which has to do with concepts. But, as Comte
has shown, there is no reason why, for purposes of special exposition,
this term should not be extended so as to embrace all operations of
the mind, in so far as these are operations of an orderly kind. For
in so far as they are orderly or adaptive—and not merely sentient or
indifferent—such operations all consist, as we have just seen, in
processes of ideal grouping, or _binding together_.[34] And therefore
I see no impropriety in using the word Logic for the special purpose
of emphasizing the fundamental identity of all ideation—so far, that
is, as its method is concerned. I object, however, to the terms “Logic
of Feelings” and “Logic of Signs.” For, on the one hand, “Feelings,”
have to do primarily with the sentient and emotional side of mental
life, as distinguished from the intellectual or ideational. And, on
the other hand, “Signs” are the _expressions_ of ideas; not the ideas
themselves. Hence, whatever method, or meaning, they may present is but
a reflection of the order, or grouping, among the ideas which they are
used to express. The logic, therefore, is neither in the feelings nor
in the signs; but in the ideas. On this account I have substituted for
the above terms what I take to be more accurate designations—namely,
the Logic of Recepts, and the Logic of Concepts.[35]

In the present chapter we have only to consider the logic of recepts,
and, in order to do so efficiently, we may first of all briefly
note that even within the region of percepts we meet with a process
of spontaneous grouping of like with like, which, in turn, leads
us downwards to the purely unconscious or mechanical grouping of
stimuli in the lower nerve-centres. So that, as fully argued out in
my previous work, on its objective face the method has everywhere
been the same: whether in the case of reflex action, of sensation,
perception, reception, conception, or reflection, on the side of the
nervous system, the method of evolution has been uniform: “it has
everywhere consisted in a progressive development of the power of
discriminating between stimuli, joined with the complementary power of
adaptive response.”[36] But although this is a most important truth
to recognize (as it appears to have been implicitly recognized—or,
rather, accidentally implied—by using a variant of the same term to
designate the lowest and the highest members of the above-named series
of faculties), for the purposes of psychological as distinguished from
physiological inquiry, it is convenient to disregard the objective side
of this continuous process, and therefore to take up our analysis at
the place where it is attended by a subjective counterpart—that is, at
Perception.

So much has already been written on what is termed the “unconscious
judgments” or “intuitive judgments” incidental to all our acts of
perception, that I feel it is needless to occupy space by dwelling at
any length upon this subject. The familiar illustration of looking
straight into a polished bowl, and alternately perceiving it as a
bowl and a sphere, is enough to show that here we _do_ have a logic
of feelings: without any act of ideation, but simply in virtue of
an automatic grouping of former percepts, the mind spontaneously
infers—or unconsciously judges—that an object, which _must either_
be a bowl or a sphere, is now one and now the other.[37] From which
we gather that all our visual perceptions are thus of the nature of
automatic inferences, based upon previous correspondencies between them
and perceptions of touch. From which, again, we gather that perceptions
of every kind depend upon previous grouping, whether between those
supplied by the same sense only, or also in combination with those
supplied by other senses.

Now, if this is so well known to be the case with percepts, obviously
it must also be the case with recepts. If we thus find by experiment
that all our perceptions are dependent on sub-conscious co-ordination
wholly automatic, much more may we be prepared to find that the
simplest of our ideas are dependent on spontaneous co-ordinations
almost equally automatic. Accordingly, it requires but a slight
analysis of our ordinary mental processes to prove that all our
simpler ideas are group-arrangements, which have been formed as I say
spontaneously, or without any of that intentionally comparing, sifting,
and combining process which is required in the higher departments of
ideational activity. The comparing, sifting, and combining is here
done, as it were, _for_ the conscious agent; not _by_ him. Recepts
are _received_: it is only concepts that require to be _conceived_.
For a recept is that kind of idea the constituent parts of which—be
they but the memories of percepts, or already more or less elaborated
as recepts—unite spontaneously as soon as they are brought together.
It matters not whether this readiness to unite is due to obvious
similarity, or to frequent repetition: the point is that there is
so strong an _affinity_ between the elementary constituents, that
the compound is formed as a consequence of their mere apposition in
consciousness. If I am crossing a street and hear behind me a sudden
shout, I do not require to wait in order to predicate to myself that
there is probably a hansom cab just about to run me down: a cry of
this kind, and in those circumstances, is so intimately associated
in my mind with its purpose, that the idea which it arouses need not
rise above the level of a recept; and the adaptive movements on my
part which that idea immediately prompts, are performed without any
intelligent reflection. Yet, on the other hand, they are neither reflex
actions nor instinctive actions: they are what may be termed receptual
actions, or actions depending on recepts.

This, of course, is an exceedingly simple illustration, and I have
used it in order to make the further remark that actions depending
on recepts, although they often thus lie near to reflex actions, are
by no means bound to do so. On the contrary, as we shall immediately
find, actions depending on recepts are often so highly “intelligent,”
that in our own case it is impossible to draw the line between them
and actions depending on concepts. That is to say, in our own case
there is a large border-land where introspection is unable to determine
whether adjustive action is due to recepts or to concepts; and hence
it is only in the case of animals that we can be certain as to the
limits of intelligent adjustment which are possible under the operation
of recepts alone. The question therefore, now arises,—How far can
this process of spontaneous or unintentional comparing, sifting, and
combining go without the intentional co-operation of the conscious
agent? To what level of ideation can recepts attain without the aid of
concepts? We have seen in the last chapter that animals display generic
or receptual ideas of Good-for-eating, Not-good-for-eating, &c.; and
we know that in our own case we “instinctively” avoid placing our
hands in a flame, without requiring to formulate any proposition upon
the properties of flame. How far, then, can this kind of unnamed or
non-conceptional ideation extend? Or, in other words, how far can mind
travel without the vehicle of Language? For the reasons already given,
I will answer this question by fastening attention exclusively on the
mind of brutes.

       *       *       *       *       *

To lead off with a few instances which have been already selected for
substantially the same purpose by Mr. Darwin:—

“Houzeau relates that, while crossing a wide and arid plain in Texas,
his two dogs suffered greatly from thirst, and that between thirty and
forty times they rushed down the hollows to search for water. These
hollows were not valleys, and there were no trees in them, or any other
difference in the vegetation; and as they were absolutely dry, there
could have been no smell of damp earth. The dogs behaved as if they
knew that a dip in the ground offered them the best chance of finding
water, and Houzeau has often witnessed the same behaviour in other
animals.”

I have myself frequently observed this association of ideas between
hollow ground and probability of finding water in the case of
setter-dogs, which require much water while working; and it is evident
that the ideas associated are of a character highly generic.

Further, Mr. Darwin writes:—“I have seen, as I dare say have others,
that when a small object is thrown on the ground beyond the reach of
one of the elephants in the Zoological Gardens, he blows through his
trunk on the ground beyond the object, so that the current reflected on
all sides may drive the object within his reach. Again, a well-known
ethnologist, Mr. Westropp, informs me that he observed in Vienna a bear
deliberately making with his paw a current in some water, which was
close to the bars of his cage, so as to draw a piece of floating bread
within his reach.”[38]

In _Animal Intelligence_ it will be seen that both these observations
are independently confirmed by letters which I have received from
correspondents; so that the facts must be accepted. And they imply a
faculty of forming generic ideas of a high order of complexity. Indeed,
these are not unlike the generic ideas of intelligent water-dogs
with reference to water-currents, which induce the animals to make
allowance for the force of the current by running in the opposite
direction to its flow before entering the water. Dogs accustomed to
tidal rivers, or to swimming in the sea, acquire a still further
generic idea of uncertainty as to the direction of the flow at any
given time; and therefore some of the more intelligent of these dogs
first ascertain the direction in which the tide is running by placing
their fore-paws in the stream, and then proceed to make their allowance
for driftway accordingly.[39]

Lastly, Mr. Darwin writes:—“When I say to my terrier in an eager voice
(and I have made the trial many times), ‘Hi, hi, where is it?’ she at
once takes it as a sign that something is to be hunted, and generally
first looks quickly all around, and then rushes into the nearest
thicket, to scout for any game, but finding nothing, she looks up into
any neighbouring tree for a squirrel. Now, do not these actions clearly
show that she had in her mind a general idea, or concept, that some
animal is to be discovered and hunted?”[40]

From the many instances which I have already given in _Animal
Intelligence_ of the high receptual capabilities of ants, it will here
be sufficient to re-state the following, which is quoted from Mr. Belt,
whose competency as an observer no one can dispute.

“A nest was made near one of our tramways, and to get to the trees the
ants had to cross the rails, over which the waggons were continually
passing and re-passing. Every time they came along a number of ants
were crushed to death. They persevered in crossing for some time, but
at last set to work and tunnelled underneath each rail. One day, when
the waggons were not running, I stopped up the tunnels with stones; but
although great numbers carrying leaves were thus cut off from the nest,
they would not cross the rails, but set to work making fresh tunnels
underneath them.”

These facts cannot be ascribed to “instinct,” seeing that tram-cars
could not have been objects of previous experience to the ancestors
of the ants; and therefore the degree of receptual intelligence, or
“practical inference,” which was displayed is highly remarkable.
Clearly, the insects must have appreciated the nature of these repeated
catastrophes, and correctly reasoned out the only way by which they
could be avoided.

As this is an important branch of my subject, I will add a few more
illustrations drawn from vertebrated animals, beginning with some
from the writings of Leroy, who had more opportunity than most men of
studying the habits of animals in a state of nature.[41]

He says of the wolf:—“When he scents a flock within its fold, memory
recalls to him the impression of the shepherd and his dog, and balances
that of the immediate neighbourhood of the sheep; he measures the
height of the fence, compares it with his own strength, takes into
account the additional difficulty of jumping it when burdened with his
prey, and thence concludes the uselessness of the attempt. Yet he will
seize one of a flock scattered over a field, under the very eyes of the
shepherd, especially if there be a wood near enough to offer him a hope
of shelter. He will resist the most tempting morsel when accompanied
by this alarming accessory [the smell of man]; and even when it is
divested of it, he is long in overcoming his suspicions. In this case
the wolf can only have an abstract idea of danger—the precise nature
of the trap laid for him being unknown.... Several nights are hardly
sufficient to give him confidence. Though the cause of his suspicions
may no longer exist, it is reproduced by memory, and the suspicion is
unremoved. The idea of man is connected with that of an unknown danger,
and makes him distrustful of the fairest appearances.”[42]

Leroy also well observes:—“Animals, like ourselves, are _forced_ to
make abstractions. A dog which has lost his master runs towards a group
of men, by virtue of a general abstract idea, which represents to him
the qualities possessed in common with these men by his master. He then
experiences in succession several less general, but still abstract
ideas of sensation, until he meets the particular sensation which he
seeks.”[43]

Again, with regard to the stag, this author writes:—“He exhausts every
variety and every design of which the action of flight consists. He
has perceived that in thickets, where the passage of his body leaves a
strong trace, the dogs follow him ardently, and without any checks; he
therefore leaves the thicket and plunges into the forests where there
is no underwood, or else skirts the high-road. Sometimes he leaves that
part of the country altogether, and depends wholly on his speed for
escape. But even when out of hearing of the dogs, he knows that they
will soon come up with him; and, instead of giving himself up to false
security, he avails himself of this respite to invent new artifices
to throw them out. He takes a straight course, returns on his steps,
and bounding from the earth many times consecutively, throws out the
sagacity of the dogs.... When hard pressed he will often drop down in
the hope that their ardour will carry them beyond the track, and should
it do so he retraces his steps. Often he seeks the company of others
of his species, and when his friend is sufficiently heated to share
the peril with, he leaves him to his fate and escapes by rapid flight.
Frequently the quarry is thus changed, and this artifice is one the
success of which is most certain.”[44]

“Often (when not being hunted at all), instead of returning home in
confidence and straightway lying down to rest, he will wander round the
spot; he enters the wood, leaves it, goes and returns on his steps
many times. Without having any immediate cause for his uneasiness, he
employs the same artifices which he would have employed to throw out
the dogs, if he were pursued by them. This foresight is an evidence of
remembered facts, and of a series of ideas and suppositions resulting
from those facts.”[45]

It is remarkable enough that an animal should seek to confuse its
trail by such devices, even when it knows that the hounds are actually
in pursuit; but it is still more so when the devices are resorted
to in order to confuse _imaginary_ hounds which may _possibly_ be
on the scent. Perhaps to some persons it may appear that such facts
argue on the part of the animals which exhibit them some powers of
representative thought, or some kind of reflection conducted without
the aid of language. Be it remembered, therefore, I am not maintaining
that they do not: I am merely conceding that the evidence is inadequate
to justify the conclusion that they do; and all I am now concerned with
is to make it certain that in animals there is a _logic_, be it a logic
of recepts only, or likewise what I shall afterwards explain as a logic
of _pre-concepts_.

Again, Leroy says of the fox:—“He smells the iron of the trap, and
this sensation has become so terrible to him, that it prevails over
every other. If he perceives that the snares become more numerous, he
departs to seek a safe neighbourhood. But sometimes, grown bold by a
nearer and oft-repeated examination, and guided by his unerring scent,
he manages, without hurt to himself, to draw the bait adroitly out of
the trap.... If all the outlets of his den are guarded by traps, the
animal scents them, recognizes them, and will suffer the most acute
hunger rather than attempt to pass them. I have known foxes keep their
dens a whole fortnight, and only then make up their minds to come out
because hunger left them no choice but as to the mode of death....
There is nothing he will not attempt in order to save himself. He will
dig till he has worn away his claws to effect his exit by a fresh
opening, and thus not unfrequently escapes the snares of the sportsman.
If a rabbit imprisoned with him gets caught in one of the snares, or if
by any other means one should go off, he infers that the machine has
done its duty, and walks boldly and securely over it.”[46]

Lastly, this author gives the case, which has since been largely
quoted—although its source is seldom given—of crows which it is
desired to shoot upon their nests, in order to destroy birds and eggs
at the same time. The crows will not return to their nests during
daylight, if they see any one waiting to shoot them. If, to lull
suspicion, a hut is made below the rookery and a man conceal himself
in it with a gun, he waits in vain if the bird has ever before been
shot at in a similar manner. “She knows that fire will issue from the
cave into which she saw a man enter.” Leroy then goes on to say:—“To
deceive this suspicious bird, the plan was hit upon of sending two men
into the watch-house, one of whom passed on while the other remained;
but the crow counted and kept her distance. The next day three went,
and again she perceived that only two returned. In fine it was found
necessary to send five or six men to the watch-house in order to put
her out of her calculation.”

Now, as Leroy is not a random writer, and as his life’s work was that
of Ranger at Versailles, we must not lightly set aside this statement
as incredible, more especially as he adds that the “phenomenon is
always to be repeated when the attempt is made,” and so is to be
regarded as “among the very commonest instances of the sagacity of
animals.”[47] If it is once granted that a bird has sagacity enough to
infer that where she has observed two men pass in and only one come
out, therefore the second man remains behind, it is only a matter
of degree how far the differential perception may extend. Of course
it would be absurd to suppose that the bird counts out the men by
any process of notation, but we know that for simple ideas of number
no symbolism in the way of figures is necessary. If we were to see
three men pass into a building and only two come out, we should not
require to calculate 3-2=1; the contrast between the simultaneous
sense-perception of A+B+C, when receptually compared with the
subsequently serial perceptions of A and B, would be sufficient for
the spontaneous inference that C must still be in the building. And
this process would in our own case continue possible up to the point at
which the simultaneous perception was not composed of too many parts to
be afterwards receptually analysed into its constituents.[48]

In this connection also I may state that, with the assistance of
the keeper, I have succeeded in teaching the Chimpanzee now at the
Zoological Gardens to count correctly as far as five. The method
adopted is to ask her for one straw, two straws, three straws, four
straws, or five straws—of course without observing any order in the
succession of such requests. If more than one straw is asked for, the
ape has been taught to hold the others in her mouth until the sum is
completed, so that she may deliver all the straws simultaneously. For
instance, if she is asked for four straws, she successively picks up
three straws and puts them in her mouth: then she picks up the fourth,
and hands over all the four together. This method prevents any possible
error arising from her interpretation of vocal tones, which might well
arise if each straw were asked for separately. Thus there can be no
doubt that the animal is able to distinguish receptually between the
numbers 1, 2, 3, 4, 5, and understands the name for each. Further than
this I have not attempted to take her. I may add that her performance
has been witnessed by the officers of the Zoological Society and also
by other naturalists, who will be satisfied with the accuracy of the
above account. But the ape is capricious, and, unless she happens to be
in a favourable mood at the time, visitors must not be disappointed if
they fail to be entertained by an exhibition of her learning.

The great physiologist Müller and the great philosopher Hegel are
quoted by Mr. Mivart as maintaining, that “to form abstract conceptions
of such operations as of something common to many under the notion of
cause and effect, is a perfect impossibility to them” (animals[49]);
and no doubt many other illustrious names might be quoted in support
of the same statement. But it seems to me that needless obscurity is
imported into this matter, by not considering in what our own idea of
causality consists. It is clear that to attain a _general_ idea of
causality as universal, &c., demands higher powers of abstract thought
than are possessed by any animals, or even by the great majority
of men; but it is no less clear that all men and most animals have
a _generic_ idea of causality, in the sense of expecting uniform
experience under uniform conditions. A cat sees a man knock at the
knocker of a door, and observes that the door is afterwards opened:
remembering this, when she herself wants to get in at that door,
she jumps at the knocker, and waits for the door to be opened.[50]
Now, can it be denied that in this act of inference, or imitation,
or whatever name we choose to call it, the cat perceives such an
association between the knocking and the opening as to feel that
the former as antecedent was in some way required to determine the
latter as consequent? And what is this but such a perception of causal
relation as is shown by a child who blows upon a watch to open the
case—thinking this to be the cause of the opening from the uniform
deception practised by its parent,—or of the savage who plants nails
and gunpowder to make them grow? And endless illustrations of such
a perception of causality might be drawn from the everyday life of
civilized man: indeed, how seldom does any one of us wait to construct
a general proposition about causality in the abstract before we act on
our practical knowledge of it. And that this practical knowledge in
the case of animals enables them to form a generic idea, or recept, of
the _equivalency_ between causes and effects—such that a perceived
equivalency is recognized by them as an _explanation_—would appear to
be rendered evident by the following fact, which I carefully observed
for the express purpose of testing the question. I quote the incident
from an already-published lecture, which was given before the British
Association at Dublin, in 1878.

“I had a setter dog which was greatly afraid of thunder. One day
a number of apples were being shot upon the wooden floor of an
apple-room, and, as each bag of apples was shot, it produced through
the rest of the house a noise resembling that of distant thunder. My
dog became terror-stricken at the sound; but as soon as I brought him
to the apple-room and showed him the true _cause_ of the noise, he
became again buoyant and cheerful as usual.”[51]

The importance of clearly perceiving that animals have a generic,
as distinguished from an abstract, idea of causation—and, indeed,
_must_ have such an idea if they are in any way at all to adjust their
actions to their circumstances—the importance of clearly perceiving
this is, that it carries with it a proof of the logic of recepts being
able to reach generic ideas of _principles_, as well as of objects,
qualities, and actions. In order to prove this important fact still
more unquestionably, I will here quote a passage from the biography
of the cebus which I kept for the express purpose of observing his
intelligence.

“To-day he obtained possession of a hearth-brush, one of the kind which
has the handle screwed into the brush. He soon found the way to unscrew
the handle, and, having done that, he immediately began to try to find
out the way to screw it in again. This he in time accomplished. At
first he put the wrong end of the handle into the hole, but turned it
round and round the right way for screwing. Finding it did not hold,
he turned the other end of the handle, carefully stuck it into the
hole, and began again to turn it the right way. It was, of course, a
very difficult feat for him to perform, for he required both his hands
to hold the handle in the proper position, and to turn it between his
hands in order to screw it in; and the long bristles of the brush
prevented it from remaining steady, or with the right side up. He held
the brush with his hind hands, but even so it was very difficult for
him to get the first turn of the screw to fit into the thread; he
worked at it, however, with the most unwearying perseverance until he
got the first turn of the screw to catch, and he then quickly turned it
round and round until it was screwed up to the end. The most remarkable
thing was that, however often he was disappointed in the beginning, he
never was induced to try turning the handle the wrong way; he always
screwed it from right to left. As soon as he had accomplished his wish,
he unscrewed it again, and then screwed it on again the second time
rather more easily than the first, and so on many times.”

The above is extracted from the diary kept by my sister. I did not
myself witness the progress of this research with the hearth-brush, as
I did so many of the other investigations successfully pursued by that
wonderful animal. But I have a perfect confidence in the accuracy of my
sister’s observation, as well as in the fidelity of her account; and,
moreover, the point with which I am about to be concerned has reference
to what followed subsequently, as to which I had abundant opportunities
for close and repeated observations. For the point is that, after
having thus discovered the mechanical _principle_ of the screw in that
one particular case, the monkey forthwith proceeded to _generalize_, or
to apply his newly gained knowledge to every other case where it was
at all probable that the mechanical principle in question was to be
met with. The consequence was that the animal became a nuisance in the
house by incessantly unscrewing the tops of fire-irons, bell-handles,
&c., &c., which he was by no means careful always to replace. Here,
therefore, I think we have unquestionable evidence of intelligent
recognition of a principle, which in the first instance was discovered
by “the most unwearying perseverance” in the way of experiment, and
afterwards sought for in multitudes of wholly dissimilar objects.[52]

To these numerous facts I will now add one other, which is sufficiently
remarkable to deserve republication for its own sake. I quote the
account from the journal _Science_, in which it appeared anonymously.
But finding on inquiry that the observer was Mr. S. P. Langley, the
well-known astronomer, and being personally assured by him that he is
certain there is no mistake about the observation, I will now give the
latter in his own words.

“The interesting description by Mr. Larkin (_Science_, No. 58) of the
lifting by a spider of a large beetle to its nest, reminds me of quite
another device by which I once saw a minute spider (hardly larger than
the head of a pin) lift a house-fly, which must have been more than
twenty times its weight, through a distance of over a foot. The fly
dangled by a single strand from the cross-bar of a window-sash, and,
when it first caught my attention, was being raised through successive
small distances of something like a tenth of an inch each; the lifts
following each other so fast, that the ascent seemed almost continuous.
It was evident that the weight must have been quite beyond the spider’s
power to stir by a ‘dead lift;’ but his motions were so quick, that
at first it was difficult to see how this apparently impossible task
was being accomplished. I shall have to resort to an illustration to
explain it; for the complexity of the scheme seems to belong less to
what we ordinarily call instinct than to intelligence, and that in a
degree we cannot all boast ourselves.

[Illustration]

“The little spider proceeded as follows:—

“_a b_ is a portion of the window-bar, to which level the fly was to
be lifted, from his original position at F vertically beneath _a_;
the spider’s first act was to descend halfway to the fly (to _d_),
and there fasten one end of an almost invisible thread; his second to
ascend to the bar and run out to _b_, where he made fast the other end,
and hauled on his guy with all his might. Evidently the previously
straight line must yield somewhat in the middle, whatever the weight
of the fly, who was, in fact, thereby brought into position F´, to the
right of the first one and a little higher. Beyond this point, it might
seem, he could not be lifted; but the guy being left fast at _b_, the
spider now went to an intermediate point _c_ directly over his victim’s
new position, and thus spun a new vertical line from _c_, which was
made fast at the bend at _d_´, after which _a d_ was cast off, so that
the fly now hung vertically below _c_, as before below _a_, but a
little higher.”

“The same operation was repeated again and again, a new guy being
occasionally spun, but the spider never descending more than about
halfway down the cord, whose elasticity was in no way involved in the
process. All was done with surprising rapidity. I watched it for some
five minutes (during which the fly was lifted perhaps six inches), and
then was called away.”

       *       *       *       *       *

Without further burdening the argument with illustrative proof, it
must now be evident that the “ore” out of which concepts are formed
is highly metalliferous: it is not merely a dull earth which bears no
resemblance to the shining substance smelted from it in the furnace
of Language; it is already sparkling to such an extent that we may
well feel there is no need of analysis to show it charged with that
substance in its pure form—that what we see in the ore is the same
kind of material as we take from the melting-pot, and differs from it
only in the degree of its agglomeration. Nevertheless, I will not yet
assume that such is the case. Before we can be perfectly sure that two
things which seem to the eye of common sense so similar are really
the same, we must submit them to a scientific analysis. Even though
it be certain that the one is extracted from the other, there still
remains a possibility that in the melting-pot some further ingredient
may have been added. Human intelligence is undoubtedly derived from
human experience, in the same way as animal intelligence is derived
from animal experience; but this does not prove that the ideation
which we have in common with brutes is not supplemented by ideation of
some other order, or kind. Presently I shall consider the arguments
which are adduced to prove that it has been, and then it will become
apparent that the supplement, if any, must have been added in the
smelting-fire of Language—a fact, be it observed, which is conceded by
all modern writers who deny the genetic continuity of mind in animal
and human intelligence. Thus far, then, I have attempted nothing more
than a preliminary clearing of the ground—first by carefully defining
my terms and impartially explaining the psychology of ideation; next
by indicating the nature of the question which has presently to be
considered; and, lastly, by showing the level to which intelligence
attains under the logic of recepts, without any possibility of
assistance from the logic of concepts.

       *       *       *       *       *

Only one other topic remains to be dealt with in the present chapter.
We continually find it assumed, and confidently stated as if the
statement did not admit of question, that the simplest or most
primitive order of ideation is that which is concerned only with
particulars, or with special objects of perception. The nascent ideas
of an infant are supposed to crystallize around the nuclei furnished
by individual percepts; the less intelligent animals—if not, indeed,
animals in general—are supposed, as Locke says, to deal “only in
particular ideas, just as they receive them from the senses.” Now, I
fully assent to this, if it is only meant (as I understand Locke to
mean) that infants and animals are not able consciously, intentionally,
or, as he says, “_of themselves_, to compound and make complex ideas.”
In order thus intentionally, or of themselves, to compound their
ideas, they would require to _think_ about their ideas _as_ ideas, or
consciously to set one idea before another as two distinct objects
of thought, _and for the known purpose of composition_. To do this
requires powers of introspective reflection; therefore it is a kind
of mental activity impossible to infants or animals, since it has to
do with concepts as distinguished from recepts. But, as we have now
so fully seen, it does not follow that because ideas cannot be thus
compounded by infants or animals _intentionally_, therefore they
cannot be compounded _at all_. Locke is very clear in recognizing that
animals do “take in and retain together several combinations of simple
ideas to make up a complex idea:” he only denies that animals “do _of
themselves_ ever compound them and make complex ideas.” Thus, Locke
plainly teaches my doctrine of recepts as distinguished from concepts;
and I do not think that any modern psychologist—more especially in
view of the foregoing evidence—will so far dispute this doctrine.
But the point now is that, in my opinion, many psychologists have
gone astray by assuming that the most primitive order of ideation is
concerned only with particulars, or that in chronological order the
memory of percepts precedes the occurrence of recepts. It appears
to me that a very little thought on the one hand, and a very little
observation on the other, is enough to make it certain that so soon as
ideas of any kind begin to be formed at all, they are formed, not only
as memories of particular percepts, but also as rudimentary recepts;
and that in the subsequent development of ideation the genesis of
recepts everywhere proceeds _pari passu_ with that of percepts. I say
that a very little thought is enough to show that this _must_ be so,
while a very little observation is enough to show that it _is_ so.
For, _a priori_, the more unformed the powers of perception, the less
able must they be to take cognizance of particulars. The development
of these powers consists in the ever-increasing efficiency of their
analysis, or _cognition_ of smaller and smaller differences of detail;
and, consequently, of their _recognition_ of these differences in
different combinations. Hence, the feebler the powers of perception,
the more must they occupy themselves with the larger or class
distinctions between objects of sensuous experience, and the less
with the smaller or more individual distinctions. Or, if we like,
what afterwards become class distinctions, are at earlier stages of
ideation the _only_ distinctions; and, therefore, all the same as
what are afterwards individual distinctions. But what follows? Surely
that—be it in the individual or the race—when these originally
individual distinctions begin to grow into class distinctions, they
leave in the mind an indelible impress of their first nativity: they
were the original recepts of memory, and if they are afterwards slowly
differentiated as they slowly become organized into many particular
parts, this does not hinder that throughout the process they never lose
their organic unity: the mind must always continue to recognize that
the parts which it subsequently perceived as successively unfolding
from what at first was known only as a whole, are parts which belong
to that whole—or, in other words, that the more newly observed
particulars are members of what is now perceived as a class. Therefore,
I say, on merely _a priori_ grounds we might banish the gratuitous
statement that the lower the order of ideation the more it is concerned
with particular distinctions, or the less with class distinctions. The
truth must be that the more primitive the recepts the larger are the
class distinctions with which they are concerned—provided, of course,
that this statement is not taken to apply beyond the region of sensuous
perception.

Accordingly we find, as a matter of fact, both in infants and in
animals, that the lower the grade of intelligence, the more is that
intelligence shut up to a perception of class distinctions. “We
pronounce the word _Papa_ before a child in its cradle, at the same
time pointing to his father. After a little, he in turn lisps the word,
and we imagine that he understands it in the same sense that we do, or
that his father’s presence only will recall the word. Not at all. When
another person—that is, one similar in appearance, with a long coat,
a beard, and loud voice—enters the room, he calls him also _Papa_.
The name was individual; he has made it general. In our case it is
applicable to one person only; in his, to a class.... A little boy,
a year old, had travelled a good deal by railway. The engine, with
its hissing sound and smoke, and the great noise of the train, struck
his attention, and the first word he learned to pronounce was _Fefer_
(chemin de fer). Then afterwards, a steam-boat, a coffee-pot with
spirit lamp—everything that hissed or smoked was a _Fefer_.”[53]

“Now, I have quoted such familiar instances from this author because
he adduces them as proof of the statement that here there appears a
delicacy of impression which is special to man.” Without waiting to
inquire whether this statement is justified by the evidence adduced,
or even whether the infant has personally distinguished his father
from among other men at the time when he first calls all men by the
same name; it is enough for my present purposes to observe the single
fact, that when a child is first able to show us the nature of its
ideation by means of speech, it furnishes us with ample evidence that
this ideation is what I have termed generic. The dress, the beard,
and the voice go to form a recept to which all men are perceived to
correspond: the most striking peculiarities of a locomotive are vividly
impressed upon the memory, so that when anything resembling them is
met with elsewhere, it is receptually classified as belonging to an
object of analogous character. Only much later, when the analytic
powers of perception have greatly developed, does the child begin
to draw its distinctions with sufficient “refinement” to perceive
that this classification is too crude—that the resemblances which
most struck its infant imagination were but accidental, and that
they have to be disregarded in favour of less striking resemblances
which were originally altogether unnoticed. But although the process
of classification is thus perpetually undergoing improvement
with advancing intelligence, from the very first it has been
_classification_—although, of course, thus far only within the region
of sensuous perception. And similarly with regard to animals, it is
sufficiently evident from such facts as those already instanced, that
the imagery on which their adaptive action depends is in large measure
generic.

Therefore, without in any way pre-judging the question as to whether
or not there is any radical distinction between a mind thus far gifted
and the conceptual thought of man, I may take it for granted that
the ideation of infants is from the first generic; and hence that
those psychologists are greatly mistaken who thoughtlessly assume
that the formation of class-ideas is a prerogative of more advanced
intelligence. No doubt their view of the matter seems plausible at
first sight, because within the region of conceptual thought we know
that progress is marked by increasing powers of _generalisation_—that
it is the easiest steps which have to do with the cognition of
particulars; the more difficult which have to do with abstractions.
But this is to confuse recepts with concepts, and so to overlook a
distinction between the two orders of generalization which it is of
the first importance to be clear about. A _generic_ idea is generic
because the particular ideas of which it is composed present such
_obvious_ points of resemblance that they spontaneously fuse together
in consciousness; but a _general_ idea is general for precisely the
opposite reason—namely, because the points of resemblance which it
has seized are _obscured_ from immediate perception, and therefore
could never have fused together in consciousness but for the aid of
intentional abstraction, or of the power of a mind knowingly to deal
with its own ideas as ideas. In other words, the kind of classification
with which recepts are concerned is that which lies nearest to the kind
of classification with which all processes of so-called “intuitive
inference” depend—such as mistaking a bowl for a sphere. But the kind
of classification with which concepts are concerned is that which
lies furthest from this purely automatic grouping of perceptions.
Classification there doubtless is in both cases; but the one order is
due to the closeness of resemblances in an act of perception, while in
the other order it is an expression of their remoteness from merely
perceptual associations.

Or, to put the matter in yet another light, if we think it sounds less
paradoxical to speak of the process of classification as everywhere
the same in kind, we must conclude that the groupings of recepts stand
to those of concepts in much the same relation as the groupings of
percepts do to those of recepts. In each case it is the lower order of
grouping which furnishes material for the higher: and the object of
this chapter has been to show, first, that the unintentional grouping
which is distinctive of recepts may be carried to a wonderful pitch
of perfection without any aid from the intentional grouping which is
distinctive of concepts; and, second, that from the very beginning
conscious ideation has been concerned with _grouping_. Not only, or
not even chiefly, has it had to do with the registration in memory
of particular percepts; but much more has it had to do with the
spontaneous sorting of such percepts, with the spontaneous arrangement
of them in ideal (or imagery) systems, and, consequently, with the
_spontaneous reflection in consciousness_ of many among the less
complex _relations_—or the less abstruse _principles_—which have
been uniformly encountered by the mind in its converse with an orderly
world.




CHAPTER IV.

LOGIC OF CONCEPTS.


The device of applying symbols to stand for ideas, and then using the
symbols as ideas, operates to the formation of more highly abstract
ideas in a manner that is easily seen. For instance, because we observe
that a great many objects present a certain quality in common, such as
redness, we find it convenient to give this quality a name; and, having
done so, we speak of redness in the abstract, or as standing apart
from any particular object. Our word “redness” then serves as a sign
or symbol of a quality, apart from any particular object of which it
may happen to be a quality; and having made this symbolic abstraction
in the case of a simple quality, such as redness, we can afterwards
compound it with other symbolic abstractions, and so on till we arrive
at verbal symbols of more and more abstract or general qualities,
as well as qualities further and further removed from immediate
perception. Thus, seeing that many other objects agree in being yellow,
others blue, and so on, we combine all these abstractions into a still
more general concept of Colour, which, _quâ_ more abstract, is further
removed from immediate perception—it being impossible that we can
ever have a percept answering to the amalgamated concept of _colour_,
although we have many percepts answering to the constituent concepts of
_colours_.

So in the analogous case of objects. The proper names Peter, Paul,
John, &c., stand in my mind as marks of my individual concepts:
the term Man serves to sum up all the points of agreement between
them—and also between all other individuals of their kind—without
regard to their points of disagreement: the word Animal takes a still
wider range, and so with nearly all words denoting objects. Like words
connoting qualities, they may be arranged in rank above rank according
to the range of their generality: and it is obvious that the wider this
range the further is their meaning withdrawn from anything that can
ever have been an object of immediate perception.

We shall afterwards find it is of the highest importance to note that
these remarks apply quite as much to actions and states as they do
to objects and qualities. Verbs, like nouns and adjectives, may be
merely the names of simple recepts, or they may be compounds of other
concepts—in either case differing from nouns and adjectives only in
that they have to do with actions and states. To sow, to dig, to spin,
&c., are names of particular actions; to labour is the name of a more
general action; to live is the symbol of a concept yet more general.
And it is obvious that here, as previously, the more general concepts
are built out of the more special.

Later on I will adduce evidence to show that, whether we look to
the growing infant or to the history of mankind as newly unearthed
by the researches of the philologist, we alike find that no one of
these divisions of simple concepts—namely, nouns, adjectives, and
verbs—appears to present priority over the others. Or, if there is any
evidence of such priority, it appears to incline in favour of nouns
and verbs. But the point on which I desire to fasten attention at
present is the enormous leverage which is furnished to the faculty of
ideation by thus using words as the mental equivalents of ideas. For by
the help of these symbols we climb into higher and higher regions of
abstraction: by thinking in verbal signs we think, as it were, with the
semblance of ideas: we dispense altogether with the necessity of actual
images, whether of precepts or of recepts: we quit the sphere of sense,
and rise to that of thought.

Take, for example, another type of abstract ideation, and one which
not only serves better than most to show the importance of signs as
substitutes for ideas, but also best illustrates the extraordinary
results to which such symbolism may lead when carried out persistently.
I refer to mathematics. Of course, before the idea of number or of
relation can arise at all, the faculty of conception must have made
great advances; but let us take this faculty at the point where the
artifice of substituting signs for ideas has gone as far as to enable
a mind to count by means of simple notation. It would clearly be
impossible to conduct the least intricate trains of reasoning which
invoke any ideas of number or proportion, were we deprived of the power
of attaching particular signs to particular ideas of number. We could
not even tell whether a clock had struck eleven or twelve, unless we
were able to mark off each successive stroke with some distinctive
sign; so that when it is said, as it often is, that an animal cannot
count, we must remember that neither could a senior wrangler count
if deprived of his symbols. “Man begins by counting things, grouping
them visibly [_i.e._ by the Logic of Recepts]. He then learns to count
simply the numbers, in the absence of things, using his fingers and
toes for symbols. He then substitutes abstract signs, and Arithmetic
begins. From this he passes to Algebra, the signs of which are not
merely abstract but general; and now he calculates numerical relations,
not numbers. From this he passes to the higher calculus of relations.”

And just as in mathematics the symbols that are employed contain in
an easily manipulated form enormous bodies of meaning—possibly,
indeed, the entire meaning of a long calculation,—so in all other
kinds of abstract ideation, the symbols which we employ—whether in
gesture, speech, or writing—contain more or less condensed masses of
signification. Or, to take another illustration, which, like the last
example, I quote from Lewes, “It is the same with the development of
commerce. Men begin by exchanging things. They pass to the exchange
of values. First money, then notes or bills, is the symbol of value.
Finally men simply debit and credit one another, so that immense
transactions are effected by means of this equation of equations. The
complicated processes of sowing, reaping, collecting, shipping, and
delivering a quantity of wheat, are condensed into the entry of a few
words in a ledger.”

Thus, without further treatment, it must be obvious that it is
impossible for us to over-estimate the importance of Language as the
handmaid of Thought. “A sign,” as Sir William Hamilton says, “is
necessary to give stability to our intellectual progress—to establish
each step in our advance as a new starting-point for our advance to
another beyond.... Words are the fortresses of thought. They enable us
to make every intellectual conquest the basis of operations for others
still beyond.” Moreover, thought and language act and react upon one
another; so that, to adopt a happy metaphor from Professor Max Müller,
the growth of thought and language is coral-like. Each shell is the
product of life, but becomes in turn the support of new life. In the
same manner each word is the product of thought, but becomes in turn a
new support for the growth of thought.

It seems needless to say more in order to show the immense importance
of sign-making to the development of ideation—the fact being one of
universal recognition by writers of every school. I will, therefore,
now pass on to the theme of the present chapter, which is that of
tracing in further detail the _logic_ of this faculty, or the _method_
of its development.

       *       *       *       *       *

From what I have already said, it may have been gathered that the
simplest concepts are merely the names of recepts; while concepts of a
higher order are the names of other concepts. Just as recepts may be
either memories of particular percepts, or the results of many percepts
(_i.e._ sundry other recepts) grouped as a class; so concepts may be
either names of particular recepts, or the results of many named
recepts (_i.e._ sundry other concepts) grouped as a class. The word
“red,” for example, is my name for a particular recept; but the word
“colour” is my name for a whole group of named recepts. And similarly
with words signifying objects, states, and actions. Hence, we may
broadly distinguish between concepts as of two orders—namely, those
which have to do with recepts, and those which have to do with other
concepts. For a concept is a concept even though it be nothing more
than a named recept; and it is still a concept, even though it stands
for the highest generalization of thought. I will make this distinction
yet more clear by means of better illustrations.

Water-fowl adopt a somewhat different mode of alighting upon land, or
even upon ice, from that which they adopt when alighting upon water;
and those kinds which dive from a height (such as terns and gannets)
never do so upon land or upon ice. These facts prove that the animals
have one recept answering to a solid substance, and another answering
to a fluid. Similarly, a man will not dive from a height over hard
ground or over ice, nor will he jump into water in the same way as
he jumps upon dry land. In other words, like the water-fowl, he has
two distinct recepts, one of which answers to solid ground, and the
other to an unresisting fluid. But, unlike the water-fowl, he is able
to bestow upon each of these recepts a name, and thus to raise them
both to the level of concepts. So far as the practical purposes of
locomotion are concerned, it is of course immaterial whether or not he
thus raises his recepts into concepts; but, as we have seen, for many
other purposes it is of the highest importance that he is able to do
this. Now, in order to do it, he must be able to set his recept before
his own mind as an object of his own thought: before he can bestow upon
these generic ideas the names of “solid” and “fluid,” he must have
_cognized_ them _as_ ideas. Prior to this act of cognition, these ideas
differed in no respect from the recepts of a water-fowl; neither for
the ordinary requirements of his locomotion is it needful that they
should: therefore, in so far as these requirements are concerned,
the man makes no call upon his higher faculties of ideation. But, in
virtue of this act of cognition, whereby he assigns a name to an idea
known as such, he has created for himself—and for purposes other than
locomotion—a priceless possession: he has formed a concept.

Nevertheless, the concept which he has formed is an extremely simple
one—amounting, in fact, to nothing more than the naming of one among
the most habitual of his recepts. But it is of the nature of concepts
that, when once formed, they admit of being intentionally compared;
and thus there arises a new possibility in the way of grouping
ideas—namely, no longer by means of sensuous associations, but by
means of symbolic representations. The names of recepts now serve
as symbols of the recepts themselves, and so admit of being grouped
without reference to the sensuous perceptions out of which they
originally sprang. No longer restricted to time, place, circumstance,
or occasion, ideas may now be called up and manipulated at pleasure;
for in this new method of ideation the mind has, as it were, acquired
an _algebra of recepts_: it is no longer necessary that the actual
recepts themselves should be present to sensuous perception, or even
to representative imagination. And as concepts are thus symbols of
recepts, they admit, as I have said, of being compared and combined
without reference to the recepts which they serve to symbolize. Thus we
become able, as it were, to calculate in concepts in a way and to an
extent that would be quite impossible in the merely perceptual medium
of recepts. Now, it is in this algebra of the imagination that all the
higher work of ideation is accomplished; and as the result of long and
elaborate syntheses of concepts we turn out mental products of enormous
intricacy—which, nevertheless, may be embodied in single words. Such
words, for example, as Virtue, Government, Mechanical Equivalent,
stand for immensely more elaborated concepts than the words Solid or
Fluid—seeing that to the former there are no possible equivalents in
the way of recepts.

Hence I say we must begin by recognizing the great reach of
intellectual territory which is covered by what are called concepts.
At the lowest level they are nothing more than named recepts; beyond
that level they become the names of other concepts; and eventually
they become the named products of the highest and most complex
co-ordinations of concepts which have been achieved by the human mind.
By the term _Lower Concepts_, then, I will understand those which are
nothing more than named recepts, while by the term _Higher Concepts_ I
will understand those which are compounded of other concepts.

       *       *       *       *       *

The next thing I wish to make clear is that concepts of the lower
order of which I speak, notwithstanding that they are the simplest
kind of concepts possible, are already something more than the names
of _particular_ ideas: they are the names of what I have called
_generic_ ideas, or recepts. We may search through the whole dictionary
of any language and not find a single word which stands as a name
for a truly particular idea—_i.e._ for the memory of a particular
percept. Proper names are those which most nearly approach this
character; but even proper names are really names of recepts (as
distinguished from particular percepts), seeing that every object to
which they are applied is a highly complex object, presenting many and
diverse qualities, all of which require to be registered in memory as
appertaining to that object if it is again to be recognized as the same.

Names, then, are not concerned with particular ideas, strictly so
called: concepts, even of the lowest order, have to do with generic
ideas. Furthermore, the generic ideas with which they have to do are
for the most part highly generic: even before a recept is old enough to
be baptized—or sufficiently far developed to be admitted as a member
of the body conceptual,—it is already a highly organized product
of ideation. We have seen in the last chapter how wonderfully far
the combining power of imagination is able to go without the aid of
language; and the consequence of this is, that before the advent of
language mind is already stored with a rich accumulation of orderly
ideas, grouped together in many systems of logical coherency. When,
therefore, the advent of language does take place, it is needless that
this work of logical grouping should be recommenced _ab initio_. What
language does is to take up the work of grouping where it has been left
by generic ideation; and if it is found expedient to name any generic
ideas, it is the more generic as well as the less generic that are
selected for the purpose. In short, immense as is the organizing power
of the Logos, it does not come upon the scene of its creative power to
find only that which is without form and void: rather does it find a
fair structure of no mean order of system, shaped by prior influences,
and, so far as thus shaped, a veritable cosmos.

Again, all concepts in their last resort depend on recepts, just as
in their turn recepts depend on percepts. This fact admits of being
abundantly proved, not only by general considerations, but also by the
etymological derivation of abstract terms. The most highly abstract
terms are derived from terms less abstract, and these from others
still less abstract, until, by two or three such steps at the most,
we are in all cases led directly back to their origin in a “lower
concept”—_i.e._ in the name of a recept. As I will prove later on,
there is no abstract word or general term in any language which, if its
origin admits of being traced at all, is not found to have its root
in the name of a recept. Concepts, therefore, are originally nothing
more than named recepts; and hence it is _a priori_ impossible that any
concept can be formed unless it does eventually rest upon the basis
of recepts. Owing to the elaboration which it subsequently undergoes
in the region of symbolism, it may, indeed, so far cease to bear any
likeness to its parentage that it is only the philologist who can
trace its lineage. When we speak of Virtue, we need no longer think
about a man, nor need we make any conscious reference to the steering
of a ship when we use the word Government. But it is none the less
obvious that both these highly abstract words have originated in the
naming of recepts (the one of an object, the other of an action); and
that their subsequent elevation in the scale of generality has been
due to a progressive widening of conceptual significance at the hands
of symbolical thought. In other words, and to revert to my previous
terminology, “higher concepts” can in no case originate _de novo_: they
can only be born of “lower concepts,” which, in turn, are the progeny
of recepts.

       *       *       *       *       *

I must now recur to a point with which we were concerned at the close
of the last chapter. I there showed that the kind of classification,
or mental grouping of ideas, which goes to constitute the logic of
recepts, differs from the mental grouping of ideas which constitutes
the logic of concepts, in that while the former has to do with
similarities which are most obvious to perception, and therefore with
analogies which most obtrude themselves upon attention, the latter
have to do with similarities which are least obvious to perception,
and therefore with analogies which are least readily apparent to the
senses. Classification there is in both cases; but while in the one it
depends on the closeness of the resemblances in an act of perception,
in the other it is expressive of their remoteness. Now, from this it
follows that the more conceptual the classification, the less obvious
to immediate perception are the similarities between the things
classified; and, consequently, the higher a generalization the greater
must be the distance by which it is removed from the merely automatic
groupings of receptual ideation.

For example, the earliest classification of the animal kingdom with
which we are acquainted, grouped together, under the common designation
of “creeping things,” articulata, mollusca, reptiles, amphibia, and
even certain mammals, such as weasels, &c. Here, it is evident, the
classification reposed only on the very superficial resemblances
which are exhibited by these various creatures in their modes of
locomotion. As yet conceptual thought had not been directed to the
anatomy of animals; and, therefore, when it undertook a classification
of animals, in the first instance it went no further than to note the
most obvious differences as to external form and movement. In other
words, this earliest conceptual classification was little more than the
verbal statement of a receptual classification. But when the science
of comparative anatomy was inaugurated by the Greeks, a much more
conceptual classification of animals emerged—although the importance
of anything like a systematic arrangement of the animal kingdom as
a whole was so little appreciated that it does not appear to have
been attempted, even by Aristotle. For, marvellous as is the advance
of conceptual grouping here displayed by him, he confined himself
to drawing anatomical comparisons between one group of animals and
another; he neither had any idea of group subordinate to group which
afterwards constituted the leading principle of taxonomic research, nor
does he anywhere give a tabular statement of his own results, such as
he could scarcely have failed to give had he appreciated the importance
of classifying the animal kingdom as a systematic whole. Lastly, since
the time of Ray the best thought of the best naturalists has been
bestowed upon this work, with the result that conceptual ideation has
continuously ascended through wider and wider generalizations, or
generalizations more and more chastened by the intentional and combined
accumulations of knowledge. How enormous, then, is the contrast between
the first simple attempt at classification as made by the early Jews,
and the elaborate body of abstract thought which is presented by the
taxonomic science of to-day.

Similar illustrations might be drawn from any of the other departments
of conceptual evolution, because everywhere such evolution essentially
consists in the achievement of ideal integrations further and further
removed from simple perceptions. Or, as Sir W. Hamilton puts it, “by a
first generalization we have obtained a number of classes of resembling
individuals. But these classes we can compare together, observe their
similarities, abstract from their differences, and bestow on their
common circumstance a common name. On the second classes we can again
perform the same operation, and thus, ascending through the scale of
general notions, throwing out of view always a greater number of
differences, and seizing always on fewer similarities in the formation
of our classes, we arrive at length at the limit of our ascent in the
notion of being or existence.”[54]

Now, the point on which I wish to be perfectly clear about is, that
this process of conceptual ideation, whereby ideas become general, must
be carefully distinguished from the processes of receptual ideation,
whereby ideas become generic. For these latter processes consist in
particular ideas, which are given immediately in sense perception,
becoming by association of similarity or contiguity automatically fused
together; so that out of a number of such associated percepts there is
formed a recept, without the need of any intentional co-operation of
the mind in the matter. On the other hand, a general idea, or concept,
can only be formed by the mind itself intentionally classifying its
recepts known as such—or, in the case of creating “higher concepts,”
performing the same process with its already acquired general ideas,
for the purpose of constructing ideas still more general. A generic
idea, then, is generalized in the sense that a naturalist speaks of a
lowly organism as generalized—_i.e._ as not yet differentiated into
the groups of higher and more specialized structures that subsequently
emanate therefrom. But a general idea is generalized in the sense of
comprising a group of such higher and more specialized structures,
already formed and named under a common designation with reference to
their points of resemblance. Classification there is in all cases; but
in the receptual order it is automatic, while in the conceptual order
it is introspective.

       *       *       *       *       *

So far as my analysis has hitherto gone, I do not anticipate criticism
or dissent from any psychologist, to whatever school he may belong. But
there is one matter of subordinate importance which I may here most
conveniently dispose of, although my views with regard to it may not
meet with universal assent.

It appears to me an obvious feature of our introspective life that we
are able to carry on elaborate processes of ideation without the aid
of words—or, to put it paradoxically, that we are able to conceive
without concepts. I am, of course, aware that this apparently obvious
power of being able to think without any mental rehearsal of verbal
signs (the _verbum mentale_ of scholasticism) is denied by several
writers of good standing—notably, for instance, by Professor Max
Müller, who seeks with much elaboration to prove that “not only to a
considerable extent, but always and altogether, we think by means of
names.”[55] Now this statement appears to me either a truism or untrue:
it is either tautological in expression, or erroneous in fact. If we
restrict the term “thought” to the operation of naming, it is merely a
truism to say that there can be no thought without language; for this
is merely to say that there can be no naming without names. But if
the term “thought” is taken to cover all processes of ideation which
we do not share with brutes, I hold that the statement is opposed to
obvious fact; and, therefore, I agree with the long array of logicians
and philosophers whom Professor Max Müller quotes as showing what he
calls “hesitation” in accepting a doctrine which in his opinion is
the inevitable conclusion of Nominalism. For to me it appears evident
that within the region of concepts, the frequent handling of those
with which the mind is familiar enables the mind to deal with them in
somewhat the same automatic manner as, on a lower plane of coordinated
action, the pianist deals with his chords and phrases. Whereas at first
it required intentional and laborious effort to perform these many
varied and complex adjustments, by practice their performance passes
more and more out of the range of conscious effort, until they come
to be executed in a manner well-nigh mechanical. So in the case of
purely mental operations, even of the highest order. At first every
link in the chain of ideation requires to be separately fastened to
attention by means of a word: every step in a process of reasoning
requires to be taken on the solid basis of a proposition. But by
frequent habit the thinking faculty ceases to be thus restricted: it
passes, so to speak, from one end of the chain to the other without
requiring to pause at every link: for its original stepping-stones it
has substituted a bridge, over which it can pass almost at a bound.
Or, again, to change the metaphor, there arises a method of short-hand
thinking, wherein even the symbols of ideas (concepts) need no longer
appear in consciousness: judgment follows judgment in logical sequence,
yet without any articulate expression by the _verbum mentale_. This, I
say, is a matter of fact which it appears to me a very small amount of
introspection is enough to verify. On reading a letter, for instance,
we may instantaneously decide upon our answer, and yet have to pause
before we are able to frame the propositions needed to express that
answer. Or, while writing an essay, how often does one feel, so to
speak, that a certain truth stands to be stated, although it is a truth
which we cannot immediately put into words. We know, in a general way,
that a truth is _there_, but we cannot supply the vehicle which is to
bring it _here_; and it is not until we have tried many devices, each
of which involve long trains of sequent propositions, that we begin
to find the satisfaction of rendering explicit in language what was
previously implicit in thought. Again, in playing a game of chess we
require to take cognizance of many and complex relations, actual and
contingent; so that to play the game as it deserves to be played, we
must make a heavy demand on our powers of abstract thinking. Yet in
doing this we do not require to preach a silent monologue as to all
that we might do, and all that may be done by our opponent. Lastly, to
give only one other illustration, in some forms of aphasia the patient
has lost every trace of verbal memory, and yet his faculties of thought
for all the practical purposes of life are not materially impaired.

On the whole, therefore, I conclude that, although language is a
needful condition to the _original construction_ of conceptional
thought, when once the building has been completed, the scaffolding
may be withdrawn, and yet leave the edifice as stable as before. In
this way familiar concepts become, as it were, degraded into recepts,
but recepts of a degree of complexity and organization which would not
have been possible but for their conceptional parentage. With Geiger we
may say, “So ist denn überall die Sprache primar, der Begriff entsteht
durch das Wort.”[56] Yet this does not hinder that with Friedrich
Müller we should add, “Sprechen ist nicht Denken, sondern es ist nur
Ausdruck des Denkens.”[57]

       *       *       *       *       *

With the exception of the last paragraph, my analysis, as already
observed, will probably not be impugned by any living psychologist,
either of the evolutionary or non-evolutionary schools; for, with the
exception of this paragraph, I have purposely arranged my argument so
as thus far to avoid debatable questions. And it will be observed that
even this paragraph has really nothing to do with the issue which lies
before us; seeing that the question with which it deals is concerned
only with intellectual processes exclusively human. But now, after
having thus fully prepared the way by a somewhat lengthy clearing of
preliminary ground, we have to proceed to the question whether it
is conceivable that the faculty of speech, with all the elaborate
structure of ideation to which it has led, can have arisen by way of a
natural genesis from the lower faculties of mind. As we have now seen,
it is on all hands agreed that the one and only distinction between
human and animal psychology consists in the former presenting this
faculty which, otherwise stated, means, as we have likewise seen, the
power of translating ideas into symbols, and using these symbols in the
stead of ideas.

This, I say, is the one distinction upon which all are agreed; the only
question is as to whether it is a distinction of kind or of degree.
Since the time when the ancient Greeks applied the same word to denote
the faculty of language and the faculty of thought, the philosophical
propriety of the identification has become more and more apparent.
Obscured as the truth may have become for a time through the fogs of
Realism, discussion of centuries has fully cleared the philosophical
atmosphere so far as this matter is concerned. Hence, in these latter
days, the only question here presented to the evolutionist is—Why has
no mere brute ever learnt to communicate with its fellows? Why has man
alone of animals been gifted with the Logos? To answer this question we
must undertake a somewhat laborious investigation of the philosophy of
Language.




CHAPTER V.

LANGUAGE.


Etymologically the word Language means sign-making by means of the
tongue, _i.e._ articulate speech. But in a wider sense the word is
habitually used to designate sign-making in general, as when we speak
of the “finger-language” of the deaf-and-dumb, the “language of
flowers,” &c. Or, as Professor Broca says, “there are several kinds
of language; every system of signs which gives expression to ideas
in a manner more or less intelligible, more or less perfect, or more
or less rapid, is a language in the general sense of the word. Thus
speech, gesture, dactylology, writing both hieroglyphic and phonetic,
are all so many kinds of language. There is, then, a general faculty of
language which presides over all these modes of expression, and which
may be defined—the faculty of establishing a constant relation between
an idea and a sign, be this a sound, a gesture, a figure, or a drawing
of any kind.”

The best classification of the sundry exhibitions of sign-making
faculty which I have met with, is one that is given by Mr. Mivart in
his _Lessons from Nature_ (p. 83). This classification, therefore, I
will render in his own words.

“We may altogether distinguish six different kinds of language:—

“1. Sounds which are neither articulate nor rational, such as cries of
pain, or the murmur of a mother to her infant.

“2. Sounds which are articulate but not rational, such as the talk of
parrots, or of certain idiots, who will repeat, without comprehending,
every phrase they hear.

“3. Sounds which are rational but not articulate, ejaculations by which
we sometimes express assent to, or dissent from, given propositions.

“4. Sounds which are both rational and articulate, constituting true
speech.

“5. Gestures which do not answer to rational conceptions, but are
merely the manifestations of emotions and feelings.

“6. Gestures which do answer to rational conceptions, and are therefore
‘external,’ but not oral manifestations of the _verbum mentale_.”

To this list of the “Categories of Language” a seventh must be added,
to contain all kinds of written signs; but with such obvious addition
I assent to the classification, as including all the species that
can possibly be included under the genus Language, and therefore as
excluding none.

Now the first thing to be noticed is, that the signs made may be made
either intentionally or unintentionally; and the next is, that the
division of intentional signs may be conveniently subdivided into two
classes—namely, intentional signs which are natural, and intentional
signs which are conventional.

The subdivision of conventional signs may further be split into
those which are due to past associations, and those which are due
to inferences from present experience. A dog which “begs” for food,
or a parrot which puts down its head to be scratched, may do so
merely because past experience has taught the animal that by so
doing it receives the gratification it desires; here is no need for
reason—_i.e._ inference—to come into play. But if the animal has had
no such previous experience, and therefore could not know by special
association that such a particular gesture, or sign, would lead to
such a particular consequence, and if under such circumstances a dog
should see another dog beg, and should imitate the gesture on observing
the result to which it led; or if under such analogous circumstances
a parrot should spontaneously depress its head for the purpose of
making an expressive gesture,—then the sign might strictly be termed a
rational one.

But it is evident that rational signs admit of almost numberless
degrees of complexity and elaboration; so that reason itself does
not present a greater variety of manifestations in this respect than
does the symbolism whereby it is expressed: an algebraical formula is
included in the same category of sign-making as the simplest gesture
whereby we intentionally communicate the simplest idea. Rational signs,
therefore, may be made by gesture, by tone, by articulation, or by
writing—using each of these words in its largest sense.[58]

The following schema may serve to show this classification in a
diagrammatic form—_i.e._ the classification which I have myself
arrived at, and which follows closely the one given by Mr. Mivart.
Indeed, there is no difference at all between the two, save that I have
endeavoured to express the distinction between signs as intentional,
unintentional, natural, conventional, emotional, and intellectual. The
subdivision of the latter into denotative, connotative, denominative,
and predicative, will be explained in Chapter VIII.


[Illustration: LANGUAGE, OR SIGN-MAKING.]

Or, neglecting the unintentional and merely initiative signs as not,
properly speaking, signs at all, every kind of intentional sign may be
represented diagrammatically as in the illustration opposite.

Now, thus far we have been dealing with matters of fact concerning
which I do not think there can be any question. That is to say, no one
can deny any of the statements which this schema serves to express;
a difference of opinion can only arise when it is asked whether the
sundry faculties (or cases) presented by the schema are developmentally
continuous with one another. To this topic, therefore, we shall now
address ourselves.

First let it be observed that there can be no dispute about one point,
namely, that all the faculties or cases presented by the schema, with
the single exception of the last (No. 7), are common to animals and
men. Therefore we may begin by taking as beyond the reach of question
the important fact that animals do present, in an unmistakable manner,
a _germ_ of the sign-making faculty. But this fact is so important in
its relation to our subject, that I shall here pause to consider the
modes and degrees in which the faculty is exhibited by animals.

       *       *       *       *       *

Huber says that when one wasp finds a store of honey, “it returns
to the nest and brings off in a short time a hundred other wasps;”
and this statement is confirmed by Dujardin. Again, the very able
observer, F. Müller, writes, in one of his letters to Mr. Darwin, that
he observed a queen bee depositing her eggs in a nest of 47 cells. In
the process she overlooked four of the cells, and when she had filled
the other 43, supposing her work to have been completed, prepared to
retire. “But as she had overlooked the four cells of the new comb, the
workers ran impatiently from this part to the queen, pushing her in an
odd manner with their heads, as they did also the other workers they
met with. In consequence, the queen began again to go round on the two
older combs; but, as she did not find any cell wanting an egg, she
tried to descend, yet everywhere she was pushed back by the workers.
This contest lasted rather a long while, till the queen escaped without
having completed her work. Thus the workers knew how to advise the
queen that something was yet to be done; but they knew not how to show
her where it had to be done.”

[Illustration]

According to De Fravière, Landois, and some other observers, bees
have a number of different notes, or tones, whereby they communicate
information to one another;[59] but there seems to be little doubt
that the means chiefly employed are gestures made with the antennæ.
For example, Huber divided a hive into two chambers by means of a
partition: great excitement prevailed in the half of the hive deprived
of the queen, and the bees set to work to build royal cells for the
creation of a new queen. Huber then divided a hive in exactly the same
manner, with the difference only that the screen, or partition, was
made of trellis work, through the openings of which the bees on either
side could pass their antennæ. Under these circumstances the bees in
the queenless half of the hive exhibited no disturbance, nor did they
construct any royal cells: the bees in the other, or separated, half of
the hive were able to inform them that the queen was safe.

Turning now to ants, the extent to which the power of communicating by
signs is here carried cannot fail to strike us as highly remarkable.
In my work on _Animal Intelligence_ I have given many observations by
different naturalists on this head, the general results of which I will
here render.

When we consider the high degree to which ants carry the principle
of co-operation, it is evident that they must have some means of
intercommunication. This is especially true of the Ecitons, which
so strangely mimic the tactics of military organization. “The army
marches in the form of a rather broad and regular column, hundreds of
yards in length. The object of the march is the capture and plunder of
other insects, &c., for food; and as the well-organized host advances,
its devastating legions set all other terrestrial life at defiance.
From the main column there are sent out smaller lateral columns, the
component individuals of which play the part of scouts, branching off
in various directions, and searching about with the utmost activity for
insects, grubs, &c., over every log, under every fallen leaf, and in
every nook and cranny where there is any chance of finding prey. When
their errand is completed, they return into the main column. If the
prey found is sufficiently small for the scouts themselves to manage,
it is immediately seized, and carried back to the main column; but if
the amount is too large for the scouts to deal with alone, messengers
are sent back to the main column, whence there is immediately
despatched a detachment large enough to cope with the requirements....
On either side of the main column there are constantly running up and
down a few individuals of smaller size and lighter colour than the
other ants, which seem to play the part of officers; for they never
leave their stations, and while running up and down the outsides of the
column, they every now and again stop to touch antennæ with some member
of the rank and file, as if to give instructions. When the scouts
discover a wasps’-nest in a tree, a strong force is sent out from the
main army, the nest is pulled to pieces, and all the larvæ carried to
the rear of the army, while the wasps fly around defenceless against
the invading multitude. Or, if the nest of any other species of ant
is found, a similarly strong force—or perhaps the whole army—is
deflected towards it, and with the utmost energy the innumerable
insects set to work to sink shafts and dig mines till the whole nest
is rifled of its contents. In these mining operations the ants work
with an extraordinary display of organized co-operation; for those low
down in the shafts do not lose time by carrying up the earth which they
excavate, but pass the pellets to those above; and the ants on the
surface, when they receive the pellets, carry them—with an appearance
of forethought which quite staggered Mr. Bates—only just far enough to
insure that they shall not roll back again into the shaft, and, after
depositing them, immediately hurry back for more. But there is not a
rigid (or merely mechanical) division of labour: the work seems to be
performed by intelligent co-operation amongst a host of eager little
creatures; for some of them act at one time as carriers of pellets, and
at another as miners, while all shortly afterwards assume the office of
conveyers of the spoil.”[60]

Mr. Belt writes:—“The Ecitons and most other ants follow each other by
scent, and I believe they can communicate the presence of danger, of
booty, or other intelligence to a distance by the different intensity
or qualities of the odours given off. I one day saw a column running
along the foot of a nearly perpendicular tramway cutting, the side
of which was about six feet high. At one point I noticed a sort of
assembly of about a dozen individuals that appeared in consultation.
Suddenly one ant left the conclave, and ran with great speed up the
perpendicular face of the cutting without stopping.... On gaining
the top of the cutting, the ants entered some brushwood suitable for
hunting. In a very short time the information was communicated to the
ants below, and a dense column rushed up in search of prey.”

Again, Mr. Bates writes:—“When I interfered with the column, or
abstracted an individual from it, news of the disturbance was quickly
communicated to a distance of several yards to the rear, and the
column at that point commenced retreating.”

On arriving at a stream of water, the marching column first endeavours
to find some natural bridge whereby to cross it. Should no such bridge
be found, “they travel along the bank of the river until they arrive
at a flat sandy shore. Each ant now seizes a bit of dry wood, pulls
it into the water and mounts thereon. The hinder rows push the front
ones farther out, holding on to the wood with their feet and to their
comrades with their jaws. In a short time the water is covered with
ants, and when the raft has grown too large to be held together by the
small creatures’ strength, a part breaks itself off, and begins the
journey across, while the ants left on the bank pull the bits of wood
into the water, and work at enlarging the ferry-boat until it breaks
again. This is repeated as long as an ant remains on shore.”[61]

So much, then, to give a general idea of the extent to which
co-operation is exhibited by Ecitons—a fact which must be taken to
depend upon some system of signs. Turning next to still more definite
evidence of communication, Mr. Hague, the geologist, writing to Mr.
Darwin from South America, says that on the mantel-shelf of his
sitting-room there were three vases habitually filled with fresh
flowers. A nest of red ants discovered these flowers, and formed a
line to them, constantly passing upwards and downwards between the
mantel-shelf and the floor, and also between the mantel-shelf and the
ceiling. For several days in succession Mr. Hague frequently brushed
the ants in great numbers from the wall to the floor, but, as they
were not killed, the line again reformed. One day, however, he killed
with his finger some of the ants upon the mantel-shelf. “The effect of
this was immediate and unexpected. As soon as those ants which were
approaching arrived near to where their fellows lay dead and suffering,
they turned and fled with all possible haste. In half an hour the wall
above the mantel-shelf was cleared of ants. During the space of an
hour or two the colony from below continued to ascend until reaching
the lower bevelled edge of the shelf, at which point the more timid
individuals, although unable to see the vase, somehow became aware of
the trouble, and turned without further investigation; while the more
daring advanced hesitatingly just to the upper edge of the shelf, when,
extending their antennæ and stretching their necks, they seemed to peep
cautiously over the edge until they beheld their suffering companions,
when they too turned and followed the others, expressing by their
behaviour great excitement and terror. An hour or two later the path
or trail leading from the lower colony to the vase was entirely free
from ants.... A curious and invariable feature of their behaviour was
that when an ant, returning in fright, met another approaching, the two
would always communicate; but each would pursue its own way, the second
ant continuing its journey to the spot where the first ant had turned
about, and then following that example. For some days after this there
were no ants visible on the wall, either above or below the shelf.
Then a few ants from the lower colony began to reappear; but instead
of visiting the vase, which had been the scene of the disaster, they
avoided it altogether, and, following the lower front edge of the shelf
to the tumbler standing near the middle, made their attack upon that
with precisely the same result.”

Lastly, Sir John Lubbock made some experiments with the express purpose
of testing the power of communication by ants. He found that if an ant
discovered a deposit of larvæ outside the nest, she would return to the
nest, and, even though she might have no larvæ to show, was able to
communicate her need of assistance—a number of friends proceeding to
follow her as a guide to the heap of larvæ which she had found.

In one very instructive experiment Sir John arranged three parallel
pieces of tape, each about two and a half feet long: one end of each
piece of tape was attached to the nest, and the other dipped into a
glass vessel. In the glass at the end of one of the tapes he placed a
considerable number of larvæ (300 to 600): in the glass at the end of
another of the pieces he put only two or three larvæ, while the third
glass he left empty. The object of the empty glass was to see whether
any of the ants would come to the glass under such circumstances by
mere accident. He then took two ants, one of which he placed in the
glass with the many larvæ, and the other in the glass with the few.
Each ant took a larva, carried it to the nest, then returned for more,
and so on. After each journey he put another larva in the glass with
the few larvæ, in order to replace the one which had been removed. The
result of the experiment was that during 47½ hours the ants which had
gone to the glass containing numerous larvæ brought 257 friends to
their assistance, while during 53 hours those which had gone to the
glass containing only two or three larvæ brought only 82 friends; and
no single ant came to the glass which contained no larva. Now, as all
the glasses were exposed to similar conditions, and as the roads to the
first two must, in the first instance at all events, have been equally
scented by the passage of ants over them, these results appear very
conclusive as proving some power of definite communication, not only
that larvæ are to be found, but even where the largest store is to be
met with.

As to the means of communication, or method of sign-making, there can
be no doubt that this in ants, as in bees, is mainly gestures made
by the antennæ; but that gestures of other kinds are also employed
is sufficiently well proved by the following observation of the
Rev. Dr. M’Cook. “I have seen an ant kneel down before another and
thrust forward the head, drooping quite under in fact, and lie there
motionless, thus expressing as plainly as sign-language could, her
desire to be cleansed. I at once understood the gesture, and so did the
supplicated ant, for she at once went to work.”

So much, then, for the power of sign-making displayed by the
Hymenoptera. As I have not much evidence of sign-making in any of the
other Invertebrata,[62] I shall pass on at once to the Vertebrata.

Ray observed the different tones used by the common hen, and found
them uniformly significant of different ideas, or emotional states;
therefore we may properly regard this as a system of language, though
of a very rudimentary form. He distinguishes altogether nine or ten
distinct tones, which are severally significant of as many distinct
emotions and ideas—namely, brooding, leading forth the brood, finding
food, alarm, seeking shelter, anger, pain, fear, joy or pride in having
laid an egg. Houzeau, who independently observed this matter, says that
the hen utters at least twelve significant sounds.[63]

Many other cases could be given among Birds, and a still greater number
among Mammals, of vocal tones being used as intentionally significant
of states of feeling and of definite ideas; but to save space I will
only render a few facts in a condensed form.

“In Paraguay, the _Cebus azaræ_ when excited utters at least six
distinct sounds, which excite in other monkeys similar emotions
(Rengger).... It is a more remarkable fact that the dog, since being
domesticated, has learned to bark in at least four or five distinct
tones: ... the bark of eagerness, as in the chase; that of anger, as
well as growling; the yelp, or howl of despair, when shut up; the
baying at night; the bark of joy when starting on a walk with his
master; and the very distinct one of demand or supplication, as when
wishing for a door or window to be opened.”[64]

I may next briefly add allusions to those instances of the use of
signs by mammals which are fully detailed in _Animal Intelligence_.

Mr. S. Goodbehere tells me of a pony which used to push back the inside
bolt of a gate in its paddock, and neigh for an ass which was loose in
the yard beyond; the ass would then come and push up the outside latch,
thus opening the gate and releasing the pony (p. 333).

With respect to gestures, Mrs. K. Addison wrote me of her
jackdaw—which lived in a garden, and which she usually supplied with a
bath—reminding her that she had forgotten to place the bath, by coming
before her and going through the movements of ablution upon the ground
(p. 316).

Youatt gives the case of a pig which was trained to point game with
great precision (pp. 339, 340), and this, as in the case of the dog,
implies a high development of the sign-making faculty. Every sportsman
must know how well a setter understands its own pointing, _and also the
pointing of other dogs_, as gesture-signs. As regards its own pointing,
if at any distance from the sportsman, the animal will look back to
see if the “point” has been noticed; and, if it has, the point will be
much more “steady” and prolonged than if the animal sees that it has
not been observed. As regards the pointing of other dogs, the “backing”
of one by another means that as soon as one dog sees another dog point
he also stands and points, whether or not he is in a position to scent
the game. In my previous work, while treating of artificial instincts,
I have shown (as Mr. Darwin had previously remarked) that in well-bred
sporting dogs a tendency to “back,” more or less pronounced, is
intuitive. But I have also observed among my own setters that even in
cases where a young dog does not show any innate disposition to “back,”
by working him with other dogs for a short time he soon acquires the
habit, without any other instruction than that which is supplied by his
own observation. I have also noticed that all sporting dogs are liable
to be deceived by the attitude which their companions strike when
defæcating; but this is probably due to their line of sight being so
much lower than that of a man, that slight differences of attitude are
not so perceptible to them as to ourselves.

Major Skinner writes of a large wild elephant which he saw on a
moonlight night coming out of a wood that skirted some water.
Cautiously advancing across the open ground to within a hundred yards
of the water, the animal stood perfectly motionless—the rest of the
herd, still concealed in the wood, being all the while so quiet and
motionless that not the least sound proceeded from them. Gradually,
after three successive advances, halting some minutes after each, he
moved up to the water’s edge, in which however he did not think proper
to quench his thirst, but remained for several minutes listening in
perfect stillness. He then returned cautiously and slowly to the point
at which he had issued from the wood, whence he came back with five
other elephants, with which he proceeded, somewhat less slowly than
before, to within a few yards of the tank, where he posted them as
patrols. He then re-entered the wood and collected the whole herd,
which must have amounted to between eighty and a hundred, and led them
across the open ground, with the most extraordinary composure and
quiet, till they came up to the five sentinels, when he left them for
a moment and again made a reconnaissance at the edge of the tank. At
last, being apparently satisfied that all was safe, he turned back,
and obviously gave the order to advance; “for in a moment,” says
Major Skinner, “the whole herd rushed to the water, with a degree of
unreserved confidence so opposite to the caution and timidity which had
marked their previous movements, that nothing will ever persuade me
that there was not rational and preconcerted co-operation throughout
the whole party”—and so, of course, some definite communication by
signs (p. 401).

With regard to the use of gesture-signs by cats, I have given such
cases as those of their imitating the begging of a terrier on observing
that the terrier received food in answer to this gesture (p. 414);
making a peculiar noise on desiring to have a door opened, which, if
not attended to, was followed up by “pulling one’s dress with its
claws, and then, having succeeded in attracting the desired attention,
it would walk to the street door and stop there, making the same cry
until let out” (p. 414); also of a cat which, on seeing her friend
the parrot “flapping its wings and struggling violently up to its
knees in dough,” ran upstairs after the cook to inform her of the
catastrophe—“mewing and making what signs she could for her to go
down,” till at last “she jumped up, seized her apron, and tried to drag
her down,” so that the cook did go down in time to save the bird from
being smothered. This gesture-sign of pulling at clothing, in order to
induce one to visit a scene of catastrophe, is of frequent occurrence
both in cats and dogs. Several instances are likewise given of cats
jumping on chairs and looking at bells when they want milk (this being
intended as a sign that they desire the bell pulled to call the servant
who brings the milk), placing their paws upon the bell as a still more
emphatic sign, or even themselves ringing the bell (p. 416).

Concerning gesture-signs made by dogs (other than pointing), I may
allude to a terrier which I had, and which when thirsty used to signify
his desire for water by begging before a wash-stand, or any other
object where he knew that water was habitually kept. And Sir John
Lefroy, F.R.S., gave me a similar, though still more striking, case of
his terrier, which it was the duty of a maid-servant to supply with
milk. One morning this servant was engaged on some needlework, and did
not supply the milk. “The dog endeavoured in every possible way to
attract her attention and draw her forth, and at last pushed aside the
curtain of a closet, and, although never having been taught to fetch or
carry, took between his teeth the cup she habitually used, and brought
it to her feet” (p. 466). Another case somewhat similar is given on the
same page.

Again, Mr. A. H. Browning wrote me:—“My attention was called to my dog
appearing in a great state of excitement, not barking (he seldom barks)
but whining, and performing all sorts of antics (in a human subject I
should have said _gesticulating_). The herdmen and myself returned to
the sty; we caught but one pig, and put him back; no sooner had we done
so, than the dog ran after each pig in succession, brought him back to
the sty by the ear, and then went after another, until the whole number
were again housed” (p. 450).

Further, I give an observation of my own (p. 445) on one terrier making
a gesture-sign to another. Terrier A being asleep in my house, and
terrier B lying on a wall outside, a strange dog, C, ran along below
the wall on the public road following a dog-cart. Immediately on seeing
C, B jumped off the wall, ran upstairs to where A was asleep, woke him
up by poking him with his nose in a determined and suggestive manner,
which A at once understood as a sign: he jumped over the wall and
pursued the dog C, although C was by that time far out of sight, round
a bend in the road.

On page 447 I give, on the authority of Dr. Beattie, the case of a
dog which saved his master’s life (who had fallen through the ice,
and was supporting himself with a gun placed across the opening), by
running into a neighbouring village, and pulling a man by the coat in
so significant a manner that he followed the animal and rescued the
gentleman. Many cases more or less similar to this one are recorded in
the anecdote books.

Concerning the use of gesture-signs by monkeys, I give on page 472 the
remarkable case recorded by James Forbes, F.R.S., of a male monkey
begging the body of a female which had just been shot. “The animal,”
says Forbes, “came to the door of the tent, and, finding threats of
no avail, began a lamentable moaning, and by the most expressive
gestures seemed to beg for the dead body. It was given him; he took it
sorrowfully in his arms and bore it away to his expecting companions.
They who were witnesses of this extraordinary scene resolved never
again to fire at one of the monkey race.”

Again, Captain Johnson writes of a monkey which he shot upon a tree,
and which then, as he says, “instantly ran down to the lowest branch of
a tree, as if he were going to fly at me, stopped suddenly, and coolly
put his paw to the part wounded, covered with blood, and held it out
for me to see. I was so much hurt at the time that it has left an
impression never to be effaced, and I have never since fired a gun at
any of the tribe. Almost immediately on my return to the party, before
I had fully described what had passed, a Syer came to inform us that
the monkey was dead. We ordered the Syer to bring it to us; but by the
time he returned the other monkeys had carried the dead one off, and
none of them could anywhere be seen” (p. 475).

And Sir William Hoste records a closely similar case. One of his
officers, coming home after a long day’s shooting, saw a female
monkey running along the rocks, with her young one in her arms. He
immediately fired, and the animal fell. On his coming up, she grasped
her little one close to her breast, and with her other hand pointed
to the wound which the ball had made, and which had entered above her
breast. Dipping her finger in the blood and holding it up, she seemed
to reproach him with having been the cause of her pain, and also of
that of the young one, to which she frequently pointed. “I never,” says
Sir William, “felt so much as when I heard the story, and I determined
never to shoot one of these animals as long as I lived” (p. 476).

Lastly, as proof that the more intelligent of the lower animals admit
of being _taught the use of signs of the most conventional character_
(or most remote from any natural expression of their feelings and
ideas), I may allude to the recent experiments by Sir John Lubbock
on “teaching animals to converse.” These experiments consisted in
writing on separate and similar cards such words as “bone,” “water,”
“out,” “pet me,” &c., and teaching a dog to bring a card bearing the
word expressive of his want at the time of bringing it. In this way
an association of ideas was established between the appearance of a
certain number and form of written signs, and the meaning which they
severally betokened. Sir John Lubbock found that his dog learnt the
correct use of those signs.[65] Of course in these experiments marks
of any other kind would have served as well as written words; for it
clearly would be absurd to suppose that the dog could read the letters,
so as mentally to construct them into the equivalent of a spoken
word, in any such way as a child would spell b-o-n-e, bone. But, all
the same, these experiments are of great interest as showing that it
falls within the mental capacity of the more intelligent animals to
appreciate the use of signs so conventional as those which constitute
a stage of writing _above_ the drawing of pictures, and _below_ the
employment of an alphabet.

       *       *       *       *       *

Enough has now been said to prove incontestably that animals present
what I have called the germ of the sign-making faculty. As the main
object of these chapters is to estimate the probability of human
language having arisen by way of a continuous development from this
germ, we may next turn to take a general survey of human language in
its largest sense, or as comprising all the manifestations of the
sign-making faculty.

Referring again to the schema (page 88), it is needless to consider
cases 1 and 2, for evidently these are on a psychological level in man
and animals. Case 3, also, especially in the direction of its branch
4, is to a large extent psychologically equivalent in men and animals:
so far as there is any difference it depends on the higher psychical
nature of man being much more rich in ideas which find their natural
expression in gestures or tones, and which, therefore, are impossible
in brutes. But it will be conceded that here there is nothing to
explain. The fact that man has a mind more richly endowed with ideas
carries with it, as a matter of course, the fact that their natural
expression is more multiplex.

The case, however, is different when we arrive at conventional signs;
for these attain so enormous a development in man as compared with
animals, that the question whether they do not really depend on some
additional mental faculty, distinct in kind, becomes fully admissible.

The first thing, then, we have to notice with regard to conventional
signs as used by man is, that no line of strict demarcation can be
drawn between them and natural signs; the latter shade off into the
former by gradations, which it becomes impossible to detect over large
numbers of individual cases. With respect to tones, for example, it
cannot be said, in many instances, whether this and that modulation,
which is now recognized as expressive of a certain state of feeling,
has always been thus expressive, or has only become so by conventional
habit; although, if we consider the different tones by which different
races of mankind express some of their similar feelings, we may be sure
that in these cases one or other of the differences must be due to
conventional habit—just as in the converse cases, in which all mankind
use the same tones to express the same feelings, we may be sure that
this mode of expression is natural. And so with gestures. Many which at
first sight we should, judging from our own feelings alone, suppose to
be natural—such, for instance, as kissing—are shown by observation
of primitive races to be conventional; while others which we should
probably regard as conventional—such, for instance, as shrugging the
shoulders—are shown by the same means to be natural.[66]

But for our present purposes it is clearly a matter of no consequence
that we should be able to classify all signs as natural or
conventional. For it is certain that animals employ both; and hence
no distinction between the brute and the man can be raised on the
question of the kind of signs which they severally employ as natural
or conventional. This distinction, therefore, may in future be
disregarded, and natural and conventional signs, _if made intentionally
as signs_, I shall consider as identical. For the sake of method,
however, I shall treat the sign-making faculty as exhibited by man in
the order of its probable evolution; and this means that I shall begin
with the most natural, or least conventional, of the systems. This is
the language of tone and gesture.




CHAPTER VI.

TONE AND GESTURE.


Tone and Gesture, considered as means of communication, may be
dealt with simultaneously. For while it cannot be said that either
historically or psychologically one is prior to the other, no more
can it be said that in the earliest phases of their development one
is more expressive than the other. All the more intelligent of the
lower animals employ both; and the hissings, spittings, growlings,
screamings, gruntings, cooings, &c., which in different species
accompany as many different kinds of gesture, are assuredly not less
expressive of the various kinds of feelings which are expressed. Again,
in our own species, tone is quite as general, and, within certain
limits, quite as expressive as gesture. Nay, even in fully developed
speech, rational meaning is largely dependent for its conveyance upon
slight differences of intonation. The five hundred words which go to
constitute the Chinese language are raised to three times that number
by the use of significant intonation; and even in the most highly
developed languages shades of meaning admit of being rendered in this
way which could not be rendered in any other.

Nevertheless, the language of tone, like the language of gesture,
clearly lies nearer to, and is more immediately expressive of the
logic of recepts, than is the language of articulation. This is easily
proved by all the facts at our disposal. We know that an infant makes
considerable advance in the language of tone and gesture before it
begins to speak; and, according to Dr. Scott, who has had a large
experience in the instruction of idiotic children, “those to whom
there is no hope of teaching more than the merest rudiments of speech,
are yet capable of receiving a considerable amount of knowledge by
means of signs, and of expressing themselves by them.”[67] Lastly,
among savages, it is notorious that tone, gesticulation, and grimace
play a much larger part in conversation than they do among ourselves.
Indeed, we have some, though not undisputed, evidence to show that
in the case of many savages gesticulation is so far a necessary aid
to articulation, that the latter without the former is but very
imperfectly intelligible. For example, “those who, like the Arapahos,
possess a very scanty vocabulary, pronounced in a quasi-intelligible
way, can hardly converse with one another in the dark.”[68] And, as Mr.
Tylor says, “the array of evidence in favour of the existence of tribes
whose language is incomplete without the help of gesture-signs, even
for things of ordinary import, is very remarkable.”[C] A fact which, as
he very properly adds, “constitutes a telling argument in favour of the
theory that the gesture-language is the original utterance of mankind
[as it is ontogenetically in the individual man], out of which speech
has developed itself more or less fully among different tribes.”[69]

In support of the same general conclusions I may here also quote the
following excellent remarks from Colonel Mallery’s laborious work on
Gesture-language:—[70]

“The wishes and emotions of very young children are conveyed in a
small number of sounds, but in a great variety of gestures and facial
expressions. A child’s gestures are intelligent long in advance of
speech; although very early and persistent attempts are made to give
it instruction in the latter but none in the former, from the time
when it begins _risu cognoscere matrem_. It learns words only as they
are taught, and learns them through the medium of signs which are
not expressly taught. Long after familiarity with speech, it consults
the gestures and facial expressions of its parents and nurses, as
if seeking thus to translate or explain their words. These facts
are important in reference to the biologic law that the order of
development of the individual is the same as that of the species....
The insane understand and obey gestures when they have no knowledge
whatever of words. It is also found that semi-idiotic children
who cannot be taught more than the merest rudiments of speech can
receive a considerable amount of information through signs, and can
express themselves by them. Sufferers from aphasia continue to use
appropriate gestures. A stammerer, too, works his arms and features as
if determined to get his thoughts out, in a manner not only suggestive
of the physical struggle, but of the use of gestures as a hereditary
expedient.”

Words, then, in so far as they are not intentionally imitative of
other sounds, and so approximate to gestures, are essentially more
conventional than are tones immediately expressive of emotions, or
bodily actions which appeal to the eye, and which, in so far as
they are intentionally significant, are made, as far as possible,
intentionally pictorial. Therefore, either to make or to understand
these more conventional signs requires a higher order of mental
evolution; and on this account it is that we everywhere find the
language of tone and gesture preceding that of articulate speech, as
at once the more simple, more natural, and therefore more _primitive_
means of conveying receptual ideas.

We find the same general truth exemplified in the fact that the
language of tone and gesture is always resorted to by men who do not
understand each others’ articulate speech; and although among the
races in which gesture-language has been carried to its highest degree
of elaboration most of the signs employed have become more or less
conventional, in the main they are still pictorial. This is directly
proved, without the need of special analysis, by the fact that the
members of such races are able to communicate with one another in a
manner so singularly complete that to an onlooker the result seems
almost magical.

Thus “the Indians who have been shown over the civilized East have
often succeeded in holding intercourse by means of their invention
and application of principles, in what may be called the voiceless
mother utterance, with white deaf-mutes, who surely have no semiotic
code more nearly connected with that attributed to the Indians than is
derived from their common humanity. They showed the greatest pleasure
in meeting deaf-mutes, precisely as travellers in a foreign country are
rejoiced to meet persons speaking their language.”[71]

Again, Tylor says, “Gesture-language is substantially the same all the
world over,” and Mallery confirms this by the remark that “the writer’s
study not only sustains it, but shows a surprising number of signs for
the same idea which are substantially identical, not only among savage
tribes, but among all peoples that use gesture-signs with any freedom.
Men, in groping for a mode of communication with each other, and using
the same general methods, have been under many varying conditions and
circumstances which have determined differently many conceptions and
their semiotic execution, but there have also been many of both which
were similar.”

Such being the case, it is a matter of interest to determine the syntax
of this language; for we may be sure that by so doing we are at work
upon the root-principles of the sign-making faculty where it arises
out of the logic of recepts, and not upon the developed ramifications
of this faculty where we find it wrought up into the more highly
conventional logic of concepts characteristic of speech. But before I
enter upon this branch of our subject, I shall say a few words to show
to what a high degree of perfection gesture-language admits of being
developed.

Tylor observes:—“As a means of communication, there is no doubt that
the Indian pantomime is not merely capable of expressing a few simple
and ordinary notions, but that to the uncultured savage, with his
few and material ideas, it is a very fair substitute for his scanty
vocabulary.”[72] And Colonel Mallery, in the admirable treatise already
referred to, shows in detail to what a surprising extent this “Indian
pantomime” is thus available as a substitute for speech. The following
may be selected from among the numerous dialogues and discourses which
he gives, and which all present the same general character. It is
communicated by Mr. Ivan Pehoff, who took notes of the conversation at
the time. The two conversers were Indians of different tribes.

“(1) _Kenaitze._—Left hand raised to height of eye, palm outward,
moved several times from right to left rapidly; fingers extended and
closed; pointing to strangers with left hand. Right hand describes a
curve from north to east.—‘Which of the north-eastern tribes is yours?’

“(2) _Tennanal._—Right hand, hollowed, lifted to mouth, then extended
and describing waving line gradually descending from right to left.
Left hand describing mountainous outline, apparently one peak rising
above the other. Said by Chalidoolts to mean, ‘Tenan-tnu-kohtana,
Mountain-river-men.’

“(3) _K._—Left hand raised to height of eye, palm outward, moved from
right to left, fingers extended. Left index describes curve from east
to west. Outline of mountain and river as in preceding sign.—‘How many
days from Mountain-river?’

“(4) _T._—Right hand raised towards index, and thumb forming first
crescent and then ring. This repeated three times.—‘Moon, new and full
three times.’

“(5) Right hand raised, palm to front, index raised and lowered at
regular intervals—‘Walked.’ Both hands imitating paddling of canoe,
alternately right and left.—‘Travelled three months on foot and by
canoe.’

“(6) Both arms crossed over breast, simulating shivering.—Cold,
winter.’

“(7) Right index pointing toward speaker.—‘I’; left hand pointing to
the west—‘travelled westward.’

“(8) Right hand lifted cup-shaped to mouth—‘Water.’ Right hand
describing waving line from right to left gradually descending,
pointing to the west.—‘River running westward.’

“(9) Right hand gradually pushed forward, palm upward, from height of
breast. Left hand shading eyes; looking at great distance.—‘Very wide.’

“(10) Left and right hands put together in shape of sloping
shelter.—‘Lodge, camp.’

“(11) Both hands lifted height of eye, palm inward, fingers
spread.—‘Many times.’

“(12) Both hands closed, palm outward, height of hips.—‘Surprised.’

“(13) Index pointing from eye forward.—‘See.’

“(14) Right hand held up, height of shoulder, three fingers extended,
left hand pointing to me.—‘Three white men.’

“(15) _K._—Right hand pointing to me, left hand held up, three fingers
extended.—‘Three white men.’

“(16) Making Russian sign of cross—‘Russians.’—‘Were the three white
men Russians?’

“(17) _T._—Left hand raised, palm inward, two fingers extended sign of
cross with right.—‘Two Russians.’

“(18) Right hand extended, height of eye, palm outward, moved outward a
little to right.—‘No.’

“(19) One finger of left hand raised.—‘One.’

“(20) Sign of cross with right.—‘Russian.’

“(21) Right hand, height of eye, fingers closed and extended, palm
outward a little to right.—‘Yes.’

“(22) Right hand carried across chest, hand extended, palm upward,
fingers and thumb closed as if holding something. Left hand in same
position carried across the right, palm downward.—‘Trade.’

“(23) Left hand upholding one finger, right pointing to me.—‘One white
man.’

“(24) Right hand held horizontally, palm downward, about four feet from
ground.—‘Small.’

“(25) Forming rings before eyes with index and thumb.—‘Eye-glasses.’

“(26) Right hand clinched, palm upward, in front of chest, thumb
pointing inward.—‘Gave one.’

“(27) Forming cup with right hand, simulating drinking.—‘Drink.’

“(28) Right hand grasping chest repeatedly, fingers curved and
spread.—‘Strong.’

“(29) Both hands pressed to temple, and head moved from side to
side.—‘Drunk, headache.’

“(30) Both index fingers placed together extended, pointing
forward.—‘Together.’

“(31) Fingers interlaced repeatedly.—‘Build.’

“(32) Left hand extended, fingers closed, placed slopingly against
left.—‘Camp.’

“(33) Both wrists placed against temples, hands curved upward and
outward, fingers spread.—‘Horns.’

“(34) Both hands horizontally lifted to height of shoulder, right
arm extended gradually full length, hand drooping a little at the
end.—‘Long back, moose.’

“(35) Both hands upright, palm outward, fingers extended and spread,
placing one before the other alternately.—‘Trees, dense forest.’

“(36) Sign of cross.—‘Russian.’

“(37) Motions of shooting again.—‘Shot.’

“(38) Sign for moose (Nos. 33, 34); showing two fingers of left
hand.—‘Two.’

“(39) Sign for camp as before (No. 10).—‘Camp.’

“(40) Right hand describing curve from east to west, twice.—‘Two days.’

“(41) Left hand lifted height of mouth, back outward, fingers closed as
if holding something; right hand simulating motion of tearing off, and
placing in mouth.—‘Eating moose meat.’

“(42) Right hand placed horizontally against heart; fingers closed,
moved forward a little and raised a little several times.—‘Glad at
heart.’

“(43) Fingers of left hand and index of right hand extended and
placed together horizontally, pointing forward height of chest. Hands
separated, right pointing eastward, and left westward.—‘Three men and
speaker parted, going west and east.’”

And so on, the conversation continuing up to 116 paragraphs. No
doubt some of these gestures appear conventional, and such is
undoubtedly the case with a great many which Colonel Mallery gives
in his _Dictionary of Indian Signs_. But this only shows that no
system of signs can be developed in any high degree without becoming
more or less conventional. The point I desire to be noticed is,
that gesture-language continues as far as possible—or as long as
possible—to be the natural expression of the logic of recepts.
As Mallery elsewhere observes, “the result of the studies, so far
as presented is, that that which is called the sign-language of
Indians is not, properly speaking, one language; but that it, and the
gesture-systems of deaf-mutes, and of all peoples, constitute together
one language—the gesture-speech of mankind—of which each system is a
dialect.” As showing this, and at the same time to give other instances
of the perfection of gesture-language, I may quote one instance of the
employment of such language by other nations, and one of its employment
by deaf-mutes. The first which I select is recorded by Alexander Dumas.

“Six weeks after this, I saw a second example of this faculty of mute
communication. This was at Naples. I was walking with a young man
of Syracuse. We passed by a sentinel. The soldier and my companion
exchanged two or three grimaces, which at another time I should not
even have noticed; but the instances I had before seen led me to give
attention. ‘Poor fellow!’ sighed my companion. ‘What did he say to
you?’ I asked. ‘Well,’ said he, ‘I thought that I recognized him as
a Sicilian, and I learned from him, as we passed, from what place he
came; he said he was from Syracuse, and that he knew me well. Then
I asked him how he liked the Neapolitan service; he said he did not
like it at all, and if his officers did not treat him better he should
certainly end by deserting. I then signified to him that if he ever
should be reduced to that extremity, he might rely upon me, and that I
would aid him all in my power. The poor fellow thanked me with all his
heart, and I have no doubt that one day or other I shall see him come.’
Three days after I was at the quarters of my Syracusan friend, when he
was told that a man asked to see him who would not give his name; he
went out and left me nearly ten minutes. ‘Well,’ said he on returning,
‘just as I said.’ ‘What?’ said I. ‘That the poor fellow would desert.’”

The instance which I select of gesture-language as employed by a
deaf-mute occurred in the National Deaf-Mute College at Washington, to
which Colonel Mallery took seven Uta Indians on March 6, 1880.

“Another deaf-mute gestured to tell us that, when he was a boy, he went
to a melon-field, tapped several melons, finding them to be green or
unripe: finally, reaching a good one, he took his knife, cut a slice
and ate it. A man made his appearance on horseback, entered the patch
on foot, found the cut melon, and, detecting the thief, threw the melon
towards him, hitting him in the back, whereupon he ran away crying. The
man mounted and rode off in an opposite direction.

“All of these signs were readily comprehended, although some of the
Indians varied very slightly in their translation. When the Indians
were asked whether, if they (the deaf-mutes) were to come to the Uta
country, they would be scalped, the answer was given, ‘Nothing would be
done to you; but we would be friends,’ as follows:—

“The palm of the right hand was brushed toward the right over that of
the left (‘nothing’), and the right made to grasp the palm of the left,
thumbs extended over, and lying upon the back of the opposing hand
(‘friends’).

“This was readily understood by the deaf-mutes. Deaf-mute sign of
milking a cow and drinking the milk was fully and quickly understood.

“The narrative of a boy going to an apple tree, hunting for ripe fruit,
and filling his pockets, being surprised by the owner and hit upon the
head with a stone, was much appreciated by the Indians and completely
understood.”

Innumerable other instances of the same kind might be given;[73]
but I have now said enough to establish the only points with which
I am here concerned—namely, that gesture-language admits of being
developed to a degree which renders it a fair substitute for spoken
language, where the ideas to be conveyed are not highly abstract; and
that it admits of being so developed without departing further from
a direct or natural expression of ideation (as distinguished from a
conventional or artificial) than allows it to be readily understood by
the sign-talkers, without any preconcerted agreement as to the meanings
to be attached to the particular signs employed.

Such being the case, it is of importance next to note that, as all
the existing races of mankind are a word-speaking race, we are not
now able to eliminate this factor, and to say how far the sign-making
faculty, as exhibited in the gesture-language of man, is indebted
to the elaborating influence produced by the constant and parallel
employment of spoken language. We can scarcely, however, entertain any
doubt that the reflex influence of speech upon gesture must have been
considerable, if not immense. Even the case of the deaf-mutes proves
nothing to the contrary; for these unfortunate individuals, although
not able themselves to speak, nevertheless inherit in their human
brains the psychological structure which has been built up by means of
speech; their sign-making _faculty_ is as well developed as in other
men, though, from a physiological accident, they are deprived of the
ordinary means of displaying it. Therefore we have no evidence to show
to what level of excellence the sign-making faculty of man would have
attained, if the race had been destitute of the faculty of speech. I
shall have to return to this consideration in the next chapter, and
only mention it here to avoid an undue estimate being prematurely
formed of the importance of gesture as a means of thought-formation, or
distinct from that of thought-expression.

I shall now proceed to analyze in some detail the syntax of
gesture-language. And here again I must depend for my facts upon the
two writers who have best studied this kind of language in a properly
scientific manner.

Mr. Tylor says:—“The gesture-language has no grammar, properly so
called; it knows no inflections of any kind, any more than the Chinese.
The same sign stands for ‘walk,’ ‘walkest,’ ‘walking,’ ‘walked,’
‘walker.’ Adjectives and verbs are not easily distinguished by the deaf
and dumb. ‘Horse, black, handsome, trot, canter,’ would be the rough
translation of the signs by which a deaf-mute would state that a black
handsome horse trots and canters. Indeed, our elaborate system of parts
of speech is but little applicable to the gesture-language, though, as
will be more fully said in another chapter, it may perhaps be possible
to trace in spoken language a Dualism, in some measure resembling that
of the Gesture-language, with its two constituent parts, the bringing
forward objects and actions in actual fact, and the mere suggestion
of them by imitation.... It has, however, a syntax which is worthy of
careful examination. The syntax of speaking man differs according to
the language he may learn, ‘equus niger,’ ‘a black horse;’ ‘hominem
amo,’ ‘j’aime l’homme.’ But the deaf-mute strings together the signs
of the various ideas he wishes to connect, in what appears to be the
natural order in which they follow one another in his mind, for it
is the same among the mutes in different countries, and is wholly
independent of the syntax which may happen to belong to the language
of their speaking friends. For instance, their usual construction is
not ‘Black horse,’ but ‘Horse black;’ not ‘Bring a black hat,’ but ‘Hat
black bring;’ not ‘I am hungry, give me bread,’ but ‘Hungry me, bread
give.’...

“The fundamental principle which regulates the order of the deaf-mutes’
signs, seems to be that enunciated by Schmalz: that which seems to
him the most important he always acts before the rest, and that which
seems to him superfluous he leaves out. For instance, to say, ‘My
father gave me an apple,’ he makes the sign for ‘apple,’ then that for
‘father,’ and then that for ‘I,’ without adding that for ‘give.’ The
following remarks, sent to me by Dr. Scott, seem to agree with this
view: With regard to the two sentences you give (I struck Tom with a
stick—Tom struck me with a stick), the sequence in the introduction
of the particular parts would in some measure depend on the part that
most attention was wished to be drawn towards. If a mere telling of
the fact was required, my opinion is that it would be arranged so,
‘I-Tom-struck-a-stick,’ and the passive form in a similar manner with
the change of ‘Tom’ first.

“Both these sentences are not generally said by the deaf-and-dumb
without their having been interested in the fact, and then, in coming
to tell of them, they first give that part they are most anxious to
impress on their hearer. Thus, if a boy had struck another boy, and the
injured party came to tell us, if he was desirous to acquaint us with
the idea that a particular boy did it, he would point to the boy first.
But if he was anxious to draw attention to his own suffering, rather
than to the person by whom it was caused, he would point to himself
and make the act of striking, and then point to the boy; or if he was
wishful to draw attention to the cause of his suffering, he might sign
the striking first, and then tell us afterwards by whom it was done.

“Dr. Scott is, so far as I know, the only person who has attempted
to lay down a set of distinct rules for the syntax of the
gesture-language. ‘The subject comes before the attribute, the object
before the action.’ A third construction is common, though not
necessary, ‘the modifier after the modified.’ The first construction,
by which the ‘horse’ is put before the ‘black,’ enables the deaf-mute
to make his syntax supply, to some extent, the distinction between
adjectives and substantives, which his imitative signs do not
themselves express.

“The other two are well exemplified by a remark of the Abbé Sicard’s:
A pupil to whom I one day put this question, ‘Who made God?’ and who
replied, ‘God made nothing,’ left me in no doubt as to this kind of
inversion, usual to the deaf-and-dumb, when I went on to ask him,
‘Who made the shoe?’ and he answered, ‘The shoe made the shoemaker.’
So when Laura Bridgman, who was blind as well as deaf-and-dumb, had
learnt to communicate ideas by spelling words on her fingers, she would
say, ‘Shut door,’ ‘Give book;’ no doubt because she had learnt these
sentences whole, but when she made sentences for herself, she would go
back to the natural deaf-and-dumb syntax, and spell out ‘Laura bread
give,’ to ask for bread to be given her, and ‘Water drink Laura,’ to
express that she wanted to drink water....

“A look of inquiry converts an assertion into a question, and fully
seems to make the difference between ‘The master is come,’ and ‘Is
the master come?’ The interrogative pronouns ‘Who?’ ‘What?’ are made
by looking or pointing about in an inquiring manner; in fact, by a
number of unsuccessful attempts to say, ‘he,’ ‘that.’ The deaf-and-dumb
child’s way of asking, ‘Who has beaten you?’ would be, ‘You beaten;
who was it?’ Though it is possible to render a great mass of simple
statements and questions, almost gesture for word, the concretism of
thought which belongs to the deaf-mute, whose mind has not been much
developed by the use of written language, and even to the educated one
when he is thinking and uttering his thoughts in his native signs,
commonly requires more complex phrases to be recast. A question so
common amongst us as, ‘What is the matter with you?’ would be put, ‘You
crying? You have been beaten?’ and so on. The deaf-and-dumb child does
not ask, ‘What did you have for dinner yesterday?’ but ‘Did you have
soup?’ ‘Did you have porridge?’ and so forth. A conjunctive sentence
he expresses by an alternative or contrast; ‘I should be punished if
I were lazy and naughty,’ would be put, ‘I lazy, naughty, no!—lazy,
naughty, I punished, yes!’ Obligation may be expressed in a similar
way; ‘I must love and honour my teacher,’ may be put, ‘Teacher, I beat,
deceive, scold, no!—I love, honour, yes!’ As Steinthal says in his
admirable essay, it is only the certainty which speech gives to a man’s
mind in holding fast ideas in all their relations, which brings him to
the shorter course of expressing only the positive side of the idea,
and dropping the negative....

“To ‘make’ is too abstract an idea for the deaf-mute; to show that
the tailor makes the coat, or that the carpenter makes the table, he
would represent the tailor sewing the coat, and the carpenter sawing
and planing the table. Such a proposition as ‘Rain makes the land
fruitful,’ would not come into his way of thinking: ‘rain fall, plants
grow,’ would be his pictorial expression.... The order of the signs by
which the Lord’s Prayer is rendered is much as follows:—‘Father our,
heaven in—name Thy hallowed—kingdom Thy come—will Thy done—earth
on, heaven in, as. Bread give us daily—trespasses our forgive us, them
trespass against us, forgive as. Temptation lead not—but evil deliver
from—Kingdom power glory thine for ever.’”[74]

I shall now add some quotations from Colonel Mallery on the same
subject.

“The reader will understand without explanation that there is in
sign-language no organized sentence such as is in the language of
civilization, and that he must not look for articles or particles,
or passive voice or case or grammatic gender, or even what appears
in those languages as a substantive or a verb, as a subject or a
predicate, or as qualifiers or inflexions. The sign radicals, without
being specifically any of our parts of speech, may be all of them
in turn. Sign-language cannot show by inflection the reciprocal
dependence of words and sentences. Degrees of motion corresponding
with vocal intonations are only used rhetorically, or for degrees of
comparison. The relations of ideas and objects are therefore expressed
by placement, and their connection is established when necessary by
the abstraction of ideas. The sign-talker is an artist, grouping
persons and things so as to show the relations, and the effect is that
which is seen in a picture. But though the artist has the advantage
in presenting in a permanent connected scene the result of several
transient signs, he can only present it as it appears at a single
moment. The sign-talker has the succession of time at his disposal,
and his scenes move and act, are localized and animated, and their
arrangement is therefore more varied and significant.”[75]

The following is the order in which the parable of the Prodigal Son
would be translated by a cultivated sign-talker, with Colonel Mallery’s
remarks thereon:—

“‘Once, man one, sons two. Son younger say, Father property your
divide: part my, me give. Father so.—Son each, part his give. Days
few after, son younger money all take, country far go, money spend,
wine drink, food nice eat. Money by and by gone all. Country everywhere
food little: son hungry very. Go seek man any, me hire. Gentleman
meet. Gentleman son send field swine feed. Son swine husks eat,
see—self husks eat want—cannot—husks him give nobody. Son thinks,
say, father my, servants many, bread enough, part give away can—I
none—starve, die. I decide: Father I go to, say I bad, God disobey,
you disobey—name my hereafter _son_, no—I unworthy. You me work give
servant like. So son begin go. Father far look: son see, pity, run,
meet, embrace. Son father say, I bad, you disobey, God disobey—name my
hereafter _son_, no—I unworthy. But father servants call, command robe
best bring, son put on, ring finger put on, shoes feet put on, calf fat
bring, kill. We all eat, merry. Why? Son this my formerly dead, now
alive: formerly lost, now found: rejoice.’

“It may be remarked, not only from this example, but from general
study, that the verb ‘to be’ as a copula or predicant does not have
any place in sign-language. It is shown, however, among deaf-mutes
as an assertion of presence or existence by a sign of stretching
the arms and hands forward and then adding the sign of affirmation.
_Time_ as referred to in the conjunctions _when_ and _then_ is not
gestured. Instead of the form, ‘When I have had a sleep I will go to
the river,’ or ‘After sleeping I will go to the river,’ both deaf-mutes
and Indians would express the intention by ‘Sleep done, I river go.’
Though time present, past, and future is readily expressed in signs,
it is done once for all in the connection to which it belongs, and
once established is not repeated by any subsequent intimation, as
is commonly the case in oral speech. Inversion, by which the object
is placed before the action, is a striking feature of the language
of deaf-mutes, and it appears to follow the natural method by which
objects and actions enter into the mental conception. In striking
a rock the natural conception is not first of the abstract idea of
striking or of sending a stroke into vacancy, seeing nothing and having
no intention of striking anything in particular, when suddenly a rock
rises up to the mental vision and receives the blow; the order is that
the man sees the rock, has the intention to strike it, and does so;
therefore he gestures, ‘I rock strike.’ For further illustration of
this subject, a deaf-mute boy, giving in signs the compound action
of a man shooting a bird from a tree, first represented the tree,
then the bird as alighting upon it, then a hunter coming toward and
looking at it, taking aim with a gun, then the report of the latter
and the falling and the dying gasps of the bird. These are undoubtedly
the successive steps that an artist would have taken in drawing the
picture, or rather successive pictures, to illustrate the story....
Degrees of comparison are frequently expressed, both by deaf-mutes and
by Indians, by adding to the generic or descriptive sign that for ‘big’
or ‘little.’ _Damp_ would be ‘wet—little’; _cool_, ‘cold—little’;
_hot_, ‘warm—much.’ The amount or force of motion also often indicates
corresponding diminution or augmentation, but sometimes expresses
a different shade of meaning, as is reported by Dr. Matthews with
reference to the sign for _bad_ and _contempt_. This change in degree
of motion is, however, often used for emphasis only, as is the raising
of the voice in speech or italicizing and capitalizing in print. The
Prince of Wied gives an instance of a comparison in his sign for
_excessively hard_, first giving that for _hard_, viz.: Open the left
hand, and strike against it several times with the right (with the
backs of the fingers). Afterwards he gives _hard, excessively_, as
follows: Sign for _hard_, then place the left index finger upon the
right shoulder, at the same time extend and raise the right arm high,
extending the index finger upward, perpendicularly.”

I have entered thus at some length into the syntax of gesture-language
because this language is, as I have before remarked, the most natural
or immediate mode of giving expression to the logic of recepts; it is
the least symbolic or conventional phase of the sign-making faculty,
and therefore a study of its method is of importance in such a general
survey of this faculty as we are endeavouring to take. The points
in the above analysis to which I would draw attention as the most
important are, the absence of the copula and of many other “parts of
speech,” the order in which ideas are expressed, the pictorial devices
by which the ideas are presented in as concrete a form as possible, and
the fact that no ideas of any high abstraction are ever expressed at
all.[76]




CHAPTER VII.

ARTICULATION.


It will be my aim in this chapter to take a broad view of Articulation
as a special development of the general faculty of sign-making,
reserving for subsequent chapters a consideration of the philosophy of
Speech.

On the threshold of articulate language, then, we have four several
cases to distinguish: first, articulation by way of meaningless
imitation; second, meaningless articulation by way of a spontaneous
or instinctive exercise of the organs of speech; third, understanding
of the signification of articulate sounds, or words; and fourth,
articulation with an intentional attribution of the meaning understood
as attaching to the words. I shall consider each of these cases
separately.

       *       *       *       *       *

The meaningless imitation of articulate sounds occurs in talking birds,
young children, not unfrequently in savages, in idiots, and in the
mentally deranged. The faculty of such meaningless imitation, however,
need not detain us; for it is evident that the mere re-echoing of a
verbal sound is of no further psychological significance than is the
mimicking of any other sound.

       *       *       *       *       *

Meaningless articulation of a spontaneous or instinctive kind occurs
in young children, in uneducated deaf-mutes, and also in idiots.[77]
Infants usually (though not invariably) begin with such syllables as
“alla,” “tata,” “mama,” and “papa” (with or without the reduplication)
before they understand the meaning of any word. One of my own children
could say all these syllables very distinctly at the age of eight
months and a half; and I could detect no evidence at that time of his
understanding words, or of his having learnt these syllabic utterances
by imitation. Another child of mine, which was very long in beginning
to speak, at fourteen and a half months old said once, and only once,
but very distinctly “Ego.” This was certainly not said in imitation
of any one having uttered the word in her presence, and therefore I
mention the incident to show that meaningless articulation in young
children is spontaneous or instinctive, as well as intentionally
imitative; for at that age the only other syllables which this child
had uttered were those having the long _[=a]_, as above mentioned. Were
it necessary, I could give many other instances of this fact; but, as
it is generally recognized by writers on infant psychology, I need not
wait to do so.

       *       *       *       *       *

We now come to the third of our divisions, or the understanding of
articulate sounds. And this is an important matter for us, because
it is evident that the faculty of appreciating the meaning of words
betokens a considerable advance in the general faculty of language. As
we have before seen, tone and gesture, being the natural expression of
the logic of recepts—and so even in their most elaborated forms being
intentionally pictorial,—are as little as possible conventional;
but words, being coined expressly for the subservience of concepts,
are always less graphic, and usually arbitrary. Therefore, although
it would of course be wrong to say that a higher faculty is required
to learn the arbitrary association between a particular verbal sound
and a particular act or phenomenon, than is required to depict an
abstract idea in gesture; this only shows that where higher faculties
are present, they are able to display themselves in gesture as well
as in speech. The consideration which I now wish to present is that
understanding a word implies (other things equal, or supposing
the gesture not to be so purely conventional as a word) a higher
development of the sign-making faculty than does the understanding of a
tone or gesture—so that, for instance, if an animal were to understand
the word “Whip,” it would show itself more intelligent in appreciating
signs than it would by understanding the gesture of threatening as with
a whip.

Now, the higher animals unquestionably do understand the meanings of
words; idiots too low in the scale themselves to speak are in the same
position; and infants learn the signification of many articulate sounds
long before they begin themselves to utter them.[78] In all these cases
it is of course important to distinguish between the understanding of
words and the understanding of tones; for, as already observed, both
in the animal kingdom and in the growing child it is evident that
the former represents a much higher grade of mental evolution than
does the latter—a fact so obvious to common observation that I need
not wait to give illustrations. But although the fact is obvious,
it is no easy matter to distinguish in particular cases whether the
understanding is due to an appreciation of words, to that of tones,
or to both combined. We may be sure, however, that words are never
understood unless tones are likewise so, and that understanding of
words may be assisted by understanding of the tones in which they are
uttered. Therefore, the only method of ascertaining where words as
such are first understood, is to find where they are first understood
irrespective of the tones in which they are uttered. This criterion—so
far, at least, as my evidence goes—excludes all cases of animals
obeying commands, answering to their names, &c., with the exception
of the higher mammalia. That is to say, while the understanding of
certain tones of the human voice extends at least through the entire
vertebrated series,[79] and occurs in infants only a few weeks old; the
understanding of words without the assistance of tones appears to occur
only in a few of the higher mammalia, and first dawns in the growing
child during the second year.[80]

The fact that the more intelligent Mammalia are able to understand
words irrespective of tones is, as I have said, important; and
therefore I shall devote a few sentences to prove it.

My friend Professor Gerald Yeo had a terrier, which was taught to
keep a morsel of food on its snout till it received the verbal signal
“Paid for;” and it was of no consequence in what tones these words
were uttered. For even if they were introduced in an ordinary stream
of conversation, the dog distinguished them, and immediately tossed
the food into his mouth. Seeing this, I thought it worth while to
try whether the animal would be able to distinguish the words “Paid
for” from others presenting a close similarity of sound; and,
therefore, while he was expecting the signal, I said “Pinafore;” the
dog gave a start, and very nearly threw the food off his nose; but
immediately arrested the movement, evidently perceiving his mistake.
This experiment was repeated many times with these two closely similar
verbal sounds, and always with the same result: the dog clearly
distinguished between them. I have more recently repeated this
experiment on another terrier, which had been taught the same trick,
and obtained exactly the same results.

The well-known anecdote told of the poet Hogg may be fitly alluded to
in this connection. A Scotch collie was able to understand many things
that his master said to him, and, as proof of his ability, his master,
while in the shepherd’s cottage, said in as calm and natural tone as
possible, “I’m thinking the cow’s in the potatoes.” Immediately the
dog, which had been lying half asleep on the floor, jumped up, ran into
the potato-field, round the house, and up the roof to take a survey;
but finding no cow in the potatoes, returned and lay down again. Some
little time afterwards his master said as quietly as before, “I’m sure
the cow’s in the potatoes,” when the same scene was repeated. But on
trying it a third time, the dog only wagged his tail. Similarly, Sir
Walter Scott, among other anecdotes of his bull terrier, says:—“The
servant at Ashestiel, when laying the cloth for dinner, would say to
the dog as he lay on the mat by the fire, ‘Camp, my good fellow, the
sheriff’s coming home by the ford,’ or ‘by the hill;’ and the poor
animal would immediately go forth to welcome his master, advancing as
far and as fast as he was able in the direction indicated by the words
addressed to him.” And numberless other anecdotes of the same kind
might be quoted.[81]

But the most remarkable display of the faculty in question on the
part of a brute which has happened to fall under my own observation,
is that which many other English naturalists must have noticed in the
case of the chimpanzee now in the Zoological Gardens. This ape has
learnt from her keeper the meanings of so many words and phrases,
that in this respect she resembles a child shortly before it begins
to speak. Moreover, it is not only particular words and particular
phrases which she has thus learnt to understand; she also understands,
to a large extent, the combination of these words and phrases in
sentences, so that the keeper is able to explain to the animal what
it is that he requests her to do. For example, she will push a straw
through any particular meshes in the network of her cage which he may
choose successively to indicate by such phrases as—“The one nearest
your foot; now the one next the key-hole; now the one above the bar,”
&c., &c. Of course there is no pointing to the places thus verbally
designated, nor is any order observed in the designation. The animal
understands what is meant by the words alone, and this even when a
particular mesh is named by the keeper remarking to her the accident of
its having a piece of straw already hanging through it.

In connection with the subject of the present treatise it appears to
me difficult to overrate the significance of these facts. The more
that my opponents maintain the fundamental nature of the connection
between speech and thought, the greater becomes the importance of
the consideration that the higher animals are able in so surprising
a degree to participate with ourselves in the understanding of
words. From the analogy of the growing child we well know that the
understanding of words precedes the utterance of them, and therefore
that the condition to the attainment of conceptual ideation is given
in this higher product of receptual ideation. Surely, then, the
fact that not a few among the lower animals (especially elephants,
dogs, and monkeys) demonstrably share with the human infant this
higher excellence of receptual capacity, is a fact of the largest
significance. For it proves at least that these animals share with an
infant those qualities of mind, which in the latter are immediately
destined to serve as the vehicle for elevating ideation from the
receptual to the conceptual sphere: the faculty of understanding
words in so considerable a degree brings us to the very borders of
the faculty of using words with an intelligent appreciation of their
meaning.

Familiarity with the facts now before us is apt to blunt this their
extraordinary significance; and therefore I invite my opponents to
reflect how differently my case would have stood, supposing that none
of the lower animals had happened to have been sufficiently intelligent
thus to understand the meanings of words. How much greater would
then have been the argumentative advantage of any one who undertook
to prove the distinctively human prerogative of the Logos. No mere
brute, it might have been urged, has ever displayed so much as the
first step in approaching to this faculty: from its commencement to
its termination the faculty belongs exclusively to mankind. But, as
matters actually stand, this cannot be urged: the lower animals share
with us the order of ideation which is concerned in the understanding
of words—and words, moreover, so definite and particular in meaning
as is involved in explaining the particular mesh in a large piece
of wire-netting through which it is required that a straw shall be
protruded. While watching this most remarkable performance on the
part of the chimpanzee, I felt more than ever disposed to agree with
the great philologist Geiger, where he says “there is scarcely a more
wonderful relationship upon the earth than this accession [_i.e._ the
understanding of words] by the intelligence of animals to that of
man.”[82]

I take it then, as certainly proved, that the germ of the sign-making
faculty which is present in the higher animals is so far developed as
to enable these animals to understand not merely conventional gestures,
but even articulate sounds, irrespective of the tones in which they
are uttered. Therefore, in view of this fact, together with the fact
previously established that these same animals frequently make use of
conventional gesture-signs themselves, I think we are justified in
concluding _a priori_, that if these animals were able to articulate,
they would employ simple words to express simple ideas. I do not say,
nor do I think, that they would form propositions; but it seems to me
little less than certain that they would use articulate sounds, as
they now use natural or conventional tones and gestures, to express
such ideas as they now express in either of these ways. For instance,
it would involve the exercise of no higher psychical faculty to say
the word “Come,” than it does to pull at a dress or a coat to convey
the same idea; or to utter the word “Open,” instead of mewing in a
conventional manner before a closed door; or, yet again, to utter the
word “Bone,” than to select and carry a card with the word written
upon it. If this is so, we must conclude that the only reason why the
higher Mammalia do not employ simple words to convey simple ideas, is
that which we may term an accidental reason, so far as their psychology
is concerned; it is an anatomical reason, depending merely on the
structure of their vocal organs not admitting of articulation.[83]

Of course at this point my attention will be called to the case of
talking birds; for it is evident that in them we have the anatomical
conditions required for speech, though assuredly occurring at a
most unlikely place in the animal series; and therefore these
animals may be properly adduced to test the validity of my _a
priori_ inference—namely, that if the more intelligent brutes could
articulate, they would make a proper use of simple verbal signs.
Let it, however, be here remembered that birds are lower in the
psychological scale than dogs, or cats, or monkeys; and, therefore,
that the inference which I drew touching the latter need not
necessarily be held as applying also to the former. Nevertheless, it
so happens that even in the case of these psychologically inferior
animals the evidence, such as it is, is not opposed to my inference: on
the contrary, there is no small body of facts which goes to support it
in a very satisfactory manner. A consideration of this evidence will
now serve to introduce us to the fourth and last case presented in the
programme at the beginning of this chapter, or the case of articulation
with attribution of the meaning understood as attaching to the words.

       *       *       *       *       *

Taking, first, the case of proper names, it is unquestionable that
many parrots know perfectly well that certain names belong to certain
persons, and that the way to call these persons is to call their
appropriate names. I knew a parrot which used thus to call its mistress
as intelligently as any other member of the household; and if she went
from home for a day, the bird became a positive nuisance from its
incessant calling for her to come.

And in a similar manner talking birds often learn correctly to assign
the names of other pet animals kept in the same house, or even the
names of inanimate objects. There can thus be no question as to the use
by talking birds of proper names and noun-substantives.

With respect to adjectives, Houzeau very properly remarks that the
apposite manner in which some parrots habitually use certain words
shows an aptitude correctly to perceive and to name qualities as well
as objects. Nor is this anything more than we might expect, seeing, on
the one hand, as already shown, that animals possess generic ideas of
many qualities, and, on the other, that an obvious quality is as much
a matter of immediate observation—and so of sensuous association—as
is the object of which it may happen to be a quality.

Again, it is no less certain that many parrots will understand the
meaning of active and passive verbs, whether as uttered by others
or by themselves. The request to “Scratch Poll” or the announcement
“Poll is thirsty,” when intentionally used as signs, show as true an
appreciation of the meaning of verbs—or rather, let us say, of verbal
signs indicative of actions and states—as is shown by the gesture-sign
of a dog or a cat in pulling one’s dress to indicate “come,” or mewing
before an open door to signify “open.”

But not only may talking birds attach appropriate significations
to nouns, adjectives, and verbs; they may even use short sentences
in a way serving to show that they appreciate—not, indeed, their
grammatical structure—but their applicability as a whole to particular
circumstances.[84] But this again is not a matter to excite surprise.
For all such instances of the apposite use of words or phrases by
talking birds are found on inquiry to be due, as antecedently we should
expect that they must, to the principle of association. The bird hears
a proper name applied to a person, and so, on learning to say the name,
henceforth associates it with that person. And similarly with phrases.
These with talking birds are mere vocal gestures, which in themselves
present but little more psychological significance than muscular
gestures. The verbal petition, “Scratch poor poll,” does not in itself
display any further psychological development than the significant
gesture already alluded to of depressing the head against the bars
of the cage; and similarly with all cases of the appropriate use of
longer phrases. Thus, supposing it to be due to association alone, a
verbal sign of any kind is not much more remarkable, or indicative of
intelligence, than is a gesture sign, or a vocal sign of any other
kind. The only respect in which it differs from such other signs is in
the fact that it is wholly arbitrary or conventional; and although,
as I have previously said, I do consider this an important point of
difference, I am not at all surprised that even the intelligence of a
bird admits of such special associations being formed, or that a wholly
arbitrary sign of any kind should here be acquired by this means, and
afterwards used as a sign.

And that the verbal signs used by talking birds are due to association,
and association only, all the evidence I have met with goes to prove.
As showing how association acts in this case, I may quote the following
remarks of Dr. Samuel Wilks, F.R.S., on his own parrot, which he
carefully observed. He says that when alone this bird used to “utter a
long catalogue of its sayings, more especially if it heard talking at a
distance, as if wishing to join in the conversation, but at other times
a particular word or phrase is only spoken when suggested by a person
or object. Thus, certain friends who have addressed the bird frequently
by some peculiar expression, or the whistling of an air, will always
be welcomed by the same words or tune, and as regards myself, when I
enter the house—for my footstep is recognized—the bird will repeat
one of my sayings. If the servants enter the room Poll will be ready
with one of their expressions, and in their own tone of voice. It is
clear that there is a close association in the bird’s mind between
certain phrases and certain persons or objects, for their presence or
voice at once suggests some special word. For instance, my coachman,
when coming for orders, has so often been told half-past two, that no
sooner does he come to the door than Poll exclaims, ‘Half-past two.’
Again, having at night found her awake, and having said, ‘Go to sleep,’
if I have approached the cage after dark the same words have been
repeated. Then, as regards objects, if certain words have been spoken
in connection with them, these are ever afterwards associated together.
For example, at dinner time the parrot, having been accustomed to have
savory morsels given to her, I taught her to say, ‘Give me a bit.’ This
she now constantly repeats, but only and appropriately at dinner-time.
The bird associates the expression with something to eat, but, of
course, knows no more than the infant the derivation of the words she
is using. Again, being very fond of cheese, she easily picked up the
word, and always asks for cheese towards the end of the dinner course,
and at no other time. Whether the bird attaches the word to the true
substance or not I cannot say, but the time of asking for it is always
correct. She is also fond of nuts, and when these are on the table
she utters a peculiar squeak; this she has not been taught, but it is
Poll’s own name for nuts, for the sound is never heard until the fruit
is in sight. Some noises which she utters have been obtained from the
objects themselves, as that of a cork-screw at the sight of a bottle of
wine, or the noise of water poured into a tumbler on seeing a bottle of
water. The passage of the servant down the hall to open the front door
suggests a noise of moving hinges, followed by a loud whistle for a
cab.”[85]

Concerning the accuracy of these observations I have no doubt, and I
could corroborate most of them were it necessary. It appears, then,
first, that talking birds may learn to associate certain words with
certain objects and qualities, certain other words or phrases with the
satisfaction of particular desires and the observation of particular
actions; words so used we may term vocal-gestures. Second, that they
may invent sounds of their own contriving, to be used in the same
way; and that these sounds may be either imitative of the objects
designated, as the sound of running fluid for “Water,” or arbitrary, as
the “particular squeak” that designated “Nuts.” Third, but that in a
much greater number of cases the sounds (verbal or otherwise) uttered
by talking birds are imitative only, without the animals attaching to
them any particular meaning. The third division, therefore, we may
neglect as presenting no psychological import; but the first and second
divisions require closer consideration.

In designating as “vocal gestures”[86] the correct use (acquired by
direct association) of proper names, noun-substantives, adjectives,
verbs, and short phrases, I do not mean to disparage the faculty which
is displayed. On the contrary, I think this faculty is precisely
the same as that whereby children first learn to talk; for, like
the parrot, the infant learns by direct association the meanings of
certain words (or sounds) as denotative of certain objects, connotative
of certain qualities, expressive of certain desires, actions, and
so on. The only difference is that, in a few months after its first
commencement in the child, this faculty develops into proportions far
surpassing those which it presents in the bird, so that the vocabulary
becomes much larger and more discriminative. But the important thing
to attend to is that at first, and for several months after its
commencement, the vocabulary of a child is always designative of
particular objects, qualities, actions, or desires, and is acquired
by direct association. The distinctive peculiarity of human speech,
which elevates it above the region of animal gesticulation, is of
later growth—the peculiarity, I mean, of using words, no longer as
stereotyped in the framework of special and direct association, but as
movable types to be arranged in any order that the meaning before the
mind may dictate. When this stage is reached, we have the faculty of
predication, or of the grammatical formation of sentences which are
no longer of the nature of vocal gestures, designative of particular
objects, qualities, actions, or states of mind: but vehicles for the
conveyance of ever-changing thoughts.

We shall presently see that this distinction between the naming and
the predicating phases of language is of the highest importance in
relation to the subject of the present treatise; but meanwhile all we
have to note is that the naming phase of spoken language occurs—in
a rudimentary form, indeed, but still unquestionably—in the animal
kingdom; and that the fact of its doing so is not surprising, if
we remember that in this stage language is nothing more than vocal
gesticulation. Psychologically considered, there is nothing more
remarkable in the fact that a bird which is able to utter an articulate
sound should learn by association to use that sound as a conventional
sign, than there is that it should learn by association similarly to
use a muscular action, as it does in the act of depressing its head
as a sign to have it scratched. Therefore we may now, I think, take
the position as established _a posteriori_ as well as _a priori_, that
it is, so to speak, a mere accident of anatomy that all the higher
animals are not able thus far to talk; and that, if dogs or monkeys
were able to do so, we have no reason to doubt that their use of words
and phrases would be even more extensive and striking than that which
occurs in birds. Or as Professor Huxley observes, “a race of dumb
men, deprived of all communication with those who could speak, would
be little indeed removed from the brutes. The moral and intellectual
differences between them and ourselves would be practically infinite,
though the naturalist should not be able to find a single shadow even
of specific structural difference.[87]

We must next briefly consider the remaining feature in the psychology
of talking birds to which Dr. Wilks has drawn attention, namely, that
of inventing sounds of their own contrivance to be used as designative
of objects and qualities, or expressive of desires—sounds which may
be either imitative of the things designated, or wholly arbitrary.
And this, I think, is a most important feature; for it serves still
more closely to connect the faculty of vocal sign-making in animals
with the faculty of speech in man. Thus, turning first to the case of
a child beginning to speak, as Dr. Wilks points out—and nearly all
writers on the philosophy of language have noticed—“baby talk” is to
a large extent onomatopoetic. And although this is in part due to an
inheritance of “nursery language,” the very fact that nursery language
has come to contain so large an element of onomatopœia is additional
proof, were any required, that this kind of word-invention appeals with
ready ease to the infant understanding. But, on the other hand, no one
can have attended to the early vocabulary of any child without having
observed a fertile tendency to the invention of words wholly arbitrary.
As this spontaneous invention of arbitrary words by young children
will be found of importance in later stages of my exposition, I will
conclude the present chapter by presenting evidence to show the extent
to which, under favourable circumstances, it may proceed. Meanwhile,
however, I desire to point out that all such cases of the invention of
arbitrary vocal signs by young children differ from the analogous cases
furnished by parrots only in that the former are usually articulate,
while the latter are usually not so. But this difference is easily
explained when we remember that hereditary tendency makes as strongly
in the direction of inarticulate sounds in the case of the bird, as in
the case of the infant it makes in the direction of articulate.

There still remains one feature in the psychology of talking birds to
which I must now draw prominent attention. So far as I can ascertain it
has not been mentioned by any previous writer, although I should think
it is one that can scarcely have escaped the notice of any attentive
observer of these animals. I allude to the aptitude which intelligent
parrots display of extending their articulate signs from one object,
quality, or action, to another which happens to be strikingly similar
in kind. For example, one of the parrots which I kept under observation
in my own house learnt to imitate the barking of a terrier, which
also lived in the house. After a time this barking was used by the
parrot as a denotative sound, or proper name, for the terrier—_i.e._
whenever the bird saw the dog it used to bark, whether or not the dog
did so. Next, the parrot ceased to apply this denotative name to that
particular dog, but invariably did so to any other, or unfamiliar, dog
which visited the house. Now, the fact that the parrot ceased to bark
when it saw my terrier after it had begun to bark when it saw other
dogs, clearly showed that it distinguished between individual dogs,
while receptually perceiving their class resemblance. In other words,
the parrot’s name for an individual dog became extended into a generic
name for all dogs. Observations of this kind might no doubt have been
largely multiplied, if observers had thought it worth while to record
such apparently trivial facts.

       *       *       *       *       *

In this general survey of articulate language, then, we have reached
these conclusions, all of which I take to be established by the
evidence of direct and adequate observation.

There are four divisions of the faculty of articulate sign-making
to be distinguished:—namely, meaningless imitation, instinctive
articulation, understanding words irrespective of tones, and
intentional use of words as signs. Cases falling under the first
division do not require consideration. Cases belonging to the second,
being due to hereditary influence, occur only in infants, uneducated
deaf-mutes and idiots. Understanding of words is shown by animals
and idiots as well as by infants, and implies, _per se_, a higher
development of the sign-making faculty than does the understanding of
tones, or gestures—unless, of course, the latter happen to be of as
purely conventional a character as words. And, lastly, concerning the
intentional use of words as signs, we have noticed the following facts.

Talking birds—which happen to be the only animals whose vocal organs
admit of uttering articulate sounds—show themselves capable of
correctly using proper names, noun-substantives, adjectives, verbs,
and appropriate phrases, although they do so by association alone, or
without appreciation of grammatical structure. Words are to them vocal
gestures, as immediately expressive of the logic of recepts as any
other signs would be. Nevertheless, it is important to observe that
this faculty of vocal gesticulation is the first phase of articulate
speech in a growing child, is the last to disappear in the descending
scale of idiocy, and is exhibited by talking birds in so considerable
a degree that the animals even invent names (whether by making
distinctive sounds, as a particular squeak for “nuts,” or by applying
words to designate objects, as “half-past-two” for the name of the
coachman)—such invention often clearly having an onomatopoetic origin,
though likewise often wholly arbitrary.

       *       *       *       *       *

I will now conclude this chapter by detailing evidence to show the
extent to which, under favourable circumstances, young children will
thus likewise invent arbitrary signs, which, however, for reasons
already mentioned, are here almost invariably of an articulate kind.
It would be easy to draw this evidence from sundry writers on the
psychogenesis of children; but it will be sufficient to give a few
quotations from an able writer who has already taken the trouble to
collect the more remarkable instances which have been recorded of the
fact in question. The writer to whom I allude is Mr. Horatio Hale, and
the paper from which I quote is published in the _Proceedings of the
American Association for the Advancement of Science_, vol. xxxv., 1886.

“In the year 1860 two children, twin boys, were born in a respectable
family residing in a suburb of Boston. They were in part of German
descent, their mother’s father having come from Germany to America at
the age of seventeen; but the German language, we are told, was never
spoken in the household. The children were so closely alike that their
grandmother, who often came to see them, could only distinguish them by
some coloured string or ribbon tied around the arm. As often happens in
such cases, an intense affection existed between them, and they were
constantly together. The remainder of their interesting story will be
best told in the words of the writer, to whose enlightened zeal for
science we are indebted for our knowledge of the facts.

“At the usual age these twins began to talk, but, strange to say, not
their ‘mother-tongue.’ They had a language of their own, and no pains
could induce them to speak anything else. It was in vain that a little
sister, five years older than they, tried to make them speak their
native language—as it would have been. They persistently refused to
utter a syllable of English. Not even the usual first words, ‘papa,’
‘mamma,’ ‘father,’ ‘mother,’ it is said, did they ever speak; and,
said the lady who gave this information to the writer,—who was an
aunt of the children, and whose home was with them,—they were never
known during this interval to call their mother by that name. They had
their own name for her, but never the English. In fact, though they
had the usual affections, were rejoiced to see their father at his
returning home each night, playing with him, &c., they would seem to
have been otherwise completely taken up, absorbed with each other....
The children had not yet been to school; for, not being able to speak
their ‘own English,’ it seemed impossible to send them from home. They
thus passed the days, playing and talking together in their own speech,
with all the liveliness and volubility of common children. Their accent
was German—as it seemed to the family. They had regular words, a few
of which the family learned sometimes to distinguish; as that, for
example, for carriage, which, on hearing one pass in the street, they
would exclaim out, and run to the window. This word for carriage, we
are told in another place, was ‘ni-si-boo-a,’ of which, it is added,
the syllables were sometimes so repeated that they made a much longer
word.”

The next case is quoted by Mr. Hale from Dr. E. R. Hun, who recorded
it in the _Monthly Journal of Psychological Medicine_, 1868.

“The subject of this observation is a girl aged four and a half years,
sprightly, intelligent, and in good health. The mother observed, when
she was two years old, that she was backward in speaking, and only
used the words ‘papa’ and ‘mamma.’ After that she began to use words
of her own invention, and though she understood readily what she said,
never employed the words used by others. Gradually she enlarged her
vocabulary until it has reached the extent described below. She has
a brother eighteen months younger than herself, who has learned her
language, so that they can talk freely together. He, however, seems
to have adopted it only because he has more intercourse with her than
the others; and in some instances he will use a proper word with his
mother, and his sister’s word with her. She, however, persists in
using only her own words, though her parents, who are uneasy about
her peculiarity of speech, make great efforts to induce her to use
proper words. As to the possibility of her having learned these words
from others, it is proper to state that her parents are persons of
cultivation, who use only the English language. The mother has learned
French, but never uses the language in conversation. The domestics,
as well as the nurses, speak English without any peculiarities, and
the child has heard even less than usual of what is called baby-talk.
Some of the words and phrases have a resemblance to the French; but
it is certain that no person using that language has frequented the
house, and it is doubtful whether the child has on any occasion heard
it spoken. There seems to be no difficulty about the vocal organs. She
uses her language readily and freely, and when she is with her brother
they converse with great rapidity and fluency.

“Dr. Hun then gives the vocabulary, which, he states, was such as he
had ‘been able at different times to compile from the child herself,
and especially from the report of her mother.’ From this statement we
may infer that the list probably did not include the whole number of
words in this child-language. It comprises, in fact, only twenty-one
distinct words, though many of these were used in a great variety of
acceptations, indicated by the order in which they were arranged, or by
compounding them in various ways....

“Three or four of the words, as Dr. Hun remarks, bear an evident
resemblance to the French, and others might, by a slight change,
be traced to that language. He was unable, it will be seen, to say
positively that the girl had never heard the language spoken; and it
seems not unlikely that, if not among the domestics, at least among
the persons who visited them, there may have been one who amused
herself, innocently enough, by teaching the child a few words of
that tongue. It is, indeed, by no means improbable that the peculiar
linguistic instinct may thus have been first aroused in the mind
of the girl, when just beginning to speak. Among the words showing
this resemblance are _feu_ (pronounced, we are expressly told, like
the French word), used to signify ‘fire, light, cigar, sun;’ _too_
(the French ‘tout’), meaning ‘all, everything;’ and _ne pa_ (whether
pronounced as in French, or otherwise, we are not told), signifying
‘not.’ _Petee-petee_, the name given to the boy by his sister, is
apparently the French ‘petit,’ little; and _ma_, ‘I,’ may be from the
French ‘moi,’ ‘me.’ If, however, the child was really able to catch
and remember so readily these foreign sounds at such an early age,
and to interweave them into a speech of her own, it would merely show
how readily and strongly in her case the language-making faculty was
developed.

“Of words formed by imitation of sounds, the language shows barely a
trace. The mewing of the cat evidently suggested the word _mea_, which
signified both ‘cat’ and ‘furs.’ For the other vocables which make up
this speech, no origin can be conjectured. We can merely notice that in
some of the words the liking which children and some races of men have
for the repetition of sounds is apparent. Thus we have _migno-migno_,
signifying ‘water, wash, bath;’ _go-go_, ‘delicacies, as sugar, candy,
or dessert,’ and _waia-waiar_, ‘black, darkness, or a negro.’ There
is, as will be seen from these examples, no special tendency to
the monosyllabic form. _Gummigar_, we are told, signifies ‘all the
substantials of the table, such as bread, meat, vegetables, &c.;’ and
the same word is used to designate the cook. The boy, it is added, does
not use this word, but uses _gna-migna_, which the girl considers as
a mistake. From which we may gather that even at their tender age the
form of their language had become with them an object of thought; and
we may infer, moreover, that the language was not invented solely by
the girl, but that both the children contributed to frame it.

“Of miscellaneous words may be mentioned _gar_, ‘horse;’ _deer_,
‘money of any kind;’ _beer_, ‘literature, books, or school;’ _peer_,
‘ball;’ _bau_, ‘soldier, music;’ _odo_, ‘to send for, to go out, to
take away;’ _keh_, ‘to soil;’ _pa-ma_, ‘to go to sleep, pillow, bed.’
The variety of acceptations which each word was capable of receiving
is exemplified in many ways. Thus _feu_ might become an adjective, as
_ne-pa-feu_, ‘not warm.’ The verb _odo_ had many meanings, according to
its position or the words which accompanied it. _Ma odo_, ‘I (want to)
go out;’ _gar odo_, ‘send for the horse;’ _too odo_, ‘all gone.’ _Gaan_
signified God; and we are told—When it rains, the children often run
to the window, and call out, _Gaan odo migno-migno, feu odo_, which
means, ‘God take away the rain, and send the sun’—_odo_ before the
object meaning ‘to take away,’ and after the object, ‘to send.’ From
this remark and example we learn, not merely that the language had—as
all real languages must have—its rules of construction, but that these
were sometimes different from the English rules. This also appears in
the form _mea waia-waiaw_, ‘dark furs’ (literally, ‘furs dark’), where
the adjective follows its substantive.

“The odd and unexpected associations which in all languages govern
the meaning of words are apparent in this brief vocabulary. We can
gather from it that the parents were Catholics, and punctual in church
observances. The words _papa_ and _mamma_ were used separately in
their ordinary sense; but when linked together in the compound term
_papa-mamma_, they signified (according to the connection, we may
presume), ‘church,’ ‘prayer-book,’ ‘cross,’ ‘priest,’ ‘to say their
prayers.’ _Bau_ was ‘soldier;’ but, we are told, from seeing the bishop
in his mitre and vestments, thinking he was a soldier, they applied the
word _bau_ to him. _Gar odo_ properly signified ‘send for the horse;’
but as the children frequently saw their father, when a carriage was
wanted, write an order and send it to the stable, they came to use the
same expression (_gar odo_) for pencil and paper.

“There is no appearance of inflection, properly speaking, in the
language; and this is only what might be expected. Very young children
rarely use inflected forms in any language. The English child of three
or four years says, ‘Mary cup,’ for ‘Mary’s cup;’ and ‘Dog bite Harry’
will represent every tense and mood. It is by no means improbable that,
if the children had continued to use their own language for a few years
longer, inflections would have been developed in it, as we see that
peculiar forms of construction and novel compounds—which are the germs
of inflection—had already made their appearance.

“These two recorded instances of child-languages have led to further
inquiries, which, though pursued only for a brief period, and in a
limited field, have shown that cases of this sort are by no means
uncommon.”

The author then proceeds to furnish other corroborative instances; but
the above quotations are, I think, sufficient for my purposes.[88] For
they show (1) that the spontaneous and to all appearances arbitrary
word-making, which is more or less observable in all children when
first beginning to speak, may, under favourable circumstances, proceed
to an astonishing degree of fulness and efficiency; (2) that although
the words, or articulate signs, thus invented are sometimes of a
plainly onomatopoetic origin, as a general rule they are not so; (3)
that the words are far from being always monosyllabic; (4) that they
admit of becoming sufficiently numerous and varied to constitute a not
inefficient language, without as yet having advanced to the inflexional
stage; and (5) that the syntax of this language presents obvious points
of resemblance to that of the gesture-languages of mankind previously
considered.




CHAPTER VIII.

RELATION OF TONE AND GESTURE TO WORDS.


We have already seen that spoken language differs from the language
of tone and gesture in being, as a system of signs, more purely
conventional. This means that for semiotic purposes articulation is
a higher product of mental evolution than either gesticulation or
intonation. It also means that as an instrument of such evolution
articulate speech is more efficient. The latter point is an important
one, so I shall proceed to deal with it at some length.

As noticed in a previous chapter, our system of coinage, bank-notes,
and bills of sale is a more convenient system of signifying value of
labour or of property, than is the more primitive and less conventional
system of actually exchanging the labour or bartering the property;
and our system of arithmetic is similarly more convenient for the
purpose of calculation than is the more natural system of counting
on the fingers. But not only are these more conventional systems
more convenient; they are likewise conducive to a higher development
of business transactions on the one hand, and of calculation on the
other. In the absence of such an improved system of signs, it would
be impossible to conduct as many or such intricate transactions and
calculations as we do conduct. Similarly with speech as distinguished
from gesture. Words, like gestures, are signs of thoughts and feelings;
but in being more conventional they are more pure as signs, and so
admit of being wrought up into a much more convenient or efficient
system, while at the same time they become more constructive in their
influence upon ideation. The great superiority of words over gestures
in both these respects may most easily be shown by the use of a few
examples.

I open Colonel Mallery’s book at random, and find the following as the
sign for a barking dog:—

“Pass the arched hand forward from the lower part of the face, to
illustrate elongated nose and mouth; then, with both forefingers
extended, remaining fingers and thumbs closed, place them upon either
side of the lower jaw, pointing upwards, to show lower canines, at the
same time accompanying the gesture with an expression of withdrawing
the lips so as to show the teeth snarling; then, with the fingers of
the right hand extended and separated throw them quickly forward and
slightly upward (voice or talking).”

Here, be it observed, how elaborate is this pictorial method of
designating a dog barking as compared with the use of two words; and
after all it is not so efficient, for the signs were misunderstood
by the Indians to whom they were shown—the meaning assigned to them
being that of a growling bear. What a large expenditure of thought is
required for the devising and the interpretation of such ideograms!
and, when they are formed and understood, how cumbersome do they
appear if contrasted with words! Colonel Mallery, indeed, says of
gesture-language that, “when highly cultivated, its rapidity on
familiar subjects exceeds that of speech, and approaches to that of
thought itself;” but, besides the important limitation “on familiar
subjects,” he adds,—“at the same time it must be admitted that great
increase in rapidity is chiefly obtained by the system of preconcerted
abbreviations before explained, and by the adoption of arbitrary forms,
in which naturalness is sacrificed and conventionality established.”[89]

But besides being cumbersome, gesture-language labours under the
more serious defect of not being so precise, and the still more
serious defect of not being so serviceable as spoken language in
the development of abstraction. We have previously seen how words,
being more or less purely conventional as signs, are not tied down,
as it were, to material objects; although they have doubtless all
originated as expressive of sensuous perceptions, not being necessarily
ideographic, they may easily pass into signs of general ideas, and
end by becoming expressive of the highest abstractions. “Words are
thus the easily manipulated counters of thought,” and so, to change
the metaphor, are the progeny of generalization. But gestures, in
being always more or less ideographic, are much more closely chained
to sensuous perceptions; and, therefore, it is only when exercised on
“familiar subjects” that they can fairly be said to rival words as a
means of expression, while they can never soar into the thinner medium
of high abstraction. No sign-talker, with any amount of time at his
disposal, could translate into the language of gesture a page of Kant.

Let it be observed that I am here speaking of gesture-language as we
actually find it. What the latent capabilities of such language may be
is another question, and one with reference to which speculation is
scarcely calculated to prove profitable. Nevertheless, as the subject
is not altogether without importance in the present connection, I may
quote the following brief passage from a recent essay by Professor
Whitney. After remarking that “the voice has won to itself the chief
and almost exclusive part in communication,” he adds:—

“This is not in the least because of any closer connection of the
thinking apparatus with the muscles that act to produce audible
sounds than with those that act to produce visible motions; not
because there are natural uttered names for conceptions, any more
than natural gestured names. It is simply a case of ‘survival of the
fittest,’ or analogous to the process by which iron has become the
exclusive material of swords, and gold and silver for money: because,
namely, experience has shown this to be the material best adapted
to this special use. The advantages of the voice are numerous and
obvious. There is first its economy, as employing a mechanism that is
available for little else, and leaving free for other purposes those
indispensable instruments, the hands. Then there is its superior
perceptibleness; its nice differences impress themselves upon the sense
at a distance at which visible motions become indistinct; they are not
hidden by intervening objects; they allow the eyes of the listeners as
well as the hands of the speaker to be employed in other useful work;
they are as plain in the dark as in the light; and they are able to
catch and command the attention of one who is not to be reached in any
other way.”[90]

To these advantages we may add that words, in being as we have seen
less essentially ideographic than gestures, must always have been
more available for purposes of abstract expression. We must remember
how greatly gesture-language, as it now appears in its most elaborate
form, is indebted to the psychologically constructing influence of
spoken language; and, thus viewed, it is a significant fact that
even now gesture language is not able to convey ideas of any high
degree of abstraction. Still, I doubt not it would be possible to
construct a wholly conventional system of gestures which should answer
to, or correspond with, all the abstract words and inflections of
a spoken language; and that then the one sign-system might replace
the other—just as the sign-system of writing is able similarly to
replace that of speech. This, however, is a widely different thing from
supposing that such a perfect system of gesture-signs could have grown
by a process of natural development; and, looking to the essentially
ideographic character of such signs, I greatly question whether, even
under circumstances of the strongest necessity (such as would have
arisen if man, or his progenitors, had been unable to articulate), the
language of gesture could have been developed into anything approaching
a substitute for the language of words.

It may tend to throw some light on this hypothetical question—which
is of some importance for us—if we consider briefly the psychological
_status_ of wholly uneducated deaf-mutes; for although it is true that
their case is not fairly parallel to that of a human race destitute
of the faculty of speech (seeing that the individual deaf-mute does
not find any elaborate system of signs prepared for him by the
exertions of dumb ancestors, as would doubtless have been the case
under the circumstances supposed), still, on the other hand, and as
a compensating consideration, we must remember that the individual
deaf-mute not only inherits a human brain, the structure of which has
been elaborated by the speech of his ancestors, but is also surrounded
by a society the whole structure of whose ideation is dependent upon
speech. So far, therefore, as the complex conditions of the question
admit of being disentangled, the case of uneducated deaf-mutes living
in a society of speaking persons affords the best criterion we can
obtain of the prospect which gesture-language would have had as a
means of thought-formation in the human race, supposing this race to
have been destitute of the faculty of speech. To show, therefore, the
psychological condition of an individual thus circumstanced, I will
quote a brief passage from a lecture of my own, which was given before
the British Association in 1878.

“It often happens that deaf and dumb children of poor parents are so
far neglected that they are never taught finger-language, or any other
system of signs, whereby to converse with their fellow-creatures. The
consequence, of course, is that these unfortunate children grow up
in a state of intellectual isolation, which is almost as complete as
that of any of the lower animals. Now, when such a child grows up and
falls into the hands of some competent teacher, it may of course be
educated, and is then in a position to record its experiences when in
its state of intellectual isolation. I have therefore obtained all the
evidence I can as to the mental condition of such persons, and I find
that their testimony is perfectly uniform. In the absence of language,
the mind is able to think in the logic of feelings; but can never
rise to any ideas of higher abstraction than those which the logic of
feelings supplies. The uneducated deaf-mutes have the same notions of
right and wrong, cause and effect, and so on, as we have already seen
that animals and idiots possess. They always think in the most concrete
forms, as shown by their telling us (when educated) that so long as
they were uneducated they always thought in pictures. Moreover, that
they cannot attain to ideas of even the lowest degree of abstraction,
is shown by the fact that in no one instance have I been able to find
evidence of a deaf-mute who, prior to education, had evolved for
himself any form of supernaturalism. And this, I think, is remarkable,
not only because we might fairly suppose that some rude form of
fetishism, or ghost-worship, would not be too abstract a system for the
unaided mind of a civilized man to elaborate; but also because the mind
in this case is _not_ wholly unaided. On the contrary, the friends of
the deaf-mute usually do their utmost to communicate to his mind some
idea of whatever form of religion they may happen to possess. Yet it is
uniformly found that, in the absence of language, no idea of this kind
can be communicated. For instance, the Rev. S. Smith tells me that one
of his pupils, previous to education, supposed the Bible to have been
printed by a printing-press in the sky, which was worked by printers
of enormous strength—this being the only interpretation the deaf-mute
could assign to the gestures whereby his parents had sought to make him
understand, that they believed the Bible to contain a revelation from
a God of power who lives in heaven. Similarly, Mr. Graham Bell informs
me of another, though similar case, in which the deaf-mute supposed the
object of going to church to be that of doing obeisance to the clergy.”

To the same effect Mr. Tylor says, in the passage already quoted,
that deaf-mutes cannot form ideas of any save the lowest degree of
abstraction, and further on he gives some interesting illustrations of
the fact. Thus, for instance, a deaf-mute who had been educated said
that before his instruction his fingers had taught him his numbers,
and that when the number was over ten, he made notches on a piece of
wood. Here we see the inherited capability of numerical computation
united with the crudest form of numerical notation, or symbolism. And
so in all other cases of deaf-mutes before instruction; they present
an inherited capacity of abstract ideation, and yet do not find their
sign-language of much service in assisting them to develop this
capacity: it is too essentially pictorial to go far beyond the region
of sensuous perception.

Thus, on the whole, although I deem it profitless to speculate on what
the language of gestures might have become in the absence of speech,
I think it is highly questionable whether it would have reached any
considerable level of excellence; and I think it is not improbable
that, in the absence of articulation, the human race would not have
made much psychological advance upon the anthropoid apes. For we must
never forget the important fact that thought is quite as much the
effect as it is the cause of language, whether of speech or of gesture;
and seeing how inferior gesture is to speech as a system of language,
especially in regard to precision and abstraction, I do not think it
probable that, in the absence of speech, gesture alone would have
supplied the exact and delicate conditions which are essential to the
growth of any highly elaborate ideation.

The next point which I desire to consider is that, although gesture
language is not in my opinion so efficient a means of developing
abstract ideation as is spoken language, it must nevertheless have been
of much service in assisting the growth of the latter, and so must
have been of much service in laying the foundation of the whole mental
fabric which has been constructed by the faculty of speech. Whether we
look to young children, to savages, or in a lesser degree to idiots, we
find that gesture plays an important part in assisting speech; and in
all cases where a vocabulary is scanty or imperfect, gesture is sure to
be employed as the natural means of supplementing speech. Therefore,
supposing speech to have had a natural mode of genesis, it is, in my
opinion, perfectly certain that its origin and development must have
been greatly assisted by gesture. In subsequent chapters I will adduce
direct evidence upon this head. At present I wish to draw attention
to another point. This is, that although gesture psychologically
precedes speech, when once articulate sounds have been devised for the
expression of ideas, the faculty of using these articulate sounds as
signs of their corresponding ideas does not involve the presence of a
higher psychological development than does the faculty of using tones
and gestures for the conveyance of similar ideas.

As already shown, it is a matter of observable fact that the only
animals which are able to articulate are able to employ nouns,
adjectives, and verbs, as expressive of concrete ideas; while animals
which are not able to articulate similarly employ tones, and in many
cases are able to understand words. Therefore, it is a matter of
observable fact that the psychological level required for using tones
as vocal gestures, understanding words as expressive of simple ideas,
and even uttering words with a correct appreciation of their meaning,
is a level not higher than that which obtains in some existing animals.

If we turn from animals to man, we find the same truth exemplified.
For in the descending grade of human intelligence as exhibited by
idiots, we see that while the use of simple gestures as signs occurs
in idiots somewhat too low in the scale to utter any articulate words,
nevertheless the interval between such an idiot and one capable
of uttering the simplest words is a short interval. Again, in the
ascending grade of human intelligence, as exhibited by the growing
child, we find the same observation to apply; although, on account
of some children requiring a longer time than others to develop the
_mechanique_ of articulation, we might by considering their cases alone
over-estimate the psychological interval which separates gesticulation
from speech.[91]

Thus all the evidence at our disposal goes to show that, while the
language of tone and gesture is distinctive, in its least-developed
form, of a comparatively low grade of mental evolution, in all but
its least-developed form it is not thus distinctive; for as soon as
the language of gesture becomes in the smallest degree conventional,
so soon is the psychological level sufficiently high to admit of the
use of articulate sounds, vocal gestures, or words expressive of
concrete ideas—always supposing that these are already supplied by the
psychological environment. Whether or not articulate sounds are then
actually made depends, of course, on conditions of a purely anatomical
kind.

And here it may be as well to remember the point previously mentioned,
namely, that although no existing quadrumanous animal has shown itself
able to articulate, we may be quite sure that this fact depends on
anatomical as distinguished from psychological conditions; for not
only are the higher monkeys much more intelligent than talking birds,
but they are likewise much more imitative of human gestures; and for
both these reasons they are the animals which, more than any others,
would be psychologically apt to learn the use of words from man, were
it not for some accident of anatomy which stands in the way of their
uttering them. And in this connection it is worth while to bear in mind
the remark of Professor Huxley, that an imperceptibly small difference
of innervation, or other anatomical character of the parts concerned,
might determine or prevent the faculty of making articulate sounds.

       *       *       *       *       *

Looking to the direction in which my argument is tending, this appears
to be the most convenient place to dispose of a criticism that is
not unlikely to arise. It may be suggested, by way of objection to my
views, that if all the foregoing discussion is accepted as paving the
way to the conclusion that human intelligence has been developed from
animal intelligence, the discussion itself is proving too much. For, if
animals possess in so conspicuous a degree the germ of the sign-making
faculty, why, it may be asked, has this germ been developed only in the
case of our own ancestors?

In answer to this question I must begin by reminding the reader,
that during the course of the present chapter I have endeavoured to
make good the following positions. First, that in the absence of
articulation, or of the power of forming verbal signs, the faculty of
language is not likely to have made much advance in the animal kingdom.
Second, seeing that words are essentially less ideographic, as well
as more precise than gestures—and, therefore, more available for the
purpose both of expressing and constructing abstract ideas,—I do not
think it is probable that in the absence of articulation the human
race would have made much psychological advance upon the anthropoid
apes. Third, that although gesture language is not so efficient a
means of developing abstract ideation as is articulate language, it
must nevertheless have been of much service in assisting the growth
of the latter; so that where the power of articulation was present,
both systems of sign-making would have co-operated in the development
of abstract thought: in the presence of articulation, gestures would
themselves gain additional influence in this respect.

From these data there follows the important consequence that only from
some species of ape which possessed the requisite anatomical conditions
could the human mind have taken its origin. In other words, the above
considerations are adduced to show the futility of arguing that, if
the human mind has been developed in virtue of the sign-making faculty
as this is exemplified in speech, we might therefore have expected
that from the same starting-point (namely, the anthropoid apes)
some comparably well-elaborated mind should have been developed in
virtue of the sign-making faculty as this is exemplified in gesture.
I maintain that we can see very good reason why (even if we suppose
all the other conditions parallel) the branch of the Primates which
presented the power—or the potentiality—of articulation should have
been able to rise in the psychological scale, as we evolutionists
believe that it has risen; while all the companion branches, being
restricted in their language to gesture, should have remained in their
original condition.

To this it may be answered that the talking birds might be looked to
as the possible—or even probable—rivals of articulating mammals in
respect of potential intelligence; and, therefore, that according to
the views which I am advocating, it might have been expected that there
should now be existing upon the earth some race of bird-like creatures
ready to dispute the supremacy of man.

This, however, would be a very shallow criticism. The veriest tyro in
natural science is aware that, if there is any truth at all in the
general theory of descent, we are everywhere compelled to see that
the conditions which determine the development of a species in any
direction are always of a complex character. Why one species should
remain constant through inconceivably enormous lapses of geological
time, while others pass through a rich and varied history of upward
change—why this should be so in any case we cannot say. We can only
say, in general terms, that the conditions which in any case determine
upward growth or stationary type are too numerous and complex to admit
of our unravelling them in detail. Now, if this is the case even as
between the structures of allied types—where there may be nothing
to indicate the difference of the conditions which have led to the
difference of results,—much more must it be the case between animals
so unlike as a parrot and an ape. I think he would be a bold man who
would affirm that even if the orang-outang had been able to articulate,
this ape would necessarily, or probably, have become the progenitor of
another human race. Absurd, then, it is to argue that, if the human
race sprang from some other species of man-like creature, and became
human in virtue of the power of articulation _plus_ all the other
conditions external and internal, therefore the talking birds ought
to have developed some similar progeny, merely because they happen to
satisfy one of these conditions.

Take a fair analogy. Flying is no doubt a very useful faculty to all
animals which present it, and it is shown to be mechanically possible
in animals so unlike one another as Insects, Reptiles, Birds, and
Mammals. We might therefore suppose that, from the fact of bats being
able to fly, many other mammals should have acquired the art. But, as
they have not done so, we can only say that the reason is because the
complex conditions leading to the growth of this faculty have been
satisfied in the bats alone. Similarly “the flight of thought” is a
most useful faculty, and it has only been developed in man. One of the
conditions required for its development—power of articulation—occurs
also in a few birds. But to argue from this that these birds ought to
have developed the faculty of thought, would be just as unwarrantable
as to argue that some other mammals ought to have developed the faculty
of flight, seeing that they all present the most important of the
needful conditions—to wit, bones and muscles actuated by nerves.
Indeed, the argument would be even more unwarranted than this; for we
can see plainly enough that the most important conditions required
for the development of thought are of a psychological and social
kind—those which are merely anatomical being but of secondary value,
even though, as I have endeavoured to indicate, they are none the less
indispensable.

In short, I am not endeavouring to argue that the influence of
articulation on the development of thought is in any way _magical_.
Therefore, the mere fact that certain birds are able to make articulate
sounds in itself furnishes no more difficulty to my argument than the
fact that they are able to imitate a variety of other sounds. For the
_psychological_ use of articulate sounds can only be developed in the
presence of many other and highly complex conditions, few if any of
which can be shown to obtain among birds. If any existing species
of anthropoid ape had proved itself capable of imitating articulate
sounds, there might have been a little more force in the apparent
difficulty; though even in that case the argument would not have been
so strong as in the above parallel with regard to the great exception
furnished by bats in the matter of flight.

So far, then, as we have yet gone, I do not anticipate that opponents
wall find it prudent to take a stand. Seeing that monkeys use their
voices more freely than any other animals in the way of intentionally
expressive intonation; that all the higher animals make use of gesture
signs; that denotative words are (psychologically considered) nothing
more than vocal gestures; that, if there is any psychological interval
between simple gesticulation and denotative articulation, the interval
is demonstrably bridged in the case alike of talking birds, infants,
and idiots;—seeing all these things, it is evident that opponents
of the doctrine of mental evolution must take their stand, not on
the faculty of _articulation_, but on that of _speech_. They must
maintain that the mere power of using denotative words implies no
real advance upon the power of using denotative gestures; that it
therefore establishes nothing to prove the possibility, or even the
probability, of articulation arising out of gesticulation; that their
position can only be attacked by showing how a sign-making faculty,
whether expressed in gesticulation or in articulation, can have become
developed into the faculty of predication; that, in short, the fortress
of their argument consists, not in the power which man displays of
using denotative words, but in his power of constructing predicative
propositions. This central position, therefore, we must next attack.
But, before doing so, I will close the present chapter by clearly
defining the exact meanings of certain terms as they will afterwards be
used by me.

By the _indicative_ stage of language, or sign-making, I will
understand the earliest stage that is exhibited by intentional
sign-making. This stage corresponds to the divisions marked four
and six in my representative scheme (p. 88), and, as we have now
so fully seen, is common to animals and human beings. Indicative
signs, then, whether in the form of gestures, tones, or words, are
intentionally significant. For the most part they are expressive of
emotional states, and simple desires. When, for example, an infant
holds out its arms to be taken by the nurse, or points to objects
in order to be taken to them, it cannot be said to be _naming_
anything; yet it is clearly _indicating_ its wants. Infants also cry
_intentionally_, or as a partly conventional sign to show discomfort,
whether bodily or mental.[92] They will likewise at an early age learn
wholly conventional signs whereby to indicate—though not yet to
name—particular feelings, objects, qualities, and actions. My son, for
instance, was taught by his nurse to shake his head for “No,” nod it
for “Yes,” and wave his hand for “Ta-ta,” or leave-taking: all these
indicative gestures he performed well and appropriately when eight and
a half months old. This indicative stage of language, or sign-making,
is universally exhibited by all the more intelligent animals, although
not to so great an extent as in infants. The parrot which depresses its
head to invite a scratching, the dog which begs before a wash-stand,
the cat which pulls one’s clothes to solicit help for her kittens in
distress—all these animals are making what I call _indicative_ signs.

Following upon the indicative stage of language there is what I have
called _denotative_ (7 A in the scheme on p. 88). This likewise occurs
both in animals and in children when first beginning to speak—talking
birds, for instance, being able to learn and correctly use names as
_notæ_, or marks, of particular objects, qualities, and actions. Yet
such _notæ_—be they verbal or otherwise—thus learned by special
association, are not, strictly speaking, _names_. By the use of such
a sign the talking bird merely affixes a vocal mark to a particular
object, quality, or action: it does not _extend_ the sign to any
other similar objects, qualities, or actions of the same class; and,
therefore, by its use of that sign does not really _connote_ anything
of the particular object, quality, or action which it _denotes_.

So much, then, for signs as _denotative_. By signs as _connotative_,
I mean signs which are in any measure _attributive_. If we call a dog
Jack, that is a denotative name: it does not attribute any quality as
belonging to that dog. But if we call the animal “Smut,” or “Swift,” or
by any other word serving to imply some quality which is distinctive
of that dog, we are thereby connoting of the dog the fact of his
presenting such a quality. Connotative names, therefore, differ from
denotative, in that they are not merely _notæ_ or marks of the things
named, but also imply some character, or characters, as belonging
to those things. And the character, or characters, which they thus
imply, by the mere fact of implication, assign the things named to a
_group_: hence these connotative names are _con-notæ_, or the marking
of one thing _along with_ another—_i.e._ express an act of nominative
_classification_. This is an important fact to remember, because, as
we shall afterwards find, all connotative terms arise from the need
which we experience of thus verbally classifying our perceptions of
likeness or analogy. Moreover, it is of even still more importance
to note that such verbal classification may be either receptual or
conceptual. For instance, the first word (after _Mamma_, _Papa_,
&c.) that one of my children learnt to say was the word _Star_. Soon
after having acquired this word, she extended its signification to
other brightly shining objects, such as candles, gas-lights, &c. Here
there was plainly a perception of likeness or analogy, and hence the
term _Star_, from having been originally denotative, began to be also
connotative. But this connotative extension of the term must evidently
have been what I term receptual. For it is impossible to suppose that
at that tender age the child was capable of thinking about the term
_as_ a term, or of setting the term before the mind as an object of
thought, distinct from the object which it served to name. Therefore,
we can only suppose that the extension of this originally denotative
name (whereby it began to be connotative) resembled the case of a
similar extension mentioned in the last chapter, where my parrot raised
its originally denotative sign for a particular dog to an incipiently
connotative value, by applying that sign to all other dogs. That is to
say, both in the case of the child and the bird, connotation within
these moderate limits was rendered possible by means of receptual
ideation alone. But, with advancing age and developing powers, the
human mind attains to conceptual ideation; and it is then in a position
to constitute the names which it uses _themselves objects of thought_.
The consequence is that connotation may then no longer represent the
merely spontaneous expression of likeness receptually perceived: it
may become the intentional expression of likeness conceptually thought
out. In the mind of an astronomer the word _Star_ presents a very
different mass of connotative meaning from that which it presented to
the child, who first extended it from a bright point in the sky to a
candle shining in a room. And the reason of this great difference is,
that the conceptual thought of the astronomer, besides having greatly
_added_ to the connotation, has also greatly _improved_ it. The only
common quality which the name served to connote when used by the child
was that of brightness; but, although the astronomer is not blind to
this point of resemblance between a star and a candle, he disregards
it in the presence of fuller knowledge, and will not apply the term
even to objects so much more closely resembling a star as a comet or a
meteor. Now, this greater _accuracy_ of connotation, quite as much as
the greater _mass_ of it, has been reached by the astronomer in virtue
of his powers of conceptual thought. It is because he has thought about
his names _as_ names that he has thus been able with so much accuracy
to define their meanings—_i.e._ to limit their connotations in some
directions, as well as to extend them in others.

Obviously, therefore, we are here in the presence of a great
distinction, and one which needs itself to be in some way connoted.
It is, indeed, but a special exhibition of the one great distinction
which I have carried through the whole course of this work—namely,
that between ideation as receptual and conceptual. But it is none the
less important to designate this special exhibition of it by means of
well-defined terms; and I can only express surprise that such should
not already have been done by logicians. The terms which I shall use
are the following.

By a connotative name I will understand the connotative extension of
a denotative name, whether such extension be great or small, and,
therefore, whether it be extended receptually or conceptually. But
for the _exclusively conceptual_ extension of a name I will reserve
the convenient term _denomination_. This term, like those previously
defined, was introduced by the schoolmen, and by them was used as
synonymous with connotation. But it is evident that they (and all
subsequent writers) only had before their minds the case of conceptual
connotation, and hence they felt no need of the distinction which
for present purposes it is obviously imperative to draw. Now, I do
not think that any two more appropriate words could be found whereby
to express this distinction than are these words _connotation_ and
_denomination_, if for the purposes of my own subsequent analysis I am
allowed to define them in accordance with their etymology. For, when so
defined, a connotative sign will mean a _classificatory_ sign, whether
conferred receptually or conceptually; while a denominative sign will
mean a connotative sign which has been conferred as such _with a truly
conceptual intention_—_i.e._ with an introspective appreciation of its
function as all that logicians understand by a _name_.

I will now sum up these sundry definitions.

By an _indicative_ sign I will understand a significant tone or
gesture intentionally expressive of a mental state; but yet not in any
sense of the word denominative.

By a _denotative_ sign I will understand the receptual marking of
particular objects, qualities, actions, &c.

By a _connotative_ sign I will understand the classificatory
attribution of qualities to objects named by the sign, whether such
attribution be due to receptual or to conceptual operations of the mind.

By a _denominative_ sign I will understand a connotative sign
consciously bestowed as such, or with a full conceptual appreciation of
its office and purpose as a name.

By a _predicative_ sign I will mean a proposition, or the conceptual
apposition of two denominative terms, expressive of the speaker’s
intention to connote something of the one by means of the other.




CHAPTER IX.

SPEECH.


We are now coming to close quarters with our subject. All the
foregoing chapters have been arranged with a view to preparing the
way for what is hereafter to follow; and, therefore, as already
remarked, I have thus far presented material over which I do not
think it is possible that any dispute can arise. But now we come to
that particular exhibition of the sign-making faculty which not only
appears to be peculiar to man, but which obviously presents so great
an advance upon all the lower phases hitherto considered, that it is
the place where my opponents have chosen to take their stand. When a
man maintains that there is a difference of kind between animal and
human intelligence, he naturally feels himself under some obligation
to indicate the point where this difference obtains. To say that it
obtains with the appearance of language, in the sense of sign-making,
is obviously too wide a statement; for, as we have now so fully seen,
language, in this widest sense, demonstrably obtains among the lower
animals. Consequently, the line must be drawn, not at language or
sign-making, but at that particular kind of sign-making which we
understand by Speech. Now the distinctive peculiarity of this kind of
sign-making—and one, therefore, which does not occur in any other
kind—consists in predication, or the using of signs as movable
types for the purpose of making propositions. It does not signify
whether or not the signs thus used are words. The gestures of Indians
and deaf-mutes admit, as we have seen, of being wrought up into a
machinery of predication which, for all purposes of practical life,
is almost as efficient as speech. The distinction, therefore, resides
in the intellectual powers; not in the symbols thereof. So that a
man _means_, it matters not by what system of signs he expresses his
meaning: the distinction between him and the brute consists in his
being able to _mean a proposition_. Now, the kind of mental act whereby
a man is thus enabled to mean a proposition is called by psychologists
an act of Judgment. Predication, or the making of a proposition, is
nothing more nor less than the expression of a judgment; and a judgment
is nothing more nor less than the apprehension of whatever meaning it
may be that a proposition serves to set forth. Therefore, it belongs to
the very essence of predication that it should involve a judgment; and
it belongs to the very essence of a judgment that it should admit of
being stated in the form of a proposition.[93]

Lastly, just as this is the place where my opponents take a stand,
so, as they freely allow, it is the only place where they _can_ take
a stand. If once this chasm of speech were bridged, there would be
no further chasm to cross. From the simplest judgment which it is
possible to make, and therefore from the simplest proposition which
it is possible to construct, it is on all hands admitted that human
intelligence displays an otherwise uniform or uninterrupted ascent
through all the grades of excellence which it afterwards presents.
Here, then, and here alone, we have what Professor Max Müller calls
the Rubicon of Mind, which separates the brute from the man, and over
which, it is alleged, the army of Science can never hope to pass.

In order to present the full difficulty which is here encountered, I
will allow it to be stated by the ablest of my opponents. As President
of the Biological Section of the British Association in 1879, Mr.
Mivart expressed his matured thought upon the subject thus:—

“The simplest element of thought seems to me to be a ‘judgment,’
with intuition of reality concerning some ‘fact,’ regarded as a fact
real or ideal. Moreover, this judgment is not itself a modified
imagination, because the imaginations which may give occasion to it
persist unmodified in the mind side by side with the judgment they
have called up. Let us take, as examples, the judgments, ‘That thing
is good to eat,’ and ‘Nothing can be and not be at the same time and
in the same sense.’ As to the former, we vaguely imagine ‘things good
to eat;’ but they must exist _beside_ the judgment, not _in_ it. They
can be recalled, compared, and seen to co-exist. So with the other
judgment, the mind is occupied with certain abstract ideas, though the
imagination has certain vague ‘images’ answering respectively to ‘a
thing being,’ and ‘a thing not being,’ and to ‘at the same time’ and
‘in the same sense;’ but the images do not _constitute_ the judgment
itself, any more than human ‘swimming’ is made up of limbs and fluid,
though without such necessary elements no such swimming could take
place.[94]

“This distinction is also shown by the fact that one and the same idea
may be suggested to, and maintained in, the mind by the help of the
most incongruous images, and very different ideas by the very same
image; this we may see to be the case with such ideas as ‘number,’
‘purpose,’ ‘motion,’ ‘identity,’ &c.

“But the distinctness of ‘thought’ from ‘imagination’ may perhaps
be made clearer by the drawing out fully what we really do when we
make some simple judgment, as, _e.g._, ‘A negro is black.’ Here, in
the first place, we directly and explicitly affirm that there is a
conformity between the external thing, ‘a negro,’ and the external
quality ‘blackness’—the negro possessing that quality. We affirm,
secondarily and implicitly, a conformity between two external entities
and two corresponding internal concepts. And thirdly, and lastly,
we also implicitly affirm the existence of a conformity between the
subjective judgment and the objective existence.”[95]

I will next allow this matter to be presented in the words of another
adversary, and one whom Mr. Mivart approvingly quotes.

“The question is, Can the sense say anything—make a judgment at all?
Can it furnish the blank formula of a judgment—the ‘is’ in ‘A is B’?
The grass of the battlefield was green, and the sense gave both the
grass and the greenness; but did it affirm that ‘the grass is green’?
It may be assumed that ‘grass’ and ‘green’ together form one complex
object, which is an object under space and time, and therefore of
sense. But against this the rejoinder at once is, that the sense may
indeed take in and report (so to speak) a complex object, but that in
this case the question is, not about the complex object, but about the
_complexity_ of the object. It is one thing to see green grass, and
evidently quite another to affirm the _greenness_ of the grass. The
difference is all the difference between seeing two things united,
and seeing them _as united_.... If a brute could think ‘is,’ brute
and man would be brothers. ‘Is,’ as the copula of a judgment, implies
the mental separation, and recombination of two terms that only exist
united in nature, and can therefore never have impressed the sense
except as one thing.[96] And ‘is,’ considered as a substantive verb, as
in the example ‘This man is,’ contains in itself the application of the
copula of judgment to the most elementary of all abstractions—‘thing’
or ‘something.’ Yet if a being has the power of thinking—‘thing,’ it
has the power of transcending space and time by dividing or decomposing
the phenomenally one. Here is the point where instinct ends and reason
begins.”[97]

It would be easy to add quotations from other writers to the same
effect as the above;[98] but these may be held sufficient to give
material for the first stage of my criticism, which is of a purely
technical character. I affirm that all writers who thus take their
stand upon the distinctively human faculty of predication are
taking their stand at the wrong place. In other words, without at
present disputing whether we have to do with a distinction of kind
or of degree, I say, and say confidently, that the distinction in
question—_i.e._ between animal and human intelligence—may be easily
proved to occur further back than at the faculty of predication, or
the forming of a proposition. The distinction occurs at the faculty of
denomination, or the bestowing of a name, known as such. “The simplest
element of thought” is _not_ a “_judgment_:” the simplest element of
thought is a _concept_. That this is the case admits of being easily
demonstrated in several different ways.

In the first place, it is evident that there could be no judgments
without concepts, just as there could be no propositions without terms.
A judgment is the result of a comparison of concepts, and this is the
reason why it can only find expression in a proposition, which sets
forth the relation between the concepts by bringing into apposition
their corresponding terms. Judgments, therefore, are _compounds_ of
thought: the _elements_ are concepts.

In the second place, given the power of conceiving, and the germ of
judgment is implied, though not expanded into the blossom of formal
predication. For whenever we bestow a name we are implicitly judging
that the thing to which we apply the name presents the attributes
connoted by that name, and thus we are virtually predicating the fact.
For example, when I call a man a “Negro,” the very term itself affirms
blackness as the distinctive quality of that individual—just as does
the equivalent nursery term, “Black-man.” To utter the name Negro,
therefore, or the name Black-man, is to form and pronounce at least
two judgments touching an individual object of sensuous perception—to
wit, that it is a man, and that he is black. The judgments so formed
and pronounced are doubtless not so explicit as is the case when both
subject and predicate are associated in the full proposition—“A negro
is black;” but in the single term Negro, or Black-man, both these
elements were already present, and _must_ have been so if the name were
in any degree at all conceptual—_i.e._ _denominative_ as distinguished
from _denotative_. In the illustration “Negro,” or “Black-man,” it so
happens that the connotation of the name is directly given by the
etymology of the name; but this circumstance is immaterial. Whether or
not the etymology of a connotative name happens to fit the particular
subject to which it is applied, the same kind of classificatory
judgment is required for any appropriate application of the same. If,
with Blumenbach, I am accustomed to call a negro an Ethiopian, when I
apply this name to any representative of that race, I am performing
the same mental act as my neighbour who calls him a Negro, or my child
who calls him a Black-man. If it should be said that in all such
cases the act of naming is so immediately due to association that no
demand is made upon the powers of judgment, the admission would be a
dangerous one for my opponents to make, since the same remark would
apply to the full proposition, “That man is black.” Moreover, the
objection admits of being easily disposed of by choosing instances of
naming where associations have not yet been definitively fixed. If I am
travelling in a strange continent, and amid all the unfamiliar flora
there encountered I suddenly perceive a plant which I think I know,
before I name it to my friend as that plant, I would submit it to close
scrutiny—_i.e._ carefully _judge_ its resemblances to the known or
familiar species. In short, all connotative names, when denominatively
applied, betoken acts of judgment, which differ from those concerned in
full predication only as regards the form of their expression. Or, as
Mill very tersely remarks, “whenever the names given to objects convey
any information, that is, whenever they have properly any meaning, the
meaning resides not in what they denote, but in what they connote.”
And although in his elaborate treatment of Names and Propositions
he omits expressly to notice the point now before us, it is clearly
implied in the above quotation. The point is that connotative names
(or denominative terms)[99] are often in themselves of predicative
value; and this point is clearly implied in the above quotation,
because, whenever “names given to objects convey any information,”
the information thus conveyed is virtually predicated: the “meaning”
connoted by the name is affirmed in the mere act of bestowing the
name, which thus in itself becomes a condensed proposition. “It is a
truism of psychology that the terms of a proposition, when closely
interrogated, turn out to be nothing but abbreviated judgments.”[100]

This view of the matter, then, is the only one that can be countenanced
by psychology. It is likewise the only one that can be countenanced
by philology, or the study of language in the making. Of this fact I
will adduce abundant evidence in a subsequent chapter, where it will be
shown, as Professor Max Müller says, that “every name was originally a
proposition.” But at present I am only concerned with one of the most
elementary points of purely psychological analysis, and will therefore
postpone the independent illumination of the whole philosophy of
predication which of late years has been so splendidly furnished by the
comparative study of languages.

From whatever point of view, therefore, we look at the matter, we
are bound to conclude, either that the term “judgment” must be
applied indifferently to the act of denominating and to the act of
predicating, or else, if it be restricted to the latter, that it must
not be regarded as “the simplest element of thought.” And thus we
are led back to the position previously gained while treating of the
Logic of Concepts. For we then found that names are the steps of the
intellectual ladder whereby we climb into higher and higher regions of
ideation; and although our progress is assisted by formal predication,
or discursive thought, this is but the muscular energy, so to speak,
which would in itself be useless but for the rungs already supplied,
and on which alone that energy can be expended. Or, to vary the
metaphor, conceptual names are the ingredients out of which is formed
the structure of propositions; and, in order that this formation should
take place, there must already be in the ingredients that element of
vitality which constitutes the _vis formativa_. Now, this element of
vitality is the element of conceptual ideation, already exhibited in
every denominative term.

Therefore, for the sake at once of clearness and of brevity, I will
hereafter speak of predication as _material_ and _formal_. By material
predication I will mean conceptual denomination, whereby, in the mere
act of bestowing a connotative term, we are virtually predicating
of the thing thus designated some fact, quality, or relation, which
the name bestowed is intended to indicate. By formal predication I
will mean the apposition of denominative terms, with the intention
of setting forth some relation which is thus expressed as subsisting
between them. But, as already observed, I regard this distinction as
artificial. Psychologically speaking, there is no line of demarcation
between these two kinds of predication. Whether I say “Fool,” or
“Thou art a fool,” I am similarly assigning the subject of my remark
to a certain category of men: I am similarly giving expression to my
judgment with regard to the qualities presented by one particular
man. The distinction, then, between what I call material and formal
predication is merely a distinction in rhetoric: as a matter of
psychology there is no distinction at all.

If to all this it should be objected, in accordance with the
psychological doctrines set forth by Mr. Mivart, above quoted, that
a judgment as embodied in a proposition differs from a concept
as embodied in a name in respect of the copula, and therefore in
presenting the idea of existence as existence; I answer, in the first
place, that every concept must necessarily present this idea however
_implicitly_; and, in the next place, that however _explicitly_ it
may be stated as a judgment, it is not of more conceptual value than
that of any other quality belonging to a subject. As regards the first
point, when an object, a quality, an action, &c., is named, it is
thereby abstracted as a distinct creation of thought, separated out
from other things, and made to stand before the mind as a distinct
entity (see Chapter IV.). Therefore, in the very act of naming we are
virtually predicating existence of the thing named: the power to
“think is” is the power concerned in the _formation_ of a _concept_,
not in the _apposing_ of concepts _when formed_. All that is done
in an act of such apposition is to bring together two ideas of two
things already conceived as existing: were it not so there could be
_no-things_ to compare.[101]

And now, as regards the second point, so far is it from being true
that the predication of existence is the essential or most important
feature even of a full or formal proposition, that it is really the
least essential or least important. For existence is the category to
which everything must belong if it is to be judged about at all, and
therefore merely to judge that _A is_ and _B is_, is to form the most
barren (or least significant) judgment that can be formed with regard
to A or B; and when we bring these two judgments (concepts) together in
the proposition _A is B_, the new judgment which we make has nothing to
do with the existence either of A or of B, nor has it really anything
to do with existence as such. The existence both of A and of B has been
already pre-supposed in the two concepts, and when these two existing
things are brought into apposition, no third existence is thereby
supposed to have been created. The copula therefore really stands, not
as a symbol of _existence_, but as the symbol of _relation_, and might
just as well be replaced by any other sign (such as =), or, indeed,
be dispensed with altogether. “As we use the verb _is_, so the Latins
use their verb _est_ and the Greeks their [Greek: esti] through all
its declensions. Whether all other nations of the world have in their
several languages a word that answereth to it, or not, I cannot tell;
but I am sure they have no need of it. For the placing of two names in
order [_i.e._ in _apposition_] may serve to signify their consequence,
if it were the custom, as well as the words _is_, _to be_, and the
like. And if it were so, that there were a language without any verb
answering to _est_, or _is_, or _be_, yet the men that used it would be
not a jot the less capable of inferring, concluding, and of all kind
of reasoning than were the Greeks and Latins.” This shrewd analysis by
Hobbes is justly said by Mill to be “the only analysis of a proposition
which is rigorously true of all propositions without exception;” and
Professor Max Müller says of it, “Hobbes, though utterly ignorant of
the historical antecedents of language, agrees with us in the most
remarkable manner.”[102]

Thus, then, upon the whole, and without further treatment, it may be
concluded that whether we look to its simplest manifestations or to
its most complex, we must alike conclude that it is the faculty of
conception, not that of judgment—the faculty of denomination, not that
of predication—which we have to regard as “the simplest element of
thought.” Of course, if it were said that these two faculties are one
in kind—that in order to conceive we must judge, and in order to name
we must predicate—I should have no objection to offer. All I am at
present engaged upon is to make it clear that the distinction between
man and brute in respect of the Logos must be drawn at the place where
this distinction first obtains; and this place is where judgment is
concerned with conception, or with the bestowing of names in the sense
previously explained as _denominative_. The subsequent working up of
names into propositions is merely a further exhibition of the self-same
faculty. It is as true of judgment when displayed in denomination
as it is of judgment when displayed in predication, that “it is not
itself a modified imagination, because the imaginations which may
give rise to it persist unmodified in the mind side by side with
it.” For, as we have seen, the act of denominating (as distinguished
from denotating) is in and of itself an act of predicating. When a
naturalist bestows a name upon a new species of plant or animal, he
has _judged_ a resemblance and _predicates_ a fact—_i.e._ that the
hitherto unnamed form belongs to certain _genus_ or _kind_. And so it
is with all other names when conceptually bestowed, because everywhere
such names are expressions of conceptual _classification_—the bringing
together of like things, or the separation of unlike. In short, all
names which present any conceptual meaning are in themselves condensed
propositions, or “material predications;” and only as such can they
afterwards become _terms_, _i.e._ constitute the essential elements of
any more extended proposition, or “formal predication.” Therefore it
is the faculty of naming wherein is first displayed—and, according to
the doctrine of Nominalism, _whereby is first attained_—that great and
distinctive characteristic of the human mind which Mr. Mivart and those
who think with him have in view; and, unless we espouse the doctrine of
Realism—which neither these nor any other psychologists with whom I
have to do are likely nowadays to countenance,—it is plain that “the
simplest element of thought” is a concept.

       *       *       *       *       *

If I do not apologize for having occupied so much space over so obvious
a point, it is only because I believe that any one who reads these
pages will sympathize with my desire to avoid ambiguity, and thus to
reduce the question before us to its naked reality. So far, it will
be observed, this question has not been touched. I am not disputing
that an immense and an extraordinary distinction obtains, and I do not
anticipate that either Mr. Mivart or any one else will take exception
to this preliminary clearing of the ground, which has been necessitated
only on account of my opponents having been careless enough to
represent the Proposition as the simplest exhibition of the Logos. But
now the time has arrived when we must tackle the distinction in serious
earnest.

Wherein does this distinction truly consist? It consists, as I believe
all my opponents will allow, in the power which the human being
displays of _objectifying ideas_, or of setting one state of mind
before another state, and contemplating the relation between them.
The power to “think is”—or, as I should prefer to state it, the
power to think at all—_is the power which is given by introspective
reflection in the light of self-consciousness_. It is because the
human mind is able, so to speak, to stand outside of itself, and thus
to constitute its own ideas the subject-matter of its own thought,
that it is capable of judgment in the technical sense above explained,
whether in the act of conception or in that of predication. For thus
it is that these ideas are enabled “to exist _beside_ the judgment,
not _in_ it;” thus it is that they may themselves become objects of
thought. We have no evidence to show that any animal is capable of
thus objectifying its own ideas; and, therefore, we have no evidence
that any animal is capable of judgment. Indeed I will go further, and
affirm that we have the best evidence which is derivable from what are
necessarily ejective sources, to prove that no animal _can possibly_
attain to these excellencies of subjective life. This evidence will
gradually unfold itself as we proceed, so at present it is enough to
say, in general terms, that it consists in a most cogent proof of the
absence in brutes of the needful _conditions_ to the occurrence of
these excellencies as they obtain in themselves. From which it follows
that the great distinction between the brute and the man really lies
behind the faculties both of conception and predication: it resides
in the conditions to the occurrence of either. What these conditions
are I will consider later on. Meanwhile, and in order that we may be
perfectly clear about the all-important distinction which is before us,
I will re-state it in other terms.

What is the difference between a recept and a concept? I cannot
answer this question more clearly or concisely than in the words of
the writer in the _Dublin Review_ before quoted. “The difference
is all the difference between seeing two things united, and seeing
them _as united_.” The difference is all the difference between
perceiving relations, and perceiving the relations _as related_, or
between cognizing a truth, and recognizing that truth _as true_. The
diving bird, which avoids a rock and fearlessly plunges into the sea,
unquestionably displays a receptual knowledge of certain “things,”
“relations,” and “truths;” but it does not know any of them _as such_:
although it knows them, it does not _know that it knows them_: however
well it knows them, it does not _think_ them, or regard the things,
the relations, and the truths which it perceives as _themselves the
objects of perception_. Now, over and above this merely receptual
knowledge, man displays conceptual, which means that he _is_ able to do
all these things that the bird cannot do: in other words, he is able
to set before his mind all the recepts which he has in common with the
bird, to think about them _as_ recepts, and by the mere fact, or in the
very act of so doing, to convert them into concepts. Concepts, then,
differ from recepts in that they are recepts which have themselves
become objects of knowledge, and the condition to their taking on
this important character is the presence of self-consciousness in the
percipient mind.[103]

I have twice stated the distinction as clearly as I am able; but, in
order to do it the fullest justice, I will now render it a third time
in the words of Mr. Mivart—some of whose terms I have borrowed in
the above paragraph, and therefore need not now repeat. He begins by
conveying the distinction as it was stated by Buffon, thus:—

“Far from denying feelings to animals, I concede to them everything
except thought and reflection.... They have sensations, but no
faculty of comparing them with one another, that is to say they have
not the power which produces ideas”—_i.e._ products of reflection.
Then, after alluding to Buffon’s views on the distinction between
“automatic memory” and “intellectual memory” (_i.e._ the distinction
which I have recognized in the Diagram attached to my previous work by
calling the former “memory” and the latter “recollection”), Mr. Mivart
adds:—“The distinction is one quite easy to perceive. That we have
automatic memory, such as animals have, is obvious: but the presence
of intellectual memory may be made evident by searching our minds (so
to speak) for something which we have fully remembered before, and
thus intellectually remember to have known, though we cannot now bring
it before the imagination. And as with memory, so with other of our
mental powers, we may, I think, distinguish between a higher and a
lower faculty of each; between our higher, self-conscious, reflective
mental acts—the acts of our intellectual faculty—and those of our
merely sensitive power. This distinction I believe to be one of the
most fundamental of all the distinctions of biology, and to be one
the apprehension of which is a necessary preliminary to a successful
investigation of animal psychology.”[104]

Were it necessary, I could quote from his work, entitled _Lessons from
Nature_, sundry further passages expressing the same distinction in
other words; but I have already been careful, even to redundancy, in
presenting this distinction, not only because it is the distinction
on which Mr. Mivart rests his whole argument for the separation of
man from the rest of the animal kingdom as a being unique in kind;
but still more because it is, as he is careful to point out, the one
real distinction which has hitherto always been drawn by philosophers
since the time of Aristotle. And, as I have already observed, it is
a distinction which I myself fully recognize, and believe to be the
most important of all distinctions in psychology. The only point of
difference, therefore, between my opinions and those—I will not say of
Mr. Mivart, but—of any other or possible opponent who understands the
psychology of this subject, is on the question whether, in view of the
light which has now been shed on psychology by the theory of evolution,
this important distinction is to be regarded as one of degree or as
one of kind. I shall now proceed to unfold the reasons which lead me
to differ on this point from Mr. Mivart, and so from all the still
extensive school of which he is, in my opinion, much the ablest
spokesman.

       *       *       *       *       *

We have seen that the distinction in question consists in the presence
or absence of the faculty now fully explained, of reflective thought,
and that of this faculty the simplest manifestation is, as alleged
by my opponents, that which is afforded by “judgment.” But we have
also seen that this faculty of judgment does not first appear in
predication, unless we extend the term so as to embrace all acts
of denomination. In other words, we have seen that judgment first
arises with conception—and necessarily so, seeing that neither of
these things can occur without the other, but both arise as direct
exhibitions of that faculty of self-conscious or reflective thought of
which they are everywhere the immediate expression. I will, therefore,
begin with a careful analysis of conceptual judgment.

We must first recur to the distinctions set forth at the close of the
last chapter, where it was shown that, without any prejudice to the
question touching the distinction between man and brute, there are five
different stages of intentional sign-making to be recognized—namely,
the indicative, the denotative, the connotative, the denominative, and
the predicative. From what has now been said regarding the essentially
predicative nature of all conceptual names, we may disregard the last
of these distinctions, and consider the denominative phase of language
as psychologically identical with the predicative. Similarly, we may
now neglect the indicative phase, as one which bears no relation to the
matters at present before us. Thus we have to fasten attention only
upon the differences between the denotative, the connotative, and the
denominative phases of language. This has already been done in general
terms; but must now be done in more detail. And for the sake of being
clear, even at the risk of being tedious, I will begin by repeating the
important distinctions already explained.

When a parrot calls a dog _Bow-wow_ (as a parrot, like a child, may
easily be taught to do), the parrot may be said, in one sense of
the word, to be _naming_ the dog; but it is not _predicating_ any
characters as belonging to a dog, or performing any act of _judgment_
with regard to a dog. Although the bird may never (or but rarely) utter
the name save when it sees a dog, this fact is attributable to the
laws of association acting only in the receptual sphere: it furnishes
no shadow of a reason for supposing that the bird _thinks_ about a
dog _as_ a dog, or sets the concept Dog before its mind as a separate
object of thought. Therefore, all my opponents must allow that in one
sense of the word there may be names without concepts: whether as
gestures or as words (vocal gestures), there may be signs of things
without these signs presenting any vestige of predicative value. Names
of this kind I have called _denotative_: they are marks affixed to
objects, qualities, actions, &c., by receptual association alone.

Next, when a denotative name has been formed and applied as the mark of
one thing, its use may be extended to denote also another thing, which
is seen to belong to the same class or kind. When denotative names
are thus extended, they become what I have called _connotative_. The
degree to which such classificatory extension of a denotative name may
take place depends, of course, on the degree in which the mind is able
to take cognizance of resemblances or analogies. Now, these degrees
are as various as are the degrees of intelligence itself. Long before
the differential engine of Conception has come to the assistance of
Mind, both animals and human beings (as previously shown) are able to
go a long way in the distinguishing of resemblances, or analogies, by
means of receptual ideation alone. When such receptual discrimination
is expressed by the corresponding extension of denotative names, the
degree of connotation which such names may thus acquire depends upon
the degree of this receptual discrimination. Even my parrot was able to
extend its denotative name for a particular dog to any other dog which
it happened to see—thus precisely resembling my child, who extended
its first denotative word _Star_ to a candle. Connotation, then, begins
in the purely receptual sphere of ideation; and although in man it is
afterwards carried up into the conceptual sphere, it is obviously most
imperative for the purposes of this analysis to draw a distinction
between connotation as receptual and as conceptual.

This distinction I have drawn by assigning the word _denomination_
to all connotation which is of a truly conceptual nature—or to the
bestowing of names _consciously recognized as such_. And I have just
shown that when connotation is thus denominative or conceptual, it is
psychologically the same as predication. Therefore it is only in this
denominative sense of the word, or in cases where conceptual ideation
is concerned, that an act of naming involves an act of judgment,
strictly so called.

Such being the psychological standing of the matter, it is evident
that the whole question before us is narrowed down to a clearing up
of the relations that obtain between connotation as receptual and
conceptual—or between connotation, that is, and connotation that is
not, denominative. To do this I will begin by quoting an instance of
un-denominative or receptual connotation in the case of a young child.

“There is this peculiar to man—the sound which has been associated in
his case with the perception of some particular individual is called up
again, not only at the sight of absolutely similar individuals, but
also by the presence of individuals strikingly different, though in
some respects comprised in the same class. In other words, analogies
which do not strike animals strike men. The child says _Bow-wow_,
first to the house-dog, then, after a little, he says _Bow-wow_ to
the terriers, mastiffs, and Newfoundlands he sees in the street. A
little later he does what an animal never does, he says _Bow-wow_ to a
paste-board dog which barks when squeezed, then to a paste-board dog
which does not bark, but runs on wheels, then to the silent motionless
bronze dog which ornaments the drawing-room, then to his little cousin
who runs about the room on all fours, then, at last, to a picture
representing a dog.”[105]

Now, in this small but typical history we have a clear exhibition, in a
simple form, of the development of a connotative name within the purely
receptual sphere. At first the word _Bow-wow_ was merely a denotative
name—or a mark affixed to a particular object of perception. But
when the child’s mind took cognizance of the resemblances between
the house-dog, terriers, mastiffs, and Newfoundlands, it expressed
the fact by extending the name _Bow-wow_ to all these dogs. The
name, from being particular, thus became generic, or indicative of
_resemblances_; and, therefore, from being merely denotative, became
truly connotative: it now served to express _common attributes_. Next,
this receptual connotation of the name was still further widened, so
as to include—or to signify—the resemblances between dogs and their
images, pictures, &c. Now, in these several and successive acts of
connotative naming, the child was obviously advancing to higher and
higher levels of receptual classification; but, no less obviously, it
would be absurd to suppose that the child was thus raising the name
_Bow-wow_ to any _conceptual_ value. All that any child in such a case
is doing is to extend its receptual appreciation of resemblance through
widening circles of generic grouping, and correspondingly to extend
the receptual connotation of a denotative name. In order to do this
(within the limits that we are now considering), there is no need for
any introspective regarding of the name as a name: there is no need to
contemplate the widening connotation of the name: there is no need to
_judge_, to _define_, to _denominate_. Such classification as is here
effected can be effected within the region of receptual consciousness
alone (as we well know from the analogous case of the parrot, and the
“practical inferences” of the lower animals generally); therefore, if
the denotative name originally assigned to a particular dog admitted
of being so assigned as merely the mark of that particular recept,
there is no reason to suppose that its subsequent extension to the
more generic recepts afterwards experienced involves any demand upon
the conceptual faculty, or implies that the child could only extend
this name from a house-dog to a terrier by first performing an act of
introspective thought—which, indeed, as we shall see later on, it is
demonstrably impossible that a child of this age can be able to do.

Nevertheless, it is evident that already the child has done more
than the parrot. For a parrot will never extend its denotative name
of a particular dog to the picture, or even to the image of a dog.
The utmost that a parrot will do is to extend the denotative name
from one particular dog to another particular dog, which, however,
may differ considerably from the former as to size, colour, and
general appearance. Still, I presume, no one will maintain that thus
far there is the faintest evidence of a difference of kind between
the connotative faculty of the bird and that of the child. All that
these facts can be held to show is that—in the words already quoted
from M. Taine while narrating these facts—“analogies which do not
strike animals strike men.” Or, in my own phraseology, the receptual
faculties of a parrot do not go further than the receptual faculties
of a very young child: consequently, the denotative name in the
case of the parrot only undergoes the first step in the process of
receptual extension—namely, from a house-dog to a terrier, a setter,
a mastiff, a Newfoundland, &c. But in the case of the child, _after
having reached this stage_, the process of extension continues, so as
to embrace images, and eventually pictures of dogs. This difference,
however, only shows an advance in the merely receptual faculties:
does not suggest that in order to carry the extension of the name
through these second and third stages, demand has yet been made on the
distinctively human powers of conceptual thought—any more than such
powers were required to carry it through the first stage in the case of
the parrot.

Hence we see again that the distinction already drawn between
denotative and connotative names is not co-extensive with the
distinction between ideas as receptual and conceptual. Or, in other
words, names may be in some measure connotative even in the absence
of self-consciousness. For if we say that a child is connoting
resemblances when it extends the name _Bow-wow_ from a particular
dog to dogs in general, clearly we must say the same thing of a
parrot when we find that thus far it goes with the child. Therefore
it is that I have distinguished between connotation as receptual and
conceptual—_i.e._ by calling the latter _denomination_. Receptual
connotation represents a higher level of ideational faculty than
mere denotation; but a lower level than conceptual connotation, or
denomination. Moreover, receptual connotation admits of many degrees
before we can discern the smallest reason for supposing that it is
even in the lowest degree conceptual. Connotation of all degrees
depending on perceptions of resemblances or analogies, the higher the
receptual life, and therefore the greater the aptitude of receptual
classification, the more will such classification become reflected in
connotative expression. Therefore it is that the child will not only
surpass the parrot in its receptual connotation from dogs to pictures
of dogs; but, as we shall afterwards see, will go much further even
than this before it gives any signs at all of conceptual connotation,
or true denomination. Thus we see that between the most rudimentary
receptual connotation which a very young child shares with a parrot,
and the fully conceptual connotation which it subsequently attains,
there is a large intervening province due to the acquisition of a
higher receptual life. Or, to put the same thing in other words, there
is a large tract of ideation lying between the highest receptual
life of a brute and the lowest conceptual life of a man: this tract
is occupied by the growing child from the time at which its ideation
surpasses that of the brute, until it begins to attain the faculty
of self-conscious reflection. This intervening tract of ideation,
therefore, may be termed “higher receptual,” in contradistinction to
the lower receptual ideation which a younger child shares with the
lower animals.

At this point I must ask the reader carefully to fasten in his mind
these various distinctions. Nor will it be difficult to do so after a
small amount of attention. It will be remembered that in Chapter IV. I
instituted a distinction between concepts as higher and lower, which
was methodically similar to that which I have now to institute between
recepts. A “lower concept” was defined to be nothing more than a “named
recept,”[106] while a “higher concept” was understood to be one that is
“compounded of other concepts”—_i.e._ the named result of a grouping
of concepts, as when we speak of the “mechanical equivalent of heat.”
So that altogether we have four stages of ideation to recognize, each
of which occupies an immensely large territory of mind. These four
stages I will present in serial order.

(1) _Lower Recepts_, comprising the mental life of all the lower
animals, and so including such powers of receptual connotation as a
child when first emerging from infancy shares with a parrot.

(2) _Higher Recepts_, comprising all the extensive tract of ideation
that belongs to a child between the time when its powers of receptual
connotation first surpass those of a parrot, up to the age at which
connotation as merely denotative begins to become also denominative.

(3) _Lower Concepts_, comprising the province of conceptual ideation
where this first emerges from the higher receptual, up to the point
where denominative connotation has to do, not merely with the naming of
recepts, but also with that of associated concepts.

(4) _Higher Concepts_, comprising all the further excellencies of human
thought.

Higher Recepts, then, are what may be conveniently termed
Pre-concepts:[107] they occupy the interval between the receptual
life of brute and the earliest dawn of the conceptual life of man. A
pre-concept, therefore, is that kind of higher recept which is not to
be met with in any brute; but which occurs in the human being after
surpassing the brute and before attaining self-consciousness. Be it
observed that in thus coining the words higher recepts or pre-concepts,
I am not in any way prejudicing the case of my opponents; I am merely
marking off a certain territory of ideation which has now for the
first time been indicated. Of course my object eventually is to show
that in the history of a growing child, just as sensations give rise
to percepts, and percepts to recepts (as they do among animals), so do
recepts give rise to pre-concepts, pre-concepts to concepts, concepts
to propositions, and propositions to syllogisms. But in now supplying
this intermediate link of pre-concepts I am not in any way pre-judging
the issue: I am merely marking out the ground for discussion. No one of
my opponents can dispute my facts, which are too obvious to admit of
question. Therefore, if they object to my classification of them so far
as the novel division of pre-concepts is concerned, it must be because
they think that by instituting this division I am surreptitiously
bringing the mind of a child nearer to that of an animal than they deem
altogether safe. What, then, I ask, would they have me do? If I fail
to institute this division, I should have to prejudice the question
indeed. Either there is some distinction between the naming powers of
a parrot and those of a young child, or else there is not. If there is
no distinction, so much the better for the purposes of my argument.
But I allow that there is a distinction, and I draw it at the first
place where it can possibly be said that the intelligence of a child
differs in any way at all from that of a parrot—_i.e._ where the
naming powers of a child demonstrably excel those of a parrot, or any
other brute. If this place happens to be before the rise of conceptual
powers, I am not responsible for the fact; nor in stating it am I at
all disparaging the position of any opponent who takes his stand upon
these powers as distinctive of man. If his position were worth anything
before, it cannot be affected by my drawing attention to the fact that,
while a parrot will extend its denotative name of a dog from a terrier
to a setter, it will not follow a child any further in the process of
receptual connotation.

Or, to put it in another way, when the child says _Bow-wow_ to a
setter, after having learnt this name for a terrier, it is either
judging a resemblance and predicating a fact, or else it is doing
neither of these things. If my opponents elect to say that the child is
doing both these things, there is an end of the only issue between us;
for in that case a parrot also is able both to judge and to predicate.
On the other hand, if my opponents adopt the wiser course, and accept
my distinction between names as receptual and conceptual, they must
also follow me in recognizing the border-land of pre-concepts as lying
between the recepts of a bird and the concepts of a man—_i.e._ the
territory which is first occupied by the higher receptual life of a
child before this passes into the conceptual life of a man,—for that
such a border-land does exist I will prove still more incontestably
later on. There is, then, as a matter of observable fact, a territory
of ideation which separates the highest recepts of a brute from the
lowest concepts of a human being; and all that my term pre-conception
is designed to do is to name this intervening territory.

Now, if this is the case with regard to naming, clearly it must also be
the case with regard to judging: if there is a stage of pre-conception,
there must also be a stage of pre-judgment. For we have seen that it is
of the essence of a judgment that it should be concerned with concepts:
if the mind be concerned merely with recepts, no act of true judgment
can be said to have been performed. When a child says _Bow-wow_ to
the picture of a dog, no one can maintain that he is actually judging
the resemblance of the picture to a dog, unless it be supposed that
for this act of receptual classification distinctively human powers
of conceptual thought are required. But, as just shown, no opponent
of mine can afford to adopt this supposition, because behind the case
of the child there stands that of the parrot. True, the parrot does
not proceed in its receptual classification further than to extend its
name for a particular dog to other living dogs; but if any one were
foolish enough to stake his whole argument on so slender a distinction
as this—to maintain that at the place where the connotation of a child
first surpasses that of a parrot we have evidence of a psychological
distinction of kind, _on the sole ground that the child has begun to
surpass the parrot_—it would be enough for me to remark that not
_every_ parrot will thus extend its denotative sign from one dog
to another of greatly unlike appearance. Different birds display
different degrees of intelligence in this respect. Most of them will
say _Bow-wow_, will bark, or utter any other denotative sign which they
may have learnt or invented, when they see dogs more or less resembling
the one to which the denotative sign was originally applied; but it
is not every parrot which will thus extend the sign from a terrier to
a mastiff or a Newfoundland. Therefore, if any one were to maintain
that the difference between the intelligence which can discern, and
one which cannot discern, the likeness of a dog in the image or the
picture of a dog, is a difference of kind, consistency should lead
him to draw a similar distinction between the intelligence which can
discern, and one which cannot discern, the likeness of a terrier to a
mastiff. But, if so, the intelligence of one parrot would be different
in kind from that of another parrot; and the child’s intelligence
at one age would differ in kind from the intelligence of that same
child when a week or two older—both of which statements would be
manifestly absurd. The truth can only be that up to the point where
the intelligence of the child surpasses that of the bird they are both
in the receptual stage of sign-making; and that the only reason why
the child does surpass the bird is not, in the first instance, because
the child there suddenly attains the power of conceptual ideation, but
because it gradually attains a higher level of receptual ideation. This
admits of direct proof from the fact that animals more intelligent
than parrots are unquestionably able to recognize sculptured and even
pictorial representations: hence there can be no doubt that if talking
birds had attained a similar level of intelligence—or if the other and
more intelligent animals had been able, like the talking birds, to use
denotative signs,—the child would not have parted company with the
brute at quite so early a stage of receptual nomenclature.[108]

What, then, are we to say about the faculty of judgment in relation to
these three stages of ideation—namely, the receptual, pre-conceptual,
and conceptual? We can only institute the parallel and consequent
distinction between judgment as receptual, pre-conceptual, and
conceptual.[109] As now so often stated, the distinguishing features of
a judgment as fully displayed in any act of formal predication, are the
bringing together in self-conscious thought of two concepts, and the
distinguishing of some relation between them as such. Therefore we do
not say that a brute judges when, without any self-conscious thought,
it brings together certain reminiscences of its past experience in the
form of recepts, and translates for us the results of its ideation
by the performance of what Mr. Mivart calls “practical inferences.”
Therefore, also, if a brute which is able to name each of two recepts
separately (as is done by a talking bird), were to name the two recepts
simultaneously when thus combined in an act of “practical inference,”
although there would then be the outward semblance of a proposition,
we should not be strictly right in calling it a proposition. It would,
indeed, be the statement of a truth _perceived_; but not the statement
of a truth perceived _as true_.[110]

Now, if all this be admitted in the case of a brute—as it must be
by any one who takes his stand on the faculty of true or conceptual
judgment,—obviously it must also be admitted in the case of the
growing child. In other words, if it can be proved that a child is able
to state a truth before it can state a truth as true, it is thereby
proved that in the psychological history of every human being there
is first the incompleted kind of judgment required for dealing with
receptual knowledge, and so for stating truths perceived, and next
the completed judgment, which deals with conceptual knowledge, and so
is enabled to state truths perceived as true. Of course the condition
to the raising of this lower kind of judgment (if for convenience
we agree so to term it) into the higher, is given by the advent of
self-consciousness; and therefore the place where _statement_ of truth
passes into _predication_ of truth must be determined by the place
at which this kind of consciousness first supervenes. Where it does
first supervene we shall presently have to consider. Meanwhile I am
but endeavouring to make clear the fact that, unless my opponents
abandon their position altogether, they must allow that there is
_some_ difference to be recognized between the connotative powers of a
parrot and the connotative powers of a man. But if they do allow this,
they must further allow that between the place where the connotative
powers of a child first surpass those of a parrot, and the place where
those powers first become truly conceptual, there is a large tract of
ideation which it is impossible to ignore. In order, therefore, not
to prejudice the question before us, I have thus far confined myself
to a mere designation of these great and obvious distinctions. But
seeing that even this preliminary step has necessitated a great deal of
explanation, I feel it may conduce to clearness if I end the present
chapter with a tabular statement of the sundry distinctions in question.

By _receptual judgments_ I will understand the same order of ideation
as Mr. Mivart expresses by his term “practical inferences of brutes,”
instances of which have already been given in Chapter III.

By _pre-conceptual judgments_ I will understand those acts of virtual
or rudimentary judgment which are performed by children subsequent to
the “practical inferences” which they share with brutes, but prior to
the advent of self-conscious reflection. These pre-conceptual judgments
may be expressed either by gestures, connotative classifications, or
by both combined. Some instances of them have already been given in
the present chapter: further and better instances will be given in the
chapters which are to follow.

By _conceptual judgments_ I will understand full and complete judgments
in the ordinary acceptation of this term.

Receptual judgment, then, has to do with recepts; pre-conceptual
judgment with pre-concepts; and true judgments with true concepts. Or,
conversely stated, receptual knowledge leads to receptual judgment
(_e.g._ when a sea-bird dives into water but alights upon land):
pre-conceptual knowledge leads to pre-conceptual judgment in the
statement of such knowledge (_e.g._ when a child, by extending the name
of a dog to the picture of a dog, virtually affirms, though it does not
conceive, the resemblance which it perceives): and, lastly, conceptual
knowledge leads to conceptual or veritable judgment, in the statement
of such knowledge known as knowledge (_e.g._ when, in virtue of his
powers of reflective thought, a man not only states a truth, but states
that truth as true).

       *       *       *       *       *

Thus far I doubt whether my opponents will find it easy to meet me.
They may, of course, cavil at some or all of the above distinctions;
but, if so, it is for them to show cause for complaint. They have
raised objections to the theory of evolution on purely psychological
grounds. I meet their objections upon these their own grounds, and
therefore the only way in which they can answer me is by showing
that there is something wrong in my psychological analysis. This I
fearlessly invite them to do. For all the distinctions which I have
made I have made out of consideration to the exigencies of their
argument. Although these distinctions may appear somewhat bewilderingly
numerous, I do not anticipate that any competent psychologist will
complain of them on account of their having been over-finely drawn.
For each of them marks off an important territory of ideation, and
all the territories so marked off must be separately noted, if the
alleged distinction of kind between one and another is to be seriously
investigated. In his essays upon the theory of evolution, Mr. Mivart
not unfrequently complains of the disregard of psychological analysis
which is betokened by any expression of opinion to the effect, that
as between one great territory of ideation and another there is only
a difference of degree. But surely this complaint comes with an ill
grace from a writer who bases an opposite opinion upon a precisely
similar neglect—or upon a bare statement of the greatest and most
obvious of all the distinctions in psychology, without so much as any
attempt to analyze it. Therefore, if my own attempt to do this has
erred on the side of overelaboration, it has done so only on account
of my desire to do full justice to the opposite side. In the result,
I claim to have shown that if it is possible to suggest a difference
of kind between any of the levels of ideation which have now been
defined, this can only be done at the last of them—or where the advent
of self-consciousness enables a mind, not only to _know_, but to _know
that it knows_; not only to _receive_ knowledge, but also to _conceive_
it; not only to _connotate_, but also to _denominate_; not only to
_state a truth_, but also to state that truth _as true_. The question,
therefore, which now lies before us is that as to the nature of this
self-consciousness—or, more accurately, whether the great and peculiar
distinction which this attribute confers upon the human intellect is
to be regarded as a distinction of degree only, or as a distinction of
kind. To answer this question we must first investigate the rise of
self-consciousness in the only place where its rise can be observed,
namely, in the psychogenesis of a child.[111]




CHAPTER X.

SELF-CONSCIOUSNESS.


My contention in this chapter will be that, given the protoplasm
of the sign-making faculty so far organized as to have reached the
denotative stage; and given also the protoplasm of judgment so far
organized as to have reached the stage of stating a truth, without the
mind being yet sufficiently developed to be conscious of itself as an
object of thought, and therefore not yet able to state to itself a
truth as true; by a confluence of these two protoplasmic elements an
act of fertilization is performed, such that the subsequent processes
of mental organization proceed apace, and soon reach the stage of
differentiation between subject and object.

And here, to avoid misapprehension, I may as well make it clear at the
outset that in all which is to follow I am in no way concerned with
the philosophy of this change, but only with its history. On the side
of its philosophy no one can have a deeper respect for the problem
of self-consciousness than I have; for no one can be more profoundly
convinced than I am that the problem on this side does not admit of
solution. In other words, so far as this aspect of the matter is
concerned, I am in complete agreement with the most advanced idealist;
and hold that in the datum of self-consciousness we each of us possess,
not merely our only ultimate knowledge, or that which only is “real
in its own right,” but likewise the mode of existence which alone the
human mind is capable of conceiving as existence, and therefore the
_conditio sine quâ non_ to the possibility of an external world.
With this aspect of the question, however, I am in no way concerned.
Just as the functions of an embryologist are confined to tracing the
mere history of developmental changes of living structure, and just
as he is thus as far as ever from throwing any light upon the deeper
questions of the how and the why of life; so in seeking to indicate
the steps whereby self-consciousness has arisen from the lower stages
of mental structure, I am as far as any one can be from throwing light
upon the intrinsic nature of that the probable genesis of which I am
endeavouring to trace. It is no less true to-day than it was in the
time of Soloman, that “as thou knowest not how the bones do grow in the
womb of her that is with child, thou knowest not what is the way of the
spirit.”

       *       *       *       *       *

If we are agreed that it is only in man that self-consciousness is to
be found at all, it follows that only to man can we look for any facts
bearing upon the question of its development. And inasmuch as it is
only during the first years of infancy that a normal human being is
destitute of self-consciousness, the statement just made implies that
only in infant psychology need we seek for the facts of which we are
in search. Further, as I maintain that self-consciousness arises out
of an admixture of the protoplasm of judgment with the protoplasm of
sign-making (according to the signification of these terms as already
explained), I have now to make good this opinion upon the basis of
facts drawn from the study of infant psychology.

Nevertheless, before I proceed to the heart of the subject, I think it
will be convenient to consider those faculties of mind which, occurring
both in the infant and in the animal, in the former case precede the
advent of self-consciousness, and, according to my view, prepare the
way for it.

It will, I suppose, on all hands be admitted that self-consciousness
consists in paying the same kind of attention to internal or psychical
processes as is habitually paid to external or physical processes—a
bringing to bear upon subjective phenomena the same powers of
perception as are brought to bear upon the objective. The degrees in
which such attention may be yielded are, of course, as various in the
one case as in the other; but this does not affect my psychological
definition of self-consciousness.

Again, I suppose it will be further admitted that in the mind of
animals and in the mind of infants there is a world of images
standing as signs of outward objects; and that the only reason why
these images are not attended to unless called up by the sensuous
associations supplied by their corresponding objects, is because the
mind is not yet able to leave the ground of such association, so as
to move through the higher and more tenuous medium of introspective
thought.[112] Nevertheless, this image world assuredly displays
an internal activity which is not wholly dependent on sensuous
associations supplied from without. That is to say, one image suggests
another, this another, and so on—although, as I have just conceded,
this cannot be due to successive acts of inward attention, or of the
self-conscious contemplation of images known as such. Nevertheless,
that an internal—though unintentional—play of ideation takes place in
the minds of brutes, without the necessity of immediate associations
supplied from present objects of sense, admits of being amply proved
from the phenomena of dreaming, hallucination, home-sickness, pining
for absent friends, &c., which, as I have fully shown in my previous
work, can only be explained by recognizing such a play of inward
ideation.[113] Now, I hold it of importance to note that such an
internal play of ideation is thus possible even in the absence of
self-consciousness, because many writers have assumed, without any
justification, that unless ideas are intentionally contemplated
as such, they must be wholly dependent for their occurrence upon
associations supplied by present objects of sense. Of course I do not
doubt that an agent who is capable of intentionally making one idea
stand as the object of another, is likewise capable of going very
much further than a brute in the way of causing one idea to start from
another irrespective of immediate stimulation from without. My point
here is merely to remark that the ideation of brutes is not wholly
dependent on such stimulation; but is capable, in a certain humble
degree, of forming independent chains of its own.

The next thing which I desire to be remembered in connection with
the ideation of brutes is, that it is not restricted to the mere
reproduction in memory of particular objects of sensuous impressions;
but, as we have so fully seen in Chapter III., admits of undergoing
that amount of mental elaboration which belongs to what I have termed
recepts.

Furthermore, the foundations of self-consciousness are largely laid in
the fact that an organism is one connected whole; all the parts are
mutually related in the unity of individual sensibility. Every stimulus
supplied from without, every movement originating from within, carries
with it the character of belonging to that which feels and moves. Hence
a brute, like a young child, has learnt to distinguish its own members,
and likewise its whole body, from all other objects; it knows how to
avoid sources of pain, how to seek those of pleasure; and it also knows
that particular movements follow from particular volitions, while in
connection with such movements it constantly experiences the same
muscular sensations. Of course such knowledge and such experience all
belong to the receptual order; but this does not hinder that they play
a most important part in laying the foundations of a consciousness of
individuality.[114]

Lastly, and I believe of still more importance in the present
connection than any of the above-named antecedents, a large
proportional number of the recepts of a brute have reference, not to
objects of sense, or even to muscular sensations, but to the _mental
states of other animals_. That is to say, the logic of recepts, even
in brutes, is sufficient to enable the mind to establish true analogies
between its own states (although these are not yet the objects of
separate attention, or of what may be termed subjective knowledge),
and the corresponding states of other minds. I need not dwell upon
this point, because I take it to be a matter of general observation
that animals habitually and accurately interpret the mental states
of other animals, while they also well know that other animals are
able similarly to interpret _theirs_—as is best proved by their
practising the arts of cunning, concealment, hypocrisy, &c.[115] From
which considerations we reach the general conclusion, that intelligent
animals recognize a world of ejects as well as a world of objects:
mental existence is known to them ejectively, though, as may be
allowed, never _thought upon_ subjectively.[116]

It is of importance further to observe that at this stage of mental
evolution the individual—whether an animal or an infant—so far
realizes its own individuality as to be informed by the logic of
recepts that it is _one of a kind_. I do not mean that at this
stage the individual realizes its own or any other individuality
as such; but merely that it recognizes the fact of its being one
among a number of similar though distinct forms of life. Alike in
conflict, rivalry, sense of liability to punishment or vengeance,
&c., the truth is continually being borne in upon the mind of an
animal that it is a separate individuality; and this though it be
conceded that the animal is never able, even in the most shadowy
manner, to think about itself as such. In this way there arises a sort
of “outward self-consciousness,” which differs from true or inward
self-consciousness only in the absence of any attention being directed
upon the inward mental states as such. This outward self-consciousness
is known to us all, even in adult life—it being but comparatively
seldom that we pause in our daily activities to contemplate the mental
processes of which these activities are the expression.

Now, if these things are so, we encounter the necessity of drawing
the same distinction in our analysis of self-consciousness, as we
have had to draw in our previous analyses of all the other faculties
of mind: there is a self-consciousness that is receptual, and a
self-consciousness that is conceptual. No doubt it is to the latter
kind of self-consciousness alone that the term is strictly applicable,
just as it is to conceptual naming or to conceptual predicating
alone that the word “judgment” is strictly applicable. Nevertheless,
here, as before, we must not ignore an important territory of mind
only because it has hitherto remained uncharted.[117] Receptual or
outward self-consciousness, then, is the practical recognition of
self as an active and a feeling agent; while conceptual or inward
self-consciousness is the introspective recognition of self as an
object of knowledge, and, therefore, as a subject. Hence, the one
form of self-consciousness differs from the other in that it is only
objective and never subjective.[118]

I take it, then, as established that true or conceptual
self-consciousness consists in paying the same kind of attention to
inward psychical processes as is habitually paid to outward physical
processes; that in the mind of animals and infants there is a world of
images standing as signs of outward objects, although we may concede
that for the most part they only admit of being revived by sensuous
association; that at this stage of mental evolution the logic of
recepts comprises an ejective as well as an objective world; and that
here we also have the recognition of individuality, so far as this is
dependent on what has been termed an outward self-consciousness, or the
consciousness of self as a feeling and an active agent, without the
consciousness of self as an object of thought, and, therefore, as a
_subject_.

Such being the mental conditions precedent to the rise of true
self-consciousness, we may next turn to the growing child for evidence
of subsequent stages in the gradual evolution of this faculty. All
observers are agreed that for a considerable time after a child is
able to use words as expressive of ideas, there is no vestige of true
self-consciousness. But, to begin our survey before this period, at a
year old even its own organism is not known to the child as part of the
self, or, more correctly, as anything specially related to feelings.
Professor Preyer observed that his boy, when more than a year old, bit
his own arm just as though it had been a foreign object; and thus may
be said to have shown even less consciousness of a limb as belonging
to “self,” than did Buffon’s parrot, which would first ask itself for
its own claw, and then comply with the request by placing the claw in
its own beak—in the same way as it would give the claw to any one else
who asked for it in the same words.

Later on, when the outward self-consciousness already explained has
begun to be developed, we find that the child, like the animal, has
learnt to associate its own organism with its own mental states, in
such wise that it recognizes its body as belonging in a peculiar
manner to the self, so far as the self is recognizable by the logic of
recepts. This is the stage that we meet with in animals. Next the child
begins to talk, and, as we might expect, this first translation of the
logic of recepts reveals the fact that as yet there is no _inward_
self-consciousness, but only outward: as yet the child has paid no
_attention_ to his own mental states, further than to feel that he
feels them; and in the result we find that the child speaks to himself
as an object, _i.e._ by his proper name or in the third person. That
is to say, “the child does not as yet set himself in opposition to all
outer objects, including all other persons, but regards himself as one
among many objects.”[119] The change of a child’s phraseology from
speaking of self as an object to speaking of self as a subject does not
take place—or but rarely so—till the third year. When it has taken
place we have definite evidence of true self-consciousness, though
still in a rudimentary stage. And it is doubtful whether this change
would take place even at so early an age as the third year, were it
not promoted by the “social environment.” For, as Mr. Sully observes,
“the relation of self and not self, including that between the I and
the You, is continually being pressed on the child’s attention by the
language of others.”[120] But, taking this great change during the
time of life when it is actually observed to be in progress, let us
endeavour to trace the phases of its development.

It will no doubt be on all hands freely conceded, that at least up to
the time when a child begins to speak it has no beginning of any true
or introspective consciousness of self; and it will further be conceded
that when this consciousness begins to dawn, the use of language
by a child may be taken as a fair exponent of all its subsequent
progress. Now we have already seen that, long before any words are
used indicative of even a dawning consciousness of self as self, the
child has already advanced so far in its use of language as to frame
implicit propositions. But lest it should be thought that my judgment
in this matter is biased by the exigencies of my argument, I may again
quote Mr. Sully as at once an impartial witness and a highly competent
authority on matters of purely psychological doctrine.

“When a child of eighteen months on seeing a dog exclaims ‘Bow-wow,’
or on taking his food exclaims ‘Ot’ (Hot), or on letting fall his toy
says ‘Dow’ (Down), he may be said to be implicitly framing a judgment:
‘That is a dog,’ ‘This milk is hot,’ ‘My plaything is down.’ The first
explicit judgments are concerned with individual objects. The child
notes something unexpected or surprising in an object, and expresses
the result of his observation in a judgment. Thus, for example, the boy
more than once referred to, whom we will call C., was first observed to
form a distinct judgment when nineteen months old, by saying ‘Dit ki’
(Sister is crying). These first judgments have to do mainly with the
child’s food, or other things of prime importance to him. Thus, among
the earliest attempts at combining words in propositions made by C.
already referred to, were the following: ‘Ka in milk,’ (Something nasty
in milk); ‘Milk dare now’ (There is still some more milk in the cup).
Towards the end of the second year quite a number of judgments is
given out having to do with the peculiarities of objects which surprise
or impress the mind, their altered position in space, &c. Among these
may be instanced the following: ‘Dat a big bow-wow’ (That is a large
dog); ‘Dit naughty’ (Sister is naughty); ‘Dit dow ga’ (sister is down
on the grass). As the observing powers grow, and the child’s interest
in things widens, the number of his judgments increases. And as his
powers of detaching relations and of uttering and combining words
develop, he ventures on more elaborate statements, _e.g._ ‘Mama naughty
say dat.’”[121]

Were it necessary, I could confirm all these statements from my own
notes on the development of children’s intelligence; but I prefer,
for the reason already given, to quote such facts from an impartial
witness. For I conceive that they are facts of the highest importance
in relation to our present subject, as I shall immediately proceed to
show.

We have now before us unquestionable evidence that in the growing
child there is a power, not only of forming, but of expressing a
pre-conceptual judgment, long before there is any evidence of the
child presenting the faintest rudiment of internal, conceptual, or
true self-consciousness. In other words, it must be admitted that long
before a human mind is sufficiently developed to perceive relations as
related, or to state a truth as true, it is able to perceive relations
and to state a truth: the logic of recepts is here concerned with
those higher receptual judgments which I have called pre-conceptual,
and is able to express such judgments in verbal signs without the
intervention of true (_i.e._ introspective) self-consciousness. It
will be remembered that I have coined these various terms in order to
acknowledge the possible objection that there can be no true judgments
without true self-consciousness. But I do not care what terms are
employed whereby to designate the different and successive phases
of development which I am now endeavouring to display. All that I
desire to make clear is that here we unquestionably have to do with a
_growth_, or with a continuous advance in degree as distinguished from
a difference of kind.

First, then, let it be observed that in these rudimentary judgments
we already have a considerable advance upon those which we have
considered as occurring in animals. For in a child between the second
and third years we have these rudimentary judgments, not only formed
by the logic of recepts, but expressed by a logic of pre-concepts
in a manner which is indistinguishable from predication, except by
the absence of self-consciousness. “Dit dow ga” is a proposition in
every respect, save in the absence of the copula; which, as I have
previously shown, is a matter of no psychological moment. The child
here perceives a certain fact, and states the perception in words, _in
order to communicate information of the fact to other minds_—just as
an animal, under similar circumstances, will use a gesture or a vocal
sign; but the child is no more able than the animal designedly to make
to its own mind the statement which it makes to another. Nevertheless,
as the child has now at its disposal a much more efficient system
of sign-making than has the animal, and moreover enjoys the double
advantage of inheriting a strong propensity to communicate perceptions
by signs, and of being surrounded by the medium of speech; we can
scarcely wonder that its practical judgments (although still unattended
by self-consciousness) should be more habitually expressed by signs
than are the practical judgments of animals. Nor need we wonder, in
view of the same considerations, that the predicative phrases as used
by a child at this age show the great advance upon similar phrases as
used by a parrot, in that subjects and predicates are no longer bound
together in particular phrases—or, to revert to a previous simile,
are no longer stereotyped in such particular phrases, but admit of
being used as movable types, in order to construct, by different
combinations, a variety of different phrases. To a talking bird a
phrase, as we have seen, is no more in point of signification than a
single word; while to the child, at the stage which we are considering,
it is very much more than this: it is the separately constructed
vehicle for the conveyance of a particular meaning, which may never
have been conveyed by that or by any other phrase before. But while we
thus attach due importance to so great an advance towards the faculty
of true predication, we must notice, on the one hand, that as yet it is
_not_ true predication in the sense of being the expression of a true
or conceptual judgment; and, on the other hand, we must notice that
the power of thus using words as movable types does not deserve to be
regarded as any wonderful or unaccountable advance in the faculty of
sign-making, when we pay due regard to the several considerations above
stated. The really important point to notice is that, notwithstanding
this great _advance_ towards the faculty of predication, this faculty
_has not yet been reached_: the propositions which are made are
still unattended by self-consciousness: they are not conceptual, but
pre-conceptual.

Given, then, this stage of mental evolution, and what follows? Be it
remembered I am not endeavouring to solve the impossible problem as to
the intrinsic nature of self-consciousness, or how it is that such a
thing is possible. I am merely accepting its existence (and therefore
its possibility) as a fact; and upon the basis of this fact I shall now
endeavour to show how, in my opinion, self-consciousness may be seen to
follow upon the stage of mental evolution which we have here reached.

       *       *       *       *       *

The child, like the animal, is supplied by its logic of recepts with a
world of images, standing as signs of outward objects; with an ejective
knowledge of other minds; and with that kind of recognition of self
as an active, suffering and accountable agent which, following Mr.
Chauncey Wright, I have called “outward self-consciousness.” But, over
and above the animal, the child has at its command, as we have just
seen, the more improved machinery of sign-making which enables it to
signify to other minds (ejectively known) the contents of its receptual
knowledge. Now, among these contents is the child’s perception of the
mental states of others as expressed in their gestures, tones, and
words. These severally receive their appropriate names, and so gain
clearness and precision as ejective images of the corresponding states
experienced by the child itself. “Mama pleased to Dodo” would have
no meaning as spoken by a child, unless the child knew from his own
feelings what is the state of mind which he thus ejectively attributes
to another. Therefore we cannot be surprised to find that at the same
stage of mental evolution the child will say, “Dodo pleased to mama.”
Yet it is evident that we here approach the very borders of true
self-consciousness. “Dodo” is no doubt still speaking of himself in
objective terminology; but he has advanced so far in the interpretation
of his own states of mind as to name them no less clearly than he names
any external objects of sense perception. Thus he is enabled to fix
these states before his mental vision as things which admit of being
_denoted_ by verbal signs, albeit he is not yet able to _denominate_.

The step from this to recognizing “Dodo” as not only the object, but
also the subject of mental changes, is not a large step. The mere act
of attaching verbal signs to inward mental states has the effect of
focussing attention upon those states; and, when attention is thus
focussed habitually, there is supplied the only further condition
required to enable the mind, through its memory of previous states, to
compare its past with its present, and so to reach that apprehension
of continuity among its own states wherein the full introspective
consciousness of self consists.

Again, as Mr. Chauncey Wright observes, “voluntary memory, or
reminiscence, is especially aided by command of language. This is a
tentative process, essentially similar to that of a search for a lost
or missing external object. Trials are made in it to revive a missing
mental image, or train of images, by means of words; and, on the other
hand, to revive a missing name by means of mental images, or even by
other words. It is not certain that this power is an exclusively human
one, as is generally believed, except in respect to the high degree of
proficiency attained by men in its use. It does not appear impossible
that an intelligent dog may be aided by its attention, purposely
directed to spontaneous necessaries, in recalling a missing fact, such
as the locality of a buried bone.”[122]

But whether or not animals possess any power of recollection as
distinguished from memory, there can be no doubt that the use of words
as signs necessarily leads to the cultivation of this faculty, and so
to the clear perception of a continuance of internal or mental states
in which consists the consciousness of an abiding self.

Further, the acquisition of language greatly advances the conception
of self, both as a suffering or feeling agent, and as an active cause;
seeing that both the feelings and the actions of the self are placed
clearly before the mind by means of denotative names, and even, as we
have just seen, by pre-conceptual propositions. Doubtless, also, the
recognition of self in each of these capacities is largely assisted by
the emotions. The expressions of affection, sympathy, praise, blame,
&c., on the part of others, and the feelings of emulation, pride,
triumph, disappointment, &c., on the part of the self, must all tend
forcibly to impress upon the growing child a sense of personality. “It
is when the child’s attention is driven inwards in an act of reflection
on his own actions, as springing from good or bad motives, that he
wakes up to a fuller consciousness of himself.”[123]

The conspiring together of all these factors leads to the gradual
attainment of self-consciousness. I say “gradual,” because the process
is throughout of the nature of a growth. Nevertheless, there is some
reason to think that when this growth has attained a certain point,
it makes, so to speak, a sudden leap of progress, which may be taken
to bear the same relation to the development of the mind as the act
of birth does to that of the body. In neither case is the development
anything like completed. Midway between the slowly evolving phases _in
utero_ and the slowly evolving phases of aftergrowth, there is in the
case of the human body a great and sudden change at the moment when it
first becomes separated from that of its parent. And so, there is some
reason to believe, it is in the case of the human mind. Midway between
the gradual evolution of receptual ideation and the no less gradual
evolution of conceptual, there appears to be a critical moment when
the soul first becomes detached from the nutrient body of its parent
perceptions, and wakes up in the new world of a consciously individual
existence. “Die Schlussprozesse, durch welche jene Trennung des Ich von
der Aussenwelt vor sich geht, geschehen allmälig. Es ist eine langsame
Arbeit, durch die sich die Scheidung bewerkstelligt. Doch diese
Scheidung selber ist stets eine plötzliche That: es ist ein bestimmter
Moment, in welchem das Ich mit einem Mal mit voller Klarheit in der
Seele aufblitzt, und es ist derselbe Moment, in welchem das bewusste
Gedächtniss beginnt, Sehr häufig ist es daher, dass gerade diesses
erste blitzähnliche Aufleuchten des Selbstbewusstseins bis in späte
Jahre noch als deutliche Erinnerung zurückbleibt.”[124]

Of course the evidence upon this point must always be more or less
unsatisfactory—first, because the powers of introspective analysis
at the particular time when they first become nascent must be most
incompetent to report upon the circumstances of their own birth;
and next, because we know how precarious it is to rely on adult
reminiscences of childhood’s experience. Therefore, I have only
mentioned this evidence for what it is worth, in order to remark
that it has no important bearing upon our present subject. Whether
or not there is in the life of every human being some particular
moment between the ages of two and three when the fact of its own
personality is revealed to the growing mind, the results of the present
analysis are in no way affected. For, even if such were supposed to
be invariably the case, it could not be supposed that the revelation
were other than low and feeble to a degree commensurate with the still
almost infantile condition of all the other mental powers. Nor could
it be doubted that this revelation needed to be led up to by that
gradual process of receptual evolution with which my analysis has been
concerned, and which in the terms of our previous analogy we may liken
to the pre-natal life of an embryo. While, on the other hand, as little
can it be doubted that such consciousness of self as is then revealed,
requires to be afterwards supplemented by another prolonged course
of mental evolution in the conceptual sphere, before those completed
faculties of introspective thought are attained, which serve to
difference the mind of a full-grown man from that of a babbling child
almost as widely as the same interval of time is found to difference
the body of an adult from that of a new-born babe.

       *       *       *       *       *

In this brief analysis of the principles which are probably concerned
in the evolution of self-consciousness, I should like to lay particular
stress upon the point in it which I do not think has been sufficiently
noticed by previous writers—namely, the ejective origin of subjective
knowledge. The logic of recepts furnishes both the infant and the
animal with a marvellously efficient store of ejective information.
Indeed, we can scarcely doubt that to a very considerable extent this
information is hereditary: witness the smile of an infant in answer
to a caressing tone, and its cry in answer to a scolding one; not to
mention the still more remarkable cases which we meet with in animals,
such as newly-hatched chickens understanding the different sounds
made to them by the hen, being terror-stricken at the voice of a
hawk, newly-born mammals knowing the voice of their mother, &c.[125]
Moreover, we find that the child, even for a considerable time after
it has begun to use words, manifests a strong tendency to regard all
objects, whether animate or inanimate, as ejects. This fact is a matter
of such general observation that I need not wait to give special
instances. I will, therefore, merely observe that the tendency is not
wholly obliterated even when the faculty of speech has been fully
acquired, and with it a general knowledge of the distinction between
objects as animate and inanimate. Mr. Sully, for instance, gives a case
of this when he records the saying of a little girl of five—“Ma, I
do think this hoop must be alive; it is so sensible; it goes wherever
I want it to.”[126] Again, we meet with the same tendency in the
psychology of uncultured man. Pages might be filled with illustrations
showing that savages all over the world both mentally and expressly
personify, or endow with psychical attributes, the inanimate objects
and forces of nature; while language, even in its most highly developed
forms, still retains the impress of an originally ejective terminology.
And, if Professor Max Müller is right in his generalization that the
personal pronoun “I” is in all languages traceable to roots equivalent
to “This one” (indicative of an accompanying gesture-sign), we have
additional and more particular evidence of the originally ejective
character of the idea of self. Nor is it too much to say that even
civilized man is still under the sway of this innate propensity to
attribute to external things the faculties of feeling and willing of
which he is conscious in himself. On the one side we have proof of this
in the universal prevalence of the hypothesis of psychism in Nature,
while on the other side we meet with further proof in the fact of
psychological analysis revealing that our idea of cause is derived from
our idea of muscular effort.

Now it is evident that in all these cases the tendency which is
shown by the human mind, in every stage of its development, to
regard external phenomena ejectively, arises from man’s intuitive
knowledge—or the knowledge which is given in the logic of recepts—of
his own existence as twofold, bodily and mental. This in his early
days leads him to regard the Ego as an eject, resembling the others
of his kind by whom he is surrounded. But as soon as the power
of pre-conceptual predication has been attained, the child is in
possession of a psychological instrument wherewith to observe his
own mental states; and as soon as attention is thus directed upon
them, there arises that which is implied in every act of such
attention—namely, the consciousness of a self as at once the subject
and object of knowledge.

I may remark that this analysis is not opposed, as at first sight it
may appear to be, to the conclusion with regard to the same subject
which is thus given by Wundt:—“It is only after the child has
distinguished by definite characteristics its own being from that of
other people, that it makes the further advance of perceiving that
these other people are also beings in or for themselves.”[127] In
other words, the attribution of personality to self is prior to the
attribution of personality to others. Now this I do not question,
although I do not think there can be much before or after in these two
concepts. But the point which I have been endeavouring to bring out is
that, prior to either of these concepts, there are two corresponding
recepts—namely, first the receptual apprehension of self as an agent,
and, second, the eject of this receptual apprehension, whereby “other
people” are recognized as agents. Out of these two recepts there
subsequently develop the corresponding concepts of personality. The
order of development, therefore, is:—

  (A) Receptual Subject.       (a) Receptual Eject.

  (B) Conceptual Subject.      (b) Conceptual Eject.

Upon the whole, then, it appears to me perfectly evident that
language is quite as much the antecedent as it is the consequent of
self-consciousness. We have seen that in its first beginnings, or
before the child is able to state a truth as true, what I have called
rudimentary or pre-conceptual predication is concerned only with
existence as objective or ejective: all these propositions, which
are made by children during the first two years of their life, have
reference to objects of sense, states of feeling, &c.; but never to
self as self, and therefore never to truths as true. But as soon
as the protoplasm of predication, or sign-making at this stage of
elaboration, begins to mix freely with the protoplasm of judgment,
or the logic of recepts at that stage of elaboration, an intimate
movement of action and reaction ensues: the judgments are rendered
clearer and more comprehensive by being thrown into the formal shape of
even rudimentary propositions, while the latter are promoted in their
development by the growing powers of judgment. And when this advancing
organization of faculties has proceeded to the extent of enabling the
mind incipiently to predicate its own states, the mental organism may
be said for the first time to be quickening into the life of true
self-consciousness.[128]




CHAPTER XI.

THE TRANSITION IN THE INDIVIDUAL.


We are now, I think, in possession of sufficient material to begin our
answer to the question with which we set out—namely, Is it conceivable
that the human mind can have arisen by way of a natural genesis from
the minds of the higher quadrumana? I maintain that the material now
before us is sufficient to show, not only that this is conceivable, but
inevitable.

First of all we must remember that we share in common with the lower
animals not only perceptual, but also what I have termed receptual
life. Thus far, no difference of kind can be even so much as suggested.
The difference then, be it one of kind or of degree, concerns only
those superadded elements of psychology which are peculiar to man,
and which, following other psychologists, I have termed conceptual.
I say advisedly the _elements_, because it is by no one disputed
that all differences of conceptual life are differences of degree,
or that from the ideation of a savage to that of a Shakespeare there
is unquestionably a continuous ascent. The only question, then, that
obtains is as to the relation between the highest recept of a brute and
the lowest concept of a man.

Now, in considering this question we must first remember to what an
extraordinarily high level of adaptive ideation the purely receptual
life of brutes is able to carry them. If we contrast the ideation of
my cebus, which honestly investigated the mechanical principle of a
screw, and then applied his specially acquired knowledge to screws in
general—if we contrast this ideation with that of palæolithic man,
who for untold thousands of years made no advance upon the chipping
of flints, we cannot say that, when gauged by the practical test of
efficiency or adaptation, the one appears to be very much in advance
of the other. Or, if we remember that these same men never hit upon
the simple expedient of attaching a chipped flint to a handle, so as
to make a hatchet out of a chisel,[129] it cannot be said that in the
matter of mechanical discovery early conceptual life displayed any
great advance upon the high receptual life of my cebus. Nevertheless, I
have allowed—nay insisted—that no matter how elaborate the structure
of receptual knowledge may be, or how wonderful the adaptive action it
may prompt, a “practical inference” or “receptual judgment” is always
separated from a conceptual inference or true judgment by the immense
distinction that it is not itself an object of knowledge. No doubt
it is a marvellous fact that by means of receptual knowledge alone a
monkey should be able to divine the mechanical principle of _a screw_,
and afterwards apply his discovery to all cases of _screws_. But even
here there is nothing to show that the monkey ever _thought_ about
the principle _as_ a principle; indeed, we may rest well assured that
he cannot possibly have done so, seeing that he was not in possession
of the intellectual instruments—and, therefore, of the _antecedent
conditions_—requisite for the purpose. All that the monkey did was
to perceive receptually certain analogies: but he did not _conceive_
them, or constitute them objects of thought _as_ analogies. He was,
therefore, unable to _predicate_ the discovery he had made, or to set
before his own mind as knowledge the knowledge which he had gained.

Or, to take another illustration, the bird which saw three men go
into a building, and inferred that one must still have remained when
only two came out, conducted the inference receptually: the only data
she had were those supplied by differential sense-perceptions. But
although these data were sufficient for the purpose of conducting
what Mr. Mivart calls a “practical inference,” and so of enabling her
to know that a man still remained behind, they were clearly not enough
to enable her to know the numerical relations _as_ relations, or in
any way to predicate to herself, 3-2=1. In order to do this, the bird
would have required to quit the region of receptual knowledge, and
rise to that of conceptual: she would have required in some form or
another to have substituted symbols for ideas. It makes no difference,
so far as this distinction is concerned, when we learn that in dealing
with certain savages “each sheep must be paid for separately: thus,
suppose two sticks of tobacco to be the rate of exchange for one sheep,
it would sorely puzzle a Dammara to take two sheep and give him four
sticks.”[130] All that such facts show is that in some respects the
higher receptual life of brutes attains almost as high a level of
ideation as the lower conceptual life of man; and although this fact
no doubt greatly lessens the difficulty which my opponents allege as
attaching to the supposition that the two were genetically continuous,
it does not in itself dispose of the psychological distinction between
a recept and a concept.

This distinction, as we have now so often seen, consists in a recept
being an idea which is not itself an object of knowledge, whereas a
concept, in virtue of having been named by a self-conscious agent, is
an idea which stands before the mind of that agent _as_ an idea, or
as a state of mind which admits of being introspectively contemplated
as such. But although we have in this distinction what I agree with
my opponents in regarding as the greatest single distinction that is
to be met with in psychology, I altogether object to their mode of
analyzing it. For what they do is to take the concept in its most
highly developed form, and then contrast this with the recept of an
animal. Nay, as we have seen, they even go beyond a concept, and allege
that “the simplest element of thought” is a judgment as bodied forth
in a proposition—_i.e._ _two_ concepts _plus_ the predication of a
relationship between them! Truly, we might as well allege that the
simplest element of matter is H_{2}SO_{4}, or the simplest element
of sound a bar of the C Minor Symphony. Obviously, therefore, or as a
mere matter of the most rudimentary psychological analysis, if we say
that the simplest element of thought is a judgment, we must extend the
meaning of this word from the mental act concerned in full predication,
to the mental act concerned in the simplest conception.

And not only so. Not only have my opponents committed the slovenly
error of regarding a predicative judgment as “the simplest element
of thought;” they have also omitted to consider that even a concept
requires to be analyzed with respect to its antecedents, before this
the really simplest element of thought can be pointed to as proving
a psychological distinction of kind in the only known intelligence
which presents it. Now, the result of my analysis of the concept has
been to show that it is preceded by what I have termed pre-concepts,
which admit of being combined into what I have termed nascent,
rudimentary, or pre-conceptual judgments. In other words, we have seen
that the receptual life of man reaches a higher level of development
than the receptual life of brutes, even before it passes into that
truly conceptual phase which is distinguished by the presence of
self-conscious reflection. In order, therefore, to mark off this higher
receptual life of a human being from the lower receptual life of a
brute, I have used the terms just mentioned.

So much, then, for these several stages of ideation, which I have now
reiterated _ad nauseam_. Turning next to my analysis of their several
modes of expression, or of their translation into their severally
equivalent systems of signs, we have seen that many of the lower
animals are able to communicate their recepts by means of gestures
significant of objects, qualities, actions, desires, &c.; and that in
the only case where they are able to articulate, they so communicate
their recepts by means of words. Therefore, in a sense, these animals
may be said to be using names; but, in order not to confuse this
kind of naming with that which is distinctive of conceptual thought,
I have adopted the scholastic terminology, and called the former
kind of naming an act of denotating, as distinguished from an act of
denominating. Furthermore, seeing that denotative language is able, as
above observed, to signify qualities and actions as well as objects,
it follows that in the higher receptual (_i.e._ pre-conceptual)
stages of ideation, denotative language is able to construct what I
have termed pre-conceptual propositions. These differ from true or
conceptual propositions in the absence of true self-consciousness on
the part of the speaker, who therefore, while communicating receptual
knowledge, or stating truths, cannot yet know his own knowledge, or
state the truths as true. But it does not appear that a pre-conceptual
proposition differs from a conceptual one in any other respect, while
it does appear that the one passes gradually into the other with the
rise of self-consciousness in every growing child. Now, if all these
things are so, we are entitled to affirm that analysis has displayed
an uninterrupted transition between the denotation of a brute and the
predication of a man. For the mere fact that it is the former phase
alone which occurs in the brute, while in the man, _after having
run a parallel course of development_, this phase passes into the
other—the mere fact that this is so cannot be quoted as evidence that
a similar transition never took place in the psychological history of
our species, unless it could be shown that when the transition takes
place in the psychological history of the individual, it does so in
such a sudden and remarkable manner as of itself to indicate that the
intellect of the individual has there and then undergone a change of
kind.

       *       *       *       *       *

Such being an outline sketch of my argument, I will now proceed to
fill in the details, taking in historical order the various stages of
ideation which I have named—_i.e._ the receptual, the pre-conceptual,
and the conceptual.

Seeing that this is, as I apprehend, the central core of the question,
I will here furnish some additional instances of receptual and
pre-conceptual ideation as expressed by denotative and connotative
signs on the part of a child which I carefully observed for the purpose.

At eighteen months old my daughter, who was late in beginning to
speak, was fond of looking at picture-books, and as already stated in
a previous chapter, derived much pleasure from naming animals therein
represented,—saying _Ba_ for a sheep, _Moo_ for a cow, uttering a
grunt for a pig, and throwing her head up and down with a bray for
a horse or an ass. These several sounds and gestures she had been
taught by the nurse as noun-substantives, and she correctly applied
them in every case, whether the picture-book happened to be one with
which she was familiar or one which she had never seen before; and she
would similarly name all kinds of animals depicted on the wall-paper,
chair-covers, &c., in strange houses, or, in short, whenever she met
with representations of objects the nursery names of which she knew.
Thus there is no doubt that, long before she could form a sentence, or
in any proper sense be said to speak, this child was able to denote
objects by voice and gesture. At this time, also, she correctly used
a limited number of denotative words significant of actions—_i.e._
active verbs.

Somewhat later by a few weeks she showed spontaneously the faculty of
expressing an adjective. Her younger brother she had called “Ilda,” and
soon afterwards she extended the name to all young children.[131] Later
still, while looking over her picture-books, whenever she came upon a
representation of a sheep with lambs, she would point to the sheep and
say _Mama-Ba_, while to the lambs she would say _Ilda-Ba_. Similarly
with ducks and ducklings, hens and chickens, and indeed with all the
animals to which she had given names. Here it is evident that _Ilda_
served to convey the generic idea of _Young_, and so, from having been
originally used as a proper or denotative name, was now employed as an
adjective or connotative name. But although it expressed a quality,
the quality was one of so sensible a kind that the adjective amounted
to virtually the same thing as substantive, so far as any faculty of
abstraction was concerned: it was equivalent to the word _Baby_, when
by connotative extension this comes to be used as an adjective in the
apposition _Baby-Ba_ for a lamb, &c.

Almost contemporaneously with the acquisition of adjectives, this child
began to learn the use of a few passive verbs, and words significant
of certain states of feeling; she also added to her vocabulary a few
prepositions indicating space relations, such as _Up_, _Down_, &c.[132]

While these advances were being made, a general progress of the
sign-making faculty was also, and even more conspicuously, shown in
another direction. For speech, in the sense of formal predication, not
having yet begun, the development in question took place in the region
of gesture. She was then (two years) able to express a great many
simple ideas by the combined use of gesture-signs, vocal-tones, and a
large connotative extension of her words. The gesture-signs, however,
were still of the simplest or most receptual order, such as pulling
one by the dress to open a door, pointing to a tumbler to signify
her desire for a drink, &c. That is to say, the indicative stage of
language largely coincided with, or overlapped, the earliest phases of
the denotative and receptually connotative. I have already said that
this indicative stage of language constituted the earliest appearance
of the sign-making faculty which I observed in my own children, at a
time when the only desire expressed seemed to be that of being taken
to the object indicated; and, so far as I can ascertain, this is
universally true of all children. But the point now is, that when the
logic recepts had become more full, the desires expressed by pointing
became of a more and more varied kind, until, at the age of two and a
half (_i.e._ after significant articulation or true word-making had
well set in), the indicative phase of language developed into regular
pantomime, as the following instance will show. Coming into the house
after having bathed in the sea for the first time, she ran to me to
narrate her novel experience. This she did by first pointing to the
shore, then pretending to take off her clothes, to walk into the sea,
and to dip: next, passing her hands up the body to her head, she
signified that the water had reached as high as her hair, which she
showed me was still wet. The whole story was told without the use of a
single articulate sound.

Now, in the case of these illustrations (and many more of the same
kind might be added if needful), we find the same general fact
exemplified—namely, that the earliest phase of language in the young
child is that which I have called the indicative,—_i.e._ tones and
gestures significant of feelings, objects, qualities, and actions.
This indicative phase of language, or sign-making, lasts much longer
in some children than in others (particularly in those who are late in
beginning to speak); and the longer it lasts the more expressive does
it become of advancing ideation. But in all cases two things have to
be observed in connection with it. The first is that, in its earliest
stages, and onwards through a considerable part of its history, it
is precisely identical with the corresponding phases of indicative
sign-making in the lower animals. Thus, for instance, Professor Preyer
observed that at sixteen months his own child—who at that age could
not speak a word—used to make a gesture significant of petitioning
with its hands (“Bittbewegung”), as indicative of desire for something
to be done. This, of course, I choose as an instance of indicative
sign-making at a comparatively high level of development; but it
is precisely paralleled by an intelligent dog which “begs” before
a water-jug to signify his desire for a drink, or before any other
object in connection with which he desires something to be done.[133]
And so it is with children who pull one’s dress towards a closed door
through which they wish to pass, significantly cry for what they want
to possess, or to have done for them, &c.: children are here doing
exactly what cats and dogs will do under similar circumstances.[134]
And although many of the gesture-signs of children at this age (_i.e._
up to about eighteen months) are not precisely paralleled by those
of the lower animals, it is easy to see that where there is any
difference it is due to different circumstances of bodily shape, social
conditions, &c.: it is not due to any difference of ideation. That
the kind of ideation which is expressed by the indicative gestures
of young children is the same as that which prompts the analogous
gestures of brutes, is further shown by the fact that, even before
any articulate words are uttered, the infant (like the animal) will
display an understanding of many articulate words when uttered in its
presence, and (also like the animal) will respond to such words by
appropriate gestures. For instance, again to quote Preyer, he found
that his hitherto speechless infant was able correctly to point to
certain colours which he named; and although, as far as I am aware, no
one has ever tried to teach an animal to do this, we know that trained
dogs will display an even better understanding of words by means of
appropriate gestures.[135]

The other point which has to be noticed in connection with these early
stages of indicative sign-making in the young child is that, sooner
or later, they begin to overlap the earliest stage of articulate
sign-making, or verbal denotation. In other words, denotative
sign-making never begins to occur until indicative sign-making has
advanced considerably; and when denotative sign-making does begin,
it advances parallel with indicative: that is to say, both kinds of
sign-making then proceed to develop simultaneously. But when the
vocabulary of denotation has been sufficiently enriched to enable
the child to dispense with the less efficient material furnished
by indication, indicative signs gradually become starved out by
denotative, and words replace gestures.

       *       *       *       *       *

So far, then, as the earliest or indicative phase of language is
concerned, no difference even of degree can be alleged between the
infant and the animal. Neither can any such difference be alleged with
respect to the earliest exhibitions of the next phases of language,
namely, the denotative and receptually connotative. For we have
seen that the only animals which happen to be capable of imitating
articulate sounds will use these sounds with a truly denotative
significance. Moreover, as we have also seen, within moderate limits
they will even extend such denotative significance to other objects
seen to belong to the same class or kind—thus raising the originally
denotative sign to an incipiently connotative value. And although these
receptually connotative powers of a parrot are soon surpassed by those
of a young child, we have further seen that this is merely owing to
the rapid advance in the _degree_ of receptual life which takes place
in the latter—or, in other words, that if a parrot resembled a dog in
being able to see the resemblance between objects and their pictures,
and also in being so much more able to understand the meanings of
words, then, without doubt, their connotative extension of names would
proceed further than it does; and hence in this matter the parallel
between a parrot and child would proceed further than it does. The
only reason, therefore, why a child thus gradually surpasses a parrot
in the matter of connotation, is because the receptual life of a child
gradually rises to that of a dog—as I have already proved by showing
that the indicative or gesture-signs used by a child after it has thus
surpassed the parrot, are psychologically identical with those which
are used by a dog. Moreover, where denotation is late in beginning and
slow in developing—as in the case of my own daughter—these indicative
signs admit, as we have seen, of becoming much more highly perfected,
so that under these circumstances a child of two years will perform
a little pantomime for the purpose of relating its experiences. Now,
this fact enables me to dispense with the imaginary comparison of a
dog that is able to talk, or of a parrot as intelligent as a dog; for
the fact furnishes me with the converse case of a child _not_ able to
talk at the usual age. No one can suggest that the intelligence of such
a child at two years old differs in kind from that of another child
of the same age, who, on account of having been earlier in acquiring
the use of words, can afford to become less proficient in the use of
gestures.[136] The case of a child late in talking may therefore be
taken as a psychological index of the development of human ideation of
the receptual order, which by accident admits of closer comparison with
that of the higher mammalia than is possible in the case of a child who
begins to talk at the usual age. But, as regards the former case, we
have already seen that the gestures begin by being much less expressive
than those of a dog, then gradually improve until they become
psychologically identical, and, lastly, continue in the same gradual
manner along the same line of advance. Therefore, if in this case no
difference of kind can be alleged _until_ the speaking age is reached,
neither can it be alleged _after_ the speaking age is reached in the
case where this happens to be earlier. Or, in the words previously
used, if a dog like a parrot were able to use verbal signs, or if a
parrot were equal in intelligence to a dog, the connotative powers of a
child would continue parallel with those of a brute through a somewhat
longer reach of psychological development than we now find to be the
case.

Remembering, then, that brutes so low in the psychological scale as
talking birds reach the level of denotating objects, qualities, &c.;
remembering that some of these birds will extend their denotative names
to objects and qualities conspicuously belonging to the same class;
remembering, further, that all children before they begin to speak have
greatly distanced the talking birds in respect of indicative language
or gesture-signs, while some children (or those late in beginning to
speak) will raise this form of language to the level of pantomime,
thus proving that the receptual ideation of infants just before they
begin to speak is invariably above that of talking birds, and often
far above that of any other animal;—remembering all these things, I
say it would indeed be a most unaccountable fact if children, soon
after they do begin to speak, did _not_ display a great advance upon
the talking birds in their use of denotative signs, and also in their
extension of such signs into connotative words. As we have seen, it
must be conceded by all prudent adversaries that, before he is able to
use any of these signs, an infant is moving in the receptual sphere of
ideation, and that this sphere is already (between one and two years)
far above that of the parrot. Yet, like the parrot, one of the first
uses that he makes of these signs is in the denotation of individual
objects, &c. Next, like the more intelligent parrots, he extends the
meaning of his denotative names to objects most obviously resembling
those which were first designated. And from that point onwards he
rapidly advances in his powers of connotative classification. But can
it be seriously maintained, in view of all the above considerations,
that this rapid advance in the powers of connotative classification
betokens any difference of kind between the ideation of the child and
that of the bird? If it is conceded (as it must be unless my opponents
commit argumentative suicide), that before he could speak at all the
infant was confined to the receptual sphere of ideation, and that
within this sphere his ideation was already superior to the ideation
of a bird,—this is merely to concede that analogies _must_ strike the
child which are somewhat too remote to strike the bird. Therefore,
while the bird will only extend its denotative name from one kind of
dog to another, the child, after having done this, will go on to apply
the name to an image, and, lastly, to the picture of a dog. Surely no
one will be fatuous enough to maintain that here, at the commencement
of articulate sign-making, there is any evidence of generic distinction
between the human mind and the mind of even so poor a representative
of animal psychology as we meet with in a parrot. But, if no such
distinction is to be asserted here, neither can it be asserted anywhere
else, until we arrive at the stage of human ideation where the mind
is able to contemplate that ideation as such. So far, therefore, as
the stages which we are now considering are concerned (_i.e._ the
denotative and receptually connotative), I submit that my case is made
out. And yet these are really the most important stages to be clear
about; for, on account of their having been ignored by nearly all
writers who argue that there is a difference of kind between man and
brute, the most important—because the initial—stages of transition
have been lost sight of, and the fully developed powers of human
thought contrasted with their low beginnings in the brute creation,
without any attention having been paid to the probable history of
their development. Hitherto, so far as I can find, no psychologist has
presented clearly the simple question whether the faculty of naming is
always and necessarily co-extensive with that of _thinking the names_;
and, therefore, the two faculties have been assumed to be one and the
same. Yet, as I have shown in an earlier chapter, even in the highest
forms of human ideation we habitually use names without waiting to
think of them as names—which proves that even in the highest regions
of ideation the two faculties are not _necessarily_ coincident.[137]
And here I have further shown that, whether we look to the brute or to
the human being, we alike find that the one faculty is in its inception
_wholly independent_ of the other—that there are connotative names
before there are any denominative thoughts, and that these connotative
names, when they first occur in brute or child, betoken no further
aptitude of ideation than is betokened by those stages in the language
of gesture which they everywhere overlap. The named recepts of a parrot
cannot be held by my opponents to be true concepts, any more than the
indicative gestures of an infant can be held by them to differ in kind
from those of a dog.

       *       *       *       *       *

I submit, then, that neither as regards the indicative, the denotative,
nor the connotative stages of sign-making is it argumentatively
possible to allege any difference of kind between animal and human
intelligence—apart, I mean, from any evidence of self-consciousness in
the latter, or so long as the intelligence of either is moving in what
I have called the receptual sphere. Let us, then, next consider what
I have called the pre-conceptual stage of ideation, or that higher
receptual life of a child which, while surpassing the receptual life of
any brute, has not yet attained to the conceptual life of a man.

From what I have already said it must, I should suppose, be now
conceded that, at the place where the receptual life of a child
first begins to surpass the receptual life of any other mammal, no
psychological difference of kind can be affirmed. Let us, therefore,
consent to tap this pre-conceptual life at a considerably higher
level, and analyze the quality of ideation which flows therefrom: let
us consider the case of a child about two years old, who is able to
frame such a rudimentary, communicative, or pre-conceptual proposition
as _Dit ki_ (Sister is crying). At this age, as already shown, there
is no consciousness of self as a thinking agent, and, therefore, no
power of stating a truth as true. _Dit_ is the denotative name of
one recept, _ki_ the denotative name of another: the object and the
action which these two recepts severally represent happen to occur
together before the child’s observation: the child therefore denotes
them both simultaneously—i.e. _brings than into apposition_. This
it does by merely following the associations previously established
between the recept of a familiar object with its denotative name
_dit_, and the recept of a frequent action with its denotative name
_ki_. The apposition in consciousness of these two recepts, with their
corresponding denotations, is thus effected _for_ the child by what
may be termed _the logic of events_: it is not effected _by_ the child
in the way of any intentional or self-conscious grouping of its ideas,
such as we have seen to constitute the distinguishing feature of the
logic of concepts.

Such being the state of the facts, I put to my opponents the following
dilemma. Either you here have judgment, or else you have not. If you
hold that this is judgment, you must also hold that animals judge,
because I have proved a ready that (according to your own doctrine
as well as mine) the only point wherein it can be alleged that the
faculty of judgment differs in animals and in man consists in the
presence or absence of self-consciousness. If, on the other hand,
you answer that here you have not judgment, inasmuch as you have not
self-consciousness, I will ask you at what stage in the subsequent
development of the child’s intelligence you would consider judgment
to arise? If to this you answer that judgment first arises when
self-consciousness arises, I will ask you to note that, as already
proved, the growth of self-consciousness is itself a gradual process;
so that, according to your present limitation of the term judgment,
it becomes impossible to say when this faculty does arise. In
point of fact, it grows by stages, _pari passu_ with the growth of
self-consciousness. But, if so, where the faculty of stating a truth
perceived passes into the higher faculty of perceiving the truth as
true, there must be a continuous series of gradations connecting the
one faculty with the other. Up to the point where this series of
gradations begins, we have seen that the mind of an animal and the
mind of a man are parallel, or not distinguishable from each other by
any one principle of psychology. Will you, then, maintain that up to
this time the two orders of psychical existence are identical in kind,
but that during its ascent through this final series of gradations the
human mind in some way becomes distinct in kind, not merely from the
mind of animals, _but also from its own previous self_? If so, I must
at this point part company with you in argument, because at this point
your argument ends in a contradiction. If A and B are affirmed to be
similar in origin or kind, and if B is affirmed to grow into C—or to
differ from both A and B only in degree,—it becomes a contradiction
further to affirm that C differs from A in kind. Therefore I submit
that, so far as the pre-conceptual stage of ideation is concerned, it
is still argumentatively impossible for my opponents to show that there
is any psychological difference of kind between man and brute.

As regards this stage of ideation, then, I claim to have shown that,
just as there is a pre-conceptual kind of naming, wherein originally
denotative words are progressively extended through considerable
degrees of connotative meaning; so there is a pre-conceptual kind
of predication, wherein denotative and connotative terms are brought
together without any conceptual cognizance of the relation thus
virtually alleged between them. For I have proved in the last chapter
that it is not until its third year that a child acquires true or
conceptual self-consciousness, and therefore attains the condition
to true or conceptual predication. Yet long before that time, as I
have also proved, the child forms what I have called rudimentary,
or pre-conceptual, and, therefore, _unthinking_ propositions. Such
propositions, then, are statements of truth made for the practical
purposes of communication; but they are not statements of truth as
true, and therefore not, strictly speaking, propositions at all.
They are translations of the logic of recepts; but not of the logic
of concepts. For neither the truth so stated, nor the idea thus
translated, can ever have been placed before the mind as itself an
object of thought. In order to have been thus placed, the mind must
have been able to dissociate this its product from the rest of its
structure—or, as Mr. Mivart says, to make the things affirmed “exist
_beside_ the judgment, not _in_ it.” And, in order to do this, the
mind must have attained to self-consciousness. But, as just remarked,
such is not yet the case with a child of the age in question; and
hence we are bound to conclude that before there is judgment or
predication in the sense understood by psychologists (conceptual),
there is judgment and predication of a lower order (pre-conceptual),
wherein truths are stated for the sake of communicating simple ideas,
while the propositions which convey them are not themselves objects of
thought. And, be it carefully observed, predication of this rudimentary
or pre-conceptual kind is accomplished by the mere apposition of
denotative signs, in accordance with the general principles of
association. _A_ being the denotative name of an object _a_, and _B_
the denotative name of a quality or action _b_, when _a b_ occur
together in nature, the relation between them is pre-conceptually
affirmed by the mere act of bringing into apposition the corresponding
denotations _A B_—an act which is rendered inevitable by the
elementary laws of psychological association.[138]

       *       *       *       *       *

The matter, then, has been reduced to the last of the three stages
of ideation which have been marked out for discussion—namely, the
conceptual. Now, whether or not there is any difference of kind
between the ideation which is capable and the ideation which is not
capable of itself becoming an object of thought, is a question which
can only be answered by studying the relations that obtain between
the two in the case of the growing child. But, as we have seen, when
we do study these relations, we find that they are clearly those of a
gradual or continuous passage of the one ideation into the other—a
passage, indeed, so gradual and continuous that it is impossible, even
by means of the closest scrutiny, to decide within wide limits where
the one begins and the other ends. Therefore I need not here recur
to this point. Having already shown that the very condition to the
occurrence of conceptual ideation (namely, self-consciousness) is of
gradual development in the growing child, it is needless to show at any
greater length that the development of conceptual out of pre-conceptual
ideation is of a similarly gradual occurrence. This fact, indeed, is in
itself sufficient to dispose of the allegation of my opponents—namely,
that there is evidence of receptual ideation differing from conceptual
in origin or kind. Only if it could be shown—either that the
receptual ideation of an infant differs in kind from that of an animal,
or that the pre-conceptual ideation of a child so differs from the
preceding receptual ideation of the same child, or lastly, that this
pre-conceptual ideation so differs from the succeeding conceptual
ideation—only if one or other of the alternatives could be proved
would my opponents be able to justify their allegation. And, as a
mere matter of logic, to prove either of the last two alternatives
would involve a complete reconstruction of their argument. For at
present their argument goes upon the assumption that throughout all
the phases of its development a human mind is one in kind—that it is
nowhere fundamentally changed from one order of existence to another.
But in case any subtle opponent should suggest that, although I have
proved the first of the above three alternatives untenable—and,
therefore, that there is no difference even of degree between the
mind of an infant and that of an animal,—I have nevertheless ignored
the possibility that in the subsequent development of every human
being a special miracle may be wrought, which regenerates that mind,
gives it a new origin, and so changes it as to kind—in case any
one should suggest this, I here entertain the two last alternatives
as logically possible. But, even so, as we have now so fully seen,
study of the child’s intelligence while passing through its several
phases of development yields no shadow of evidence in favour of
any of these alternatives; while, on the contrary, it most clearly
reveals the fact that transition from each of the levels of ideation
to the next above it is of so gradual and continuous a character that
it is practically impossible to draw any real lines of demarcation
between them. This, then, I say is in itself enough to dispose of the
allegation of my opponents, seeing that it shows the allegation to be,
not only gratuitous, but opposed to the whole body of evidence which
is furnished by a study of the facts. Nevertheless, still restricting
ourselves to grounds of psychology alone, there remains two general
and important considerations of an independent or supplementary kind,
which tend strongly to support my side of the argument. These two
considerations, therefore, I will next adduce.

       *       *       *       *       *

The first consideration is, that although the advance to
self-consciousness from lower grades of mental development is no doubt
a very great and important matter, it is not so great and important
in comparison with what this development is afterwards destined to
become, as to make us feel that it constitutes any distinction _sui
generis_—or even, perhaps, the principal distinction—between the man
and the brute. For while, on the one hand, we have now fully seen that,
given the protoplasm of judgment and of predication as these occur in
the young child (or as they may be supposed to have occurred in our
semi-human ancestors), and self-consciousness must needs arise; on the
other hand, there is evidence to show that when self-consciousness
does arise, and even when it is fairly well developed, the powers
of the human mind are still in an almost infantile condition. Thus,
for instance, I have observed in my own children that, while before
their third birthday they employed appropriately and always correctly
the terms “I,” “my,” “self,” “myself,” at that age their powers of
reasoning were so poorly developed as scarcely to be in advance of
those which are exhibited by an intelligent animal. To give only one
instance of this. My little girl when four and a half years old—or
nearly two years after she had correctly used the terms indicative
of true self-consciousness—wished to know what room was beneath the
drawing-room of a house in which she had lived from the time of her
birth. When she asked me to inform her, I told her to try to think out
the problem for herself. She first suggested the bath-room, which was
not only above the drawing-room, but also at the opposite side of the
house; next she suggested the dining-room, which, although below the
drawing-room, was also at the other side of the house; and so on, the
child clearly having no power to think out so simple a problem as the
one which she had spontaneously desired to solve. From which (as from
many other instances on my notes in this connection) I conclude that
the genesis of self-consciousness marks a comparatively low level in
the evolution of the human mind—as we might expect that it should, if
its genesis depends on the not unintelligible conditions which I have
endeavoured to explain in the last chapters. But, if so, does it not
follow that great as the importance of self-consciousness afterwards
proves to be as a condition to the higher development of ideation,
in itself, or in its first beginning, it does not betoken any very
perceptible improvement upon those powers of pre-conceptual ideation
which it immediately follows? In other words, there is thus shown to
be even less reason to regard the advent of self-consciousness as
marking a psychological difference of kind, than there would be so
to regard the advent of those higher powers of conceptual ideation
which subsequently—though as gradually—supervene between early
childhood and youth. Yet no one has hitherto ventured to suggest that
the intelligence of a child and the intelligence of a youth display a
difference of kind.

Or, otherwise stated, the psychological interval between my cebus and
my child (when the former successfully investigated the mechanical
principle of the screw by means of his highly developed receptual
faculties, while the latter unsuccessfully attempted to solve a most
simple topographical problem by means of her lowly developed conceptual
faculties), was assuredly much less than that which afterwards
separated the intelligence of my child from this level of its own
previous self. Therefore, on merely psychological grounds, I conclude
that there would be better—or _less bad_—reasons for alleging that
there is an observable difference of kind between the lowest and the
highest levels of conceptual ideation, than there is to allege that any
such difference obtains between the lowest level of conceptual ideation
and the highest level of receptual.

“The greatest of all distinctions in biology,” when it first arises,
is thus seen to lie in its _potentiality_ rather than in its _origin_.
Self-consciousness is, indeed, the condition to an immeasurable
change in the mind which presents it; but, in order to become so, it
must be itself conditioned: it must itself undergo a long and gradual
development under the guiding principles of a natural evolution.

       *       *       *       *       *

And, now, lastly, the second supplementary consideration which I have
to adduce is, that even in the case of a fully developed self-conscious
intelligence, both receptual and pre-conceptual ideation continue to
play an important part. That is to say, even in the full-summed powers
of the human intellect, the three descriptions of ideation which I have
distinguished are so constantly and so intimately blended together,
that analysis of the adult mind corroborates the fact already yielded
by analysis of the infantile mind, namely, that the distinctions (which
I have been obliged to draw in order to examine the allegations of my
opponents) are all essentially or intrinsically artificial. My position
is that Mind is everywhere continuous, and if for purposes of analysis
or classification we require to draw lines of demarcation between the
lower and the higher faculties thereof, I contend that we should only
do so as an evolutionist classifies his animal or vegetable species:
higher or lower do not betoken differences of _origin_, but differences
of _development_. And just as the naturalist finds a general
corroboration of this view in the fact that structural and functional
characters are carried upwards from lower to higher forms of life, thus
knitting them all together in the bonds of organic evolution; so may
the psychologist find that even the highest forms of human intelligence
unmistakably share the more essential characters met with in the lower,
thus bearing testimony to their own lineage in a continuous system of
mental evolution.

Let us, then, briefly contemplate the relations that obtain in the
adult human mind between the boasted faculties of conceptual judgment,
and the lower faculties of non-conceptual. Although I agree with my
opponents in holding that predication (in the strict sense of the
term) is dependent on introspection, I further hold that not every
statement made by adult man is a predication in this sense: the vast
majority of our verbal propositions are made for the practical purposes
of communication, or without the mind pausing to contemplate the
propositions as such in the light of self-consciousness. When I say “A
negro is black,” I do not require to think all the formidable array of
things that Mr. Mivart says I affirm[139]; and, on the other hand, when
I perform an act of conscious introspection, I do not always require
to perform an act of mental predication. No doubt in many cases, or in
those where highly abstract ideation is concerned, this independence of
the two faculties arises from each having undergone so much elaboration
by the assistance which it has derived from the other, that both are
now, so to speak, in possession of a large body of organized material
on which to operate, without requiring, whensoever they are exercised,
to build up the structure of this material _ab initio_. Thus, to take
an example, when I say “Heat is a mode of motion,” I am using what
is now to me a merely verbal sign which expresses an external fact:
I do not require to examine my own ideas upon the abstract terms in
the abstract relation which the proposition sets forth. But for the
_original attainment_ of these ideas I had to exercise many and complex
efforts of conceptual thought, without the previous occurrence of which
I should not now have been able to use, with full understanding of its
import, this verbal sign. Thus all such predications, however habitual
and mechanical they may become, must at some time have required the
mind to examine the ideas which they announce. And, similarly, all acts
of such mental examination—_i.e._ all acts of introspection,—however
superfluous they may now appear when their known product is used for
further acts of mental examination, must originally have required the
mind to pause before them and make to itself a definite statement or
predication of their meaning.[140]

But although I hold this to be the true explanation of the _apparent_
independence of predication and introspection in all cases of highly
abstract thought, I am firmly convinced that in all cases where
those lower orders of ideation to which I have so often referred as
receptual and pre-conceptual are concerned, the independence is not
only _apparent_, but _real_. This, indeed, I have already proved
_must_ be the case with the pre-conceptual propositions of a young
child, inasmuch as such propositions are then made in the absence of
self-consciousness, or of the necessary condition to their being _in
any degree_ introspective. But the point now is, that even in the
adult human mind non-conceptual predication is habitual, and that, in
cases where only receptual ideation is concerned, predication of this
kind need _never have been_ conceptual. For, as Mill very truly says,
“it will be admitted that, by asserting the proposition, we wish to
communicate information of that physical fact (namely, that the summit
of Chimborazo is white), and are not thinking of the names, except as
the necessary means of making that communication. The meaning of the
proposition, therefore, is that the individual thing denoted by the
subject has the attributes connoted by the predicate.”[141]

Now, if it is thus true that even in ordinary predication we may not
require to take conceptual cognizance of the matter predicated—having
to do only with the apposition of names immediately suggested by
association,—the ideation concerned becomes so closely affiliated
with that which is expressed in the lower levels of sign-making, that
even if the connecting links were not supplied by the growing child,
no one would be justified, on psychological grounds alone, in alleging
any difference of kind between one level and another. The object of
all sign-making is primarily that of communication, and from our
study of the lower animals we know that communication first has to do
exclusively with recepts, while from our study of the growing child we
know that it is the signs used in the communication of recepts which
first lead to the formation of concepts. For concepts are first of all
named recepts, known as such; and we have seen in previous chapters
that this kind of knowledge (_i.e._ of names as names) is rendered
possible by introspection, which, in turn is reached by the naming of
self as an agent. But even after the power of conceptual introspection
has been fully reached, demand is not always made upon it for the
communication of merely receptual knowledge; and therefore it is that
not every proposition requires to be introspectively contemplated as
such before it can be made. Given the power of denotative nomination on
the one hand, and the power of even the lowest degree of connotative
nomination on the other, and all the conditions are furnished to
the formation of non-conceptual statements, which differ from true
propositions only in that they do not themselves become objects of
thought. And the only difference between such a statement when made by
a young child, and the same statement when similarly made by a grown
man, is that in the former case it is not even _potentially_ capable of
itself becoming an object of thought.

       *       *       *       *       *

Here, then, the psychological examination of my opponents’ position
comes to an end. And, in the result, I claim to have shown that in
whatever way we regard the distinctively human faculty of conceptual
predication, it is proved to be but a higher development of that
faculty of receptual communication, the ascending degrees of which
admit of being traced through the brute creation up to the level which
they attain in a child during the first part of its second year,—after
which they continue to advance uninterruptedly through the still
higher receptual life of the child, until by further though not less
imperceptible growth they pass into the incipiently conceptual life
of a human mind—which, nevertheless, is not even then nearly so far
removed from the intelligence of the lower animals, as it is from that
which in the course of its own subsequent evolution it is eventually
destined to become.




CHAPTER XII.

COMPARATIVE PHILOLOGY.


We have now repeatedly seen that there is only one argument in favour
of the view that the elsewhere continuous and universal process of
evolution—mental as well as organic—was interrupted at its terminal
phase, and that this argument stands on the ground of psychology. But
we have also seen that even upon this its own ground the argument
admits of abundant refutation. In order the more clearly to show that
such is the case, I have hitherto designedly kept my discussion within
the limits of psychological science. The time, however, has now come
when I can afford to take a new point of departure. It is to Language
that my opponents appeal: to Language they shall go.

In previous chapters I have more than once remarked that the science
of historical psychology is destitute of fossils: unlike pre-historic
structures, pre-historic ideas leave behind them no record of their
existence. But now a partial exception must be taken to this general
statement. For the new science of Comparative Philology has revealed
the important fact that, if on the one hand speech gives _ex_pression
_to_ ideas, on the other hand it receives _im_pression _from_ them,
and that the impressions thus stamped are surprisingly persistent.
The consequence is that in philology we possess the same kind of
unconscious record of the growth and decay of ideas, as is furnished by
palæontology of the growth and decay of species. Thus viewed, language
may be regarded as the stratified deposit of thoughts, wherein they lie
embedded ready to be unearthed by the labours of the man of science.

In now turning to this important branch of my subject, I may remark
_in limine_ that, like all the sciences, philology can be cultivated
only by those who devote themselves specially to the purpose. My
function, therefore, will here be that of merely putting together the
main results of philological research, so far as this has hitherto
proceeded, and so far as these results appear to me to have any bearing
upon the “origin of human faculty.” Being thus myself obliged to rely
upon authority, where I find that authorities are in conflict—which,
I need hardly say, is often the case—I will either avoid the points
of disagreement, or else state what has to be said on both sides of
the question. But where I find that all competent authorities are in
substantial agreement, I will not burden my exposition by tautological
quotations.

Among the earlier students of language it was a moot question whether
the faculty had its origin in Divine inspiration or in human invention.
So long as the question touching the origin of language was supposed
to be restricted to one or other of these alternatives, the special
creationists in this department of thought may be regarded as having
had the best of the argument. And this for the following reasons. Their
opponents, for the most part, were unfairly handicapped by a general
assumption of special creation as regards the origin of man, and also
by a general belief in the confusion of tongues at the Tower of Babel.
The theory of evolution having been as yet unformulated, there was an
antecedent presumption in favour of the Divine origin of speech, since
it appeared in the last degree improbable that Adam and Eve should
have been created “with full-summed powers” of intellect, without the
means of communicating their ideas to one another. And even where
scientific investigators were not expressly dominated by acceptance
of the biblical cosmology, many of them were nevertheless implicitly
influenced by it, to the extent of supposing that if language were not
the result of direct inspiration, it can only have been the result of
deliberate invention. But against this supposition of language having
been deliberately invented, it was easy for orthodox opponents to
answer—“Daily experience informs us, that men who have not learned to
articulate in their childhood, never afterwards acquire the faculty of
speech but by such helps as savages cannot obtain; and therefore, if
speech were invented at all, it must have been either by children who
were incapable of invention, or by men who were incapable of speech.
A thousand, nay, a million, of children could not think of inventing
a language. While the organs are pliable, there is not understanding
enough to frame the conception of a language; and by the time that
there is understanding, the organs are become too stiff for the task,
and therefore, say the advocates for the Divine origin of language,
reason as well as history intimates that mankind in all ages must have
been speaking animals—the young having constantly acquired this art by
imitating those who are older; and we may warrantably conclude that our
first parents received it by immediate inspiration.”[142]

There remained, however, the alternative that language might have been
the result neither of Divine inspiration nor of human invention; but of
natural growth. And although this alternative was clearly perceived by
some of the earlier philologists, its full significance could not be
appreciated before the advent of the general theory of evolution.[143]
Nevertheless, it is here of interest to observe that the theory of
evolution was clearly educed from, and applied to, the study of
languages by some of the more scientific philologists, before it had
been clearly enunciated by naturalists. Thus, for instance, Dr. Latham,
while criticizing the passage above quoted, wrote in 1857:—“In the
actual field of language, the lines of demarcation are less definitely
marked than in the preceding sketch. The phenomena of growth, however,
are, upon the whole, what it suggests.... In order to account for the
existing lines of demarcation, which are broad and definite, we must
bear in mind a fresh phenomenon, viz. the spread of one dialect at the
expense of others, a fact which obliterates intermediate forms, and
brings extreme ones into geographical juxtaposition.”[144]

Now, at the present day—owing partly to the establishment of the
doctrine of evolution in the science of biology, but much more to
direct evidence furnished by the science of philology itself—students
of language are unanimous in their adoption of the developmental
theory. Even Professor Max Müller insists that “no student of the
science of language can be anything but an evolutionist, for, wherever
he looks, he sees nothing but evolution going on all around him;”[145]
while Schleicher goes so far as to say that “the development of new
forms from preceding forms can be much more easily traced, and this on
even a larger scale, in the province of words, than in that of plants
and animals.”[146]

Here, however, it is needful to distinguish between language and
languages. A philologist may be firmly convinced that all languages
have developed by way of natural growth from those simplest elements,
or “roots,” which we shall presently have to consider. But he may
nevertheless hesitate to conclude, with anything like equal certainty,
that these simplest elements were themselves developed from still
lower ingredients of the sign-making faculty; and hence that not only
all languages in particular, but the faculty of language in general,
has been the result of a natural evolution.

Here then, let it be noted, we are in the presence of exactly the
same distinction with regard to the origin of language, as we were at
the beginning of this treatise with regard to the origin of man. For
we there saw that while we have the most cogent historical evidence
in proof of the principles of evolution having governed the progress
of civilization, we have no such direct evidence of the descent of
man from a brutal ancestry. And here also we find that, so long as
the light of history is able to guide us, there can be no doubt that
the principles of evolution have determined the gradual development
of languages, in a manner strictly analogous to that in which they
have determined the ever-increasing refinement and complexity of
social organization. Now, in the latter case we saw that such direct
evidence of evolution from lower to higher levels of culture renders
it well-nigh certain that the method must have extended backwards
beyond the historical period; and hence, that such direct evidence
of evolution uniformly pervading the historical period, in itself
furnishes a strong _primâ facie_ presumption that this period was
itself reached by means of a similarly gradual development of human
faculty. And thus, also, it is in the case of language. If philology
is able to prove the fact of evolution in all known languages as far
back as the primitive roots out of which they have severally grown, the
presumption becomes exceedingly strong that these earliest and simplest
elements, like their later and more complex products, were the result
of a natural growth.

Nevertheless, as I have said, it is important to distinguish between
demonstrated fact and speculative inference, however strong; and,
therefore, I will begin by briefly stating the stages of evolution
through which languages are now generally recognized by philologists to
have passed, without at present considering the more difficult question
as to the origin of roots.

Supposing we take such a word as “uncostliness.” Obviously here the
“un” the “li” and the “ness” are derivative appendages, demonstrative
elements, suffixes and affixes, or whatever else we care to call
_modifying constants_ which the speakers of a language are in the
habit of adding to their root-words, for the sake of ringing upon
those words whatever changes of meaning occasion may require. These
modifying constants, of course, have all had a history, which often
admits of being traced. Thus, for instance, in the above illustration,
we know that the “li” is an abbreviation of what used to be pronounced
as “like;” the “ness,” however, being older than the English language;
while the “un” dates back still further. The word “cost,” then, is here
the root, as far as English is concerned—though it can be followed
(through the Latin _con-sta_) to an Aryan root, signifying “stand.”

These modifying constants, moreover, are not restricted to suffixes,
infixes, and affixes attached to roots, so as to constitute single (or
compound) words: they also occur as themselves separate words, which
admit of being built into the structure of sentences as pronouns,
adverbs, prepositions, &c. And they may occur likewise as so-called
“auxiliary verbs,” in the case of some languages, while in the case of
others their functions are served by grammatical “inflection” of the
words themselves. Thus, according to the “genius” of a language, its
roots are made to lend themselves to significant treatment in different
ways, or according to different methods. But in all cases the roots are
present, and serve as what may be termed the back-bone of a language:
the demonstrative elements, in whatever form they appear, are merely
what I have termed modifying constants.

From this general fact we may be prepared to expect, on the theory
of evolution, that in all languages the roots should be the
oldest elements; those elements which serve only the function of
“demonstrating” the particular meaning which is to be assigned to the
roots on particular occasions, we should expect to have been of later
growth. For they serve only the function of giving specific meanings
to the general meanings already present in the roots; and, therefore,
in the absence of the roots would themselves present no meaning at
all. Consequently, as I have said, we should antecedently expect to
find that the roots are the earliest discoverable (though not on this
account necessarily the most primitive) elements of all languages. And
this, as a general rule, is what we do find. In tracing back the family
tree of any group of languages, different demonstrative elements are
found on different branches, though all these branches proceed from
(_i.e._ are found to contain) the same roots. Of course these roots
may be variously modified, both as to sound and the groups of words
to which in the different branches they have given origin; but such
divergent evolution merely tends to corroborate the proof of a common
descent among all the branches concerned.[147]

I have said that all philologists now agree in accepting the doctrine
of evolution as applied to languages in general; while there is no such
universal agreement touching the precise method or history of evolution
in the case of particular languages. I will, therefore, first give a
brief statement of the main facts of language-structure, and afterwards
render an equally short account of the different views which are
entertained upon the question of language-development. Or, to borrow
terms from another science, I will first deal with the morphology
of the main divisions of the language-kingdom, and then proceed to
consider the question of their phylogeny.

       *       *       *       *       *

More than a thousand languages exist as “living” languages, no one
of which is intelligible to the speakers of another. These separate
languages, however, are obviously divisible into families—all the
members of each family being more or less closely allied, while members
of different families do not present any such evidence of genetic
affinity. The test of genetic affinity is resemblance in structure,
grammar, and roots. Judged by this test, the thousand or more living
languages are classified by Professor Friedrich Müller under “about one
hundred families.”[148] Therefore, again to borrow biological terms,
we may say that there are about one thousand existing “species” of
language, which fall into about one hundred “genera”—all the species
in each genus being undoubtedly connected by the ties of genetic
affinity.

But besides these species and genera of language, there are what may
be termed “orders”—or much larger divisions, each comprising many of
the genera. By philologists these orders are usually called “groups,”
and whether or not there is any genetic relation among them is still
an unsettled question. From the very earliest days of true linguistic
research, three of these groups have been recognized, and called
respectively, (1) the Isolating, (2) the Agglutinative, and (3) the
Inflectional. I will first explain the meaning which these names are
intended to bear, and then proceed to consider the results of more
recent research upon the question of their phylogeny.

In the _Isolating_ forms of language every word stands by itself,
without being capable of inflectional change for purposes of
grammatical construction, and without admitting of much assistance
for such purposes from demonstrative elements, or modifying constants.
Languages of this kind are often called _Monosyllabic_, from the fact
that the isolated words usually occur in the form of single syllables.
They have also been called _Radical_, from the resemblance which
their monosyllabic and isolated words present to the primitive roots
of languages of other types—roots which, as already indicated, have
been unearthed by the labours of the comparative philologist. Thus,
upon the whole, the best idea of an isolating language may be gained
by comparing it with the “nursery-language” of our own children, who
naturally express themselves, when first beginning to speak, by using
monosyllabic and isolated words, which further resemble the languages
in question by not clearly distinguishing between what we understand
as “parts of speech.” For in isolating tongues such variations of
grammatical meaning as the words are capable of conveying are mainly
produced, either by differences of intonation, or by changing the
positions which words occupy in a sentence. Of course these expedients
obtain more or less in languages of both the other types; but in the
isolating group they have been wrought up into a much greater variety
and nicety of usage, so as to become fairly good substitutes for
modifying constants on the one hand, and inflectional change on the
other. Nevertheless, although inflectional change is wholly absent,
modifying constants in the form of auxiliary words are not so. In
Chinese, for example, there are what the native grammarians call
“full words,” and “empty words.” The full words are the monosyllabic
terms, which, when standing by themselves, present meanings of such
vague generality as to include, for instance, _a ball_, _round_,
_to make round_, _in a circle_: that is to say, the full words when
standing alone do not belong to any one part of speech more than to
another. Moreover, one such word may present many totally different
meanings, such as _to be_, _truly_, _he_, _the letter_, _thus_. In
order, therefore, to notify the particular meaning which a full word is
intended to convey, the empty words are used as aids supplementary to
the devices of intonation and syntax. It is probable that all these
empty words were once themselves full words, the meanings of which
gradually became obscured, until they acquired a purely arbitrary use
for the purpose of defining the sense in which other words were to be
understood—just as our word “like,” in its degenerated form of “ly,”
is now employed to give adjectives the force of adverbs; although, of
course, there is the difference that in isolating tongues the empty
or defining words are not fused into the full ones, but themselves
remain isolated. In the opinion of many philologists, however, “the
use of accessory words, in order to impart the required precision to
the principal terms, is the path that leads from monosyllabic to the
agglutinative state.”[149]

This _Agglutinative_, or, as it is sometimes called, _Agglomerative_
state belongs to languages of the second order. Here the words which
serve the purpose of modifying constants, or marks of relationship,
become fusible with the words which they serve to modify or define, so
as to constitute single though polysyllabic compounds, as in the above
example, “_un-cost-li-ness_.” I have already remarked that by long
usage many of these modifying constants have had their own original
meanings as independent words so completely obscured as to baffle the
researches of philologists.

If all our words had been formed on the type of this example
_un-cost-li-ness_, English would have been an agglutinative language.
But, as a matter of fact, English, like the rest of the group to
which it mainly belongs, has adopted the device of inflecting many
of its words (or, rather, has inherited this device from some of its
progenitors), and thus belongs to the third order of languages which
I have mentioned, namely, the _Inflective_. Languages of this type
are also often termed _Transpositive_, because the words now admit
of being shifted about as to their relative positions in a sentence,
without the meaning being thereby affected. That is to say, relations
between words are now marked much less by syntax, and much more
by individual change. In languages of this kind the principle of
agglutination has been so perfected that the original composition
is more or less obscured, and the resulting words therefore admit
of being themselves twisted into a variety of shapes significant of
finer grades of meaning, in the way of declension, conjugation, &c.
Or, to state the case as it has been stated by some philologists,
in agglutinative tongues the welded elements are not sufficiently
welded to admit of flexion: they are too loosely joined together, or
still too independent one of another. But when the union has grown
more intimate, the structure allows of more artistic treatment at the
hands of language-makers: the “amalgamation” of elements having become
complete, the resulting alloy can be manipulated in a variety of ways
without involving its disintegration. Moreover, this principle of
inflection may extend from the component parts to the root itself; not
only suffixes and prefixes, but even the word which these modify, may
undergo inflectional change. So that, upon the whole, the best general
idea of these various types of language-structure may perhaps be given
by the following formulæ, which I take from Hovelacque.[150]

In the isolating type the formula of a word is simply R, and that of
a sentence R+R+R, &c., where R stands for “root.” If we represent by
r those roots whose sense has become obscured so as to pass into the
state of prefixes and suffixes significant only of relationship between
other words, we shall have a formula of agglutination, Rr, Rrr, rR,
rRr, &c. Lastly, the essence of an inflecting language consists in
the power of a root to express, by modification of its own form, its
various relations to other roots. Not that the roots of all words are
necessarily modified; for they often remain as they do in agglutinating
tongues. But they _may_ be modified, and “languages in which relations
may be thus expressed, not only by suffixes and prefixes, but also by
a modification of the form of the roots, are inflectional languages.”
Therefore, if we represent this power of inflectional change on the
part of the root itself by the symbol ^x, the agglutinating formula
Rr may become R^{xr}. Moreover, the modifying elements may also be
inflected, words thus yielding such formulæ as Rr^x, Rrr^x, &c.

Such, then, are the three main groups or orders of language. But
in addition to them we must notice three others, which have been
shown to be clearly separable. These three additional groups are the
Polysynthetic, the Incorporating, and the Analytic.

The _Polysynthetic_ (= _Incapsulating_) order is found among certain
savages, especially on the continent of America, where, according
to Duponceau, more or less distinctive adherence to this type is
to be met with from Greenland to Chili. The peculiarity of such
languages consists in the indefinite composition of words by syncope
and ellipsis. That is to say, sentences are formed by the running
together of compound words of inordinate length, and in the process of
fusion the constituent words are so much abbreviated as often to be
represented by no more than a single intercalated letter. For example,
the Greenland _aulisariartorasuarpok_, “he-hastened-to-go-afishing,”
is made up of _aulisar_, “to fish,” _peartor_, “to be engaged in
anything,” _pinnesuarpok_, “he hastens:” and the Chippeway _totoccabo_,
“wine,” is formed of _toto_, “milk,” with _chominabo_, “a bunch of
grapes.” Thus, polysynthesis consists of fusion with contraction,
some of the component words losing their first, and others their last
syllables. Moreover, composition of this kind further differs from that
which occurs in many other types of language (_e.g._ our adjectival
_never-to-be-forgotten_), in that the constituent parts may never have
attained the rank of independent words, which can be set apart and
employed by themselves.

The _Incorporating_ order is merely a subdivision of the agglutinative,
and represents an earlier stage of it, wherein the speakers had not
yet begun to analyze their sentences, and so still retain in their
sentences subordinate words in cumbersome variety, as, for example,
“House-I-it-built;” “They-have-them-their-books.”

Again, the _Analytic_ order is merely a subdivision of the
inflectional, and represents a later stage of it. “One by one the
grammatical relations implied in an inflectional compound are brought
out into full relief, and provided with special forms in which to be
expressed.” Thus, in English, for example, inflections have largely
given place to the use of “auxiliary” words, whereby most of the
advantages of refined distinction are retained, while the machinery of
expression is considerably simplified.

So that, on the whole, we may classify the Language-kingdom thus:—

 Order I. Isolating.

 Order II. Agglutinative: (Sub-orders, Polysynthetic and Incorporating).

 Order III. Inflectional: (Sub-order, Analytic).

In the opinion of some philologists, however, the Polysynthetic type
deserves to be regarded, not as a sub-order of the Agglutinative,
but as itself independent of all the other three, and therefore
constituting a fourth order. Thus, on the one hand, we have it said
that polysynthetic languages must “simply be placed last in the
ascending order of the agglutinating series;”[151] while, on the other
hand, it is said, “the conception of the sentence that underlies
the polysynthetic dialects is the precise converse of that which
underlies the isolating or the agglutinative types; the several ideas
into which the sentence may be analyzed, instead of being made equal
or independent, are combined, like a piece of mosaic, into a single
whole.”[152]

These two representative quotations may serve to show how accentuated
is the difference of teaching with regard to this particular group
of languages. As a mere matter of classification, of course, the
question would not be of any importance for us; but as the question
of classification involves one of phylogeny, the matter does acquire
considerable interest in relation to our subject.

       *       *       *       *       *

Turning, then, from the classification of language-types to their
phylogeny, no one disputes that what I have called the sub-order
Incorporating is genetically connected with the order Agglutinative;
or that the sub-order Analytic is similarly connected with the order
Inflectional. Indeed, these sub-orders are merely branches of these two
respective trunks. The question before us, therefore, reduces itself
to the relations between the three orders _inter se_, and also between
the polysynthetic type and Order II. I will deal with these two cases
separately.

On the one hand it is argued that the isolating, monosyllabic, or
“nursery” type of speech must be regarded as the most primitive—in
fact, that it presents to actual observation the continued “survival”
of that embryonic or “radical” stage of development out of which all
the subsequent growths of language have arisen. Again, the proved fact
of agglutination is seen to represent a long course of development,
wherein words previously isolated were run together into compounds for
the purpose of securing that higher differentiation of language-growth
which we know as parts of speech. Similarly, the inflectional stage
is taken to have been a further elaboration of the agglutinative, in
the manner already explained; while, lastly, the use of auxiliary
words in analytic tongues is regarded as the final consummation of
language-growth.

The theory thus briefly sketched is still maintained by many
philologists; and, indeed, in some of its parts is not a theory at all,
but a matter of demonstrable fact. Thus, it is manifestly impossible
that the phenomena of agglutination can be presented before there are
elements to agglutinate: these elements, therefore, must have preceded
that process of fusion wherein the “genius” of agglutinated speech
consists. Similarly, of course, agglutination must have preceded the
inflection of already agglutinated words; while the use of auxiliaries
can be proved to have been historically subsequent to inflection.
Nevertheless, other philologists have shown good ground for questioning
our right to regard these facts as justifying so universal a theory
as that the law of language-growth is always to be found in these
particular lines, or that all languages of one type must have passed
through the lower phase, or phases, before reaching that in which they
now appear. The most recent argument on this side of the question is by
Professor Sayce, whom, therefore, I will quote.

“We are apt to assume that inflectional languages are more highly
advanced than agglutinative ones, and agglutinative languages than
isolating ones, and hence that isolation is the lowest stage of the
three, at the top of which stands flection. But what we really mean
when we say that one language is more advanced than another, is that
it is better adapted to express thought, and that the thought to be
expressed is itself better. Now, it is a grave question whether from
this point of view the three classes of language can really be set the
one against the other.”[153]

He then proceeds to argue that isolating languages have an advantage
over all other forms in “the attainment of terseness and vividness;”
that “the agglutinative languages are in advance of the inflectional
in one important point, that, namely, of analyzing the sentence into
its component parts, and distinguishing the relations of grammar one
from another.... In fact, when we examine closely the principle upon
which flection rests, we shall find that it implies an inferior logical
faculty to that implied by agglutination.”[154]

Elsewhere he says, “As for the primeval root-language, we have no
proof that it ever existed, and to confound it with a modern isolating
language is simply erroneous. Equally unproved is the belief that
isolating languages develop into agglutinative, and agglutinative into
inflectional. At all events, the continued existence of isolating
tongues like the Chinese, or of agglutinative like the Magyar and
Turkish, shows that the development is not a necessary one.”[155]

I could quote other passages to the same effect; but the above are
sufficient to show that we must not unreservedly accept the earlier
doctrines previously sketched. There is, indeed, no question about the
fact of language-growth as regards particular languages; the question
here is as to the evolution of language-types one from another. And I
have given prominence to this question in order to make the following
remarks upon it.

When we are told that “the continued existence of isolating tongues
like the Chinese, or of agglutinative tongues like the Magyar and
Turkish, shows that the development is not a necessary one,” we of
course at once perceive the unquestionable truth of the statement.
But the fact is without relevance to the only question in debate. The
continued existence of the Protozoa unquestionably proves that their
development into the Metazoa is not necessary; but this fact raises no
presumption at all against the doctrine that all the Metazoa have been
evolved from the Protozoa.

Similarly, when we are told that “what we really mean when we say
that one language is more advanced than another, is that it is better
adapted to express thought,” we are again being shunted from the
question. The question is whether one type of language-structure
_develops_ into another: not whether, when developed, it is “_more
advanced_” than another in the sense of being “better adapted to
express thought.” This it may or may not be; but in either case the
question of its efficiency as a language has no necessary connection
with the question of its development as a language. For it may very
well be that from the same origin two or more lines of development
may occur in different directions. It is doubtless perfectly true,
as Professor Sayce says, that modern Chinese is a higher product
of evolution than ancient Chinese along the line of isolating
condensation; but this is no proof that the agglutinative languages
did not start from an isolating type, and thereafter proceed on a
different line of development in accordance with their different
“genius,” or method of growth. Naturalists entertain no doubt that
two different types of morphological structure, _b_ and [Greek: b],
are both descended from a common parent form B, even though _b_ has
“advanced” in one line of change and [Greek: b] in another, so that
both are now equally efficient from a morphological point of view. Why,
then, should a philologist dispute genetic relationship in what appears
to be a precisely analogous case, on the sole ground that _b_ is, to
his thinking, no less psychologically efficient a language than [Greek:
b]?

Lastly, as I have before indicated, it appears to me impossible to
dispute that every agglutinative language, in whatever measure it can
be proved to be agglutinative, in that measure is thereby proved to
have been derived from a language less agglutinative, and therefore
more isolating. And, similarly, in whatever measure an inflective
language can be proved to inflect its agglutinated words, in that
measure is it thereby proved to have been derived from a language less
inflective, or a language whose agglutinations had not yet undergone so
much of the inflective modification.

On the other hand, as there is no necessary reason why an isolating
language should develop into an agglutinative, or an agglutinative
into an inflectional, it may very well be that the higher evolution
of isolating tongues has proceeded collaterally with that of
agglutinative, while the higher evolution of agglutinative has
proceeded collaterally with that of inflectional. If this were so, both
the schools of philology which we are considering would be equally
right, and equally wrong: each would represent a different side of the
same truth.

Thus it appears to me that, so far as the purposes of the present
treatise are concerned, we may neglect the question of phylogenesis
as between these three orders of languages. For, so long as it is on
all hands agreed that the principles of evolution are universally
concerned in the genesis of every language, it will make no difference
to my future argument whether these principles have obtained in one
or in more lines of development. There can be no reasonable doubt that
in some greater or less degree the three orders are connected: in
what precise degree this connection obtains is doubtless a question
of high importance to the science of philology: it is of scarcely any
importance to the problems which we shall presently have to consider.

But the issue touching the relation between the polysynthetic and
other types of language is of more importance for us, inasmuch as
it involves the question whether or not we have here to do with the
most primitive type of language. In the opinion of some philologists,
“these polysynthetic languages are an interesting survival of the
early condition of language everywhere, and are but a fresh proof
that America is in truth ‘the new world:’ primitive forms of speech
that have elsewhere perished long ago still survive there, like the
armadillo, to bear record of a bygone past.”[156] On the other hand,
it is with equal certainty affirmed that “polysynthesis is not a
primitive feature, but an expansion, or, if you will, a second phase of
agglutination.”[157]

Of course in dealing with this issue I can only do so as an amateur,
quite destitute of authority in matters pertaining to philology; but
the points on which I am about to speak have reference to principles so
general, that in trying them the lay mind may not be without its uses
in the jury-box. Moreover, philologists themselves are at present so
ill-informed touching the facts of polysynthetic language, that there
is less presumption here than elsewhere in any outsider offering his
opinion upon the matters in dispute.[158] It is however, undesirable
to occupy space with any tedious rehearsal of the facts on which, after
reading the more important literature of the subject, my judgment is
based. For what it is worth, this judgment is as follows.

In the first place, it appears to me that those experts have an
overwhelmingly strong case who argue in favour of the polysynthetic
languages as presenting a highly primitive form of speech. Indeed, so
undifferentiated do I think they prove this type of language-structure
to be, that I agree with them in concluding that it probably brings
us nearer “the origin of speech” than any other type now extant.
Furthermore, looking to the wide contrast between this type and
that which is presented by the isolating tongues, it appears to me
impossible that the one can be genetically connected with the other.
For it appears to me that the experts on the opposite side have no
less completely proved, that the isolating tongues also present
evidence of a highly primitive origin; and, therefore, that whatever
amount of evolution and subsequent degeneration (“phonetic decay”)
the Chinese language, for instance, may be proved to have undergone,
this only goes to show that it has throughout remained true to the
isolating principle—just as the Protozoa, through all their long
history of evolution, have remained true to _their_ “isolating” type,
notwithstanding that some of their branches must long ago have given
origin to the “agglutinated” Metazoa. In other words, it appears to me
that the experts on this side of the question have been able to place
the isolating type of speech on as low a level of development—and,
therefore, presumably on as high a level of antiquity—as experts on
the other side have been able to claim for the polysynthetic.

If I am right in this opinion, it follows that there must have been at
least two points of origin from which all existing languages arose—or
rather, let me say, at least two types of language-formation upon
which the earliest materials of speech were moulded. For even the
strongest advocates of the polysynthetic origin of speech do not
venture to question the highly primitive nature of the monosyllabic
type. Thus, for instance, Professor Sayce is the principal upholder
of the polysynthetic view, and yet he quotes the isolating forms
of Chinese and Taic as furnishing “excellent illustrations of the
early days of speech;”[159] and he adduces them as “examples from
the far East to show us the way in which our words first came into
existence.”[160] But if this is allowed to be so even by the leading
advocate of the polysynthetic view, I cannot conceive the possibility
of the one type having become so completely transformed into the other
as to have left no trace in the isolating type of its polysynthetic
origin. For, in view of the above admissions, we are left to conclude
that the transformation must have taken place soon after the birth
of language in any form—notwithstanding that, as Professor Sayce
elsewhere insists (in the passage already quoted), “the conception of
the sentence which underlies the polysynthetic dialects is the precise
converse of that which underlies the isolating or the agglutinative
type.”

In view of these statements, therefore, by Professor Sayce himself, I
do not think it is necessary for me to go further in justification of
the opinion already expressed—namely, that we must recognize at least
two types of language-formation upon which the earliest materials of
speech were moulded. It is probable enough that both these types of
language-formation were independently originated in many parts of the
earth’s surface at different times; and it is possible that yet other
types may have arisen, which are now either extinct, or fused with
some of the later developments of the two which have survived. But, be
these things as they may, I believe that both the schools of philology
which we are considering have made out their respective cases; and,
therefore, that they both err in so often assuming that these cases
are mutually exclusive.

It will thus be apparent that I am altogether in favour of the
polyphylectic theory of language-development. Even if it were not for
the specially philological considerations just adduced, on grounds of
merely general reasoning it would appear to me much more probable that
so useful a sociological instrument as that of articulate sign-making
should have been evolved from the sign-making of tone and gesture,
wherever the psychological powers of mankind were far enough advanced
to admit of the evolution. And, if this is so, it clearly becomes
probable that any aboriginal races which were geographically separated
would have slowly and independently elaborated their primitive forms
of utterance—supposing, of course, that mankind had become segregated
while still in the speechless state, which, as I will subsequently
explain, seems to me the most probable supposition. And, if this were
the case, it appears to me highly improbable that languages which
originated and developed independently of one another should all have
been under the necessity of starting either on the monosyllabic,
the polysynthetic, or any other type exclusively. That the existing
languages of the earth did originate in more than one centre is now the
almost universal belief of competent authorities.[161] But too many
of these authorities are still bound by what appears to me the wholly
gratuitous and highly improbable assumption, that although various
languages thus originated in different centres, they must all have been
born with an exact family resemblance to one another, so far as type or
“genius” is concerned. But there is no basis for such an assumption,
either in the physiology or the psychology of mankind. On the contrary,
if we look to the nearest analogue of the case, namely, the growing
child, we may find abundant evidence of the fact that the earliest
attempts at articulate utterance may occur on different types, as we
saw so strikingly proved by quotations from Dr. Hale in a previous
chapter.

       *       *       *       *       *

In this connection I would like to conclude the present chapter by
giving prominence to an interesting and ingenious hypothesis, which has
been suggested by Dr. Hale on the basis of the facts just alluded to.

In order that the merits of this suggestion may be appreciated, it
is desirable to remind the reader that the languages now spoken by
the native tribes of the American continent present so many and such
radical differences among themselves, that, with regard to a large
proportion of them, philologists are unable so much as to suggest any
philological classification. Thus, to quote Professor Whitney, “as
regards the material of expression, it is fully confessed that there
is irreconcilable diversity among them. There are a very considerable
number of groups, between whose significant signs exist no more
apparent correspondencies than between those of English, Hungarian,
and Malay; none, namely, which may not be merely fortuitous.”[162]
And, what is most curious, these immense differences may obtain
between neighbouring tribes who are to all appearance ethnologically
identical—as, for instance, the Algonkin, Iroquois, and Dakota groups.
Moreover, this diversity of language-structure in some cases goes so
far as to reach the very roots of language-growth; “the polysynthetic
structure does not belong in the same degree to all American languages:
on the contrary, it seems to be altogether effaced, or originally
wanting, in some.”[163] Nay, even the isolating type of language
has gained a footing, and this in its properly monosyllabic and
uninflective form.

Such being the state of matters on the American continent (and also,
though to a lesser extent, in the Southern parts of the African), Dr.
Hale suggests the following hypothesis by way of explanation. To me
it certainly appears a plausible one, and if it should eventually be
found to furnish a key for unlocking the mysteries of language-growth
in the New World, it would obviously become available as a sufficient
explanation of radical diversities of language elsewhere.

Starting from the facts which I have already quoted from his paper at
the close of my chapter on Articulation, he argues that if children
will thus spontaneously devise a language of their own in a wholly
arbitrary manner, even when surrounded by the spoken language of a
civilized community, much more would children be likely to do this if
they should be accidentally separated from human society, and thus
thrown upon their own resources in an isolated condition. Now, “if,
under such circumstances, disease or the casualties of a hunter’s life
should carry off the parents, the survival of the children would, it is
evident, depend mainly upon the nature of the climate and the ease with
which food could be procured at all seasons of the year. In ancient
Europe, after the present climatical conditions were established, it
is doubtful if a family of children under ten years of age could have
lived through a single winter. We are not, therefore, surprised to find
that no more than four or five linguistic stocks are represented in
Europe, and that all of them, except the Basque, are believed, on good
evidence, to have been of comparatively late introduction. Even the
Basque is traced by some, with much probability, to a source in North
Africa. Of Northern America, east of the Rocky Mountains and north of
the tropics, the same may be said. The climate and the scarcity of food
in winter forbid us to suppose that a brood of orphan children could
have survived, except possibly, by a fortunate chance, in some favoured
spot on the shore of the Mexican Gulf, where shell-fish, berries, and
edible roots are abundant and easy of access.

“But there is one region where Nature seems to offer herself as the
willing nurse and bountiful step-mother of the feeble and unprotected.
Of all countries on the globe, there is probably not one in which a
little flock of very young children would find the means of sustaining
existence more readily than in California. Its wonderful climate, mild
and equable beyond example, is well known. Mr. Cronise, in his volume
on the ‘Natural Wealth of California,’ tells us, that ‘the monthly mean
of the thermometer at San Francisco in December, the coldest month, is
50°; in September, the warmest month, 61°.’ And he adds:—‘Although
the State reaches to the latitude of Plymouth Bay on the north, the
climate, for its whole length, is as mild as that of the regions near
the topics. Half the months are rainless. Snow and ice are almost
strangers, except in the high altitudes. There are fully two hundred
cloudless days in every year. Roses bloom in the open air through
all seasons.’ Not less remarkable than this exquisite climate is the
astonishing variety of food, of kinds which seem to offer themselves
to the tender hands of children. Berries of many sorts—strawberries,
blackberries, currants, raspberries, and salmon-berries—are indigenous
and abundant. Large fruits and edible nuts on low and pendent boughs
may be said, in Milton’s phrase, to ‘hang amiable.’ Mr. Cronise
enumerates, among others, the wild cherry and plum, which ‘grow on
bushes;’ the barberry, or false grape (_Berberis herbosa_), a ‘low
shrub,’ which bears edible fruit; and the Californian horse-chestnut
(_Æsculus Californica_), ‘a low, spreading tree or shrub, seldom
exceeding fifteen feet high,’ which ‘bears abundant fruit much used
by the Indians.’ Then there are nutritious roots of various kinds,
maturing at different seasons. Fish swarm in the rivers, and are taken
by the simplest means. In the spring, Mr. Powers informs us, the
whitefish ‘crowd the creeks in such vast numbers that the Indians,
by simply throwing in a little brushwood to impede their motion,
can literally scoop them out.’ Shell-fish and grubs abound, and are
greedily eaten by the natives. Earthworms, which are found everywhere
and at all seasons, are a favourite article of diet. As to clothing,
we are told by the authority just cited that ‘on the plains all adult
males and all children up to ten or twelve went perfectly naked, while
the women wore only a narrow strip of deer-skin around the waist.’ Need
we wonder that, in such a mild and fruitful region, a great number
of separate tribes were found, speaking languages which a careful
investigation has classed in nineteen distinct linguistic stocks?

“The climate of the Oregon coast region, though colder than that of
California, is still far milder and more equable than that of the
same latitude in the east; and the abundance of edible fruits, roots,
river-fish, and other food of easy attainment, is very great. A family
of young children, if one of them were old enough to take care of the
rest, could easily be reared to maturity in a sheltered nook of this
genial and fruitful land. We are not, therefore, surprised to find that
the number of linguistic stocks in this narrow district, though less
than in California, is more than twice as large as in the whole of
Europe, and that the greater portion of these stocks are clustered near
the Californian boundary....

“Some reminiscences of the parental speech would probably remain with
the older children, and be revived and strengthened as their faculties
gained force. Thus we may account for the fact, which has perplexed
all inquirers, that certain unexpected and sporadic resemblances,
both in grammar and in vocabulary, which can hardly be deemed purely
accidental, sometimes crop up between the most dissimilar languages....

“A glance at other linguistic provinces will show how aptly this
explanation of the origin of language-stocks everywhere applies.
Tropical Brazil is a region which combines perpetual summer with a
profusion of edible fruits and other varieties of food, not less
abundant than in California. Here, if anywhere, there should be a great
number of totally distinct languages. We learn on the best authority,
that of Baron J. J. von Tschudi, in the Introduction to his recent
work on the Khetshua Language, that this is the fact. He says:—‘I
possess a collection made by the well-known naturalist, J. Natterer,
during his residence of many years in Brazil, of more than a hundred
languages, lexically completely distinct, from the interior of Brazil.’
And he adds:—‘The number of so-called isolated languages—that is, of
such as, according to our present information, show no relationship
to any other, and which therefore form distinct stocks of greater
or less extent—is in South America very large, and must, on an
approximate estimate, amount to many hundreds. It will perhaps be
possible hereafter to include many of them in larger families, but
there must still remain a considerable number for which this will not
be possible.’”

I have quoted this hypothesis, as previously remarked, because it
appears to me philologically interesting; but whatever may be thought
of it by professional authorities, the evidence which the American
continent furnishes of a polygenetic and polytypic origin of the native
languages remains the same. And if there is good reason for concluding
in favour of polygenetic origins of different types as regards the
languages on that continent, of course the probability arises that
radical differences of structure among languages of the Old World
admit of being explained by their having been derived from similarly
independent sources.[164]




CHAPTER XIII.

ROOTS OF LANGUAGE.


In the last chapter my treatment of the classification and phylogeny
of languages may have led the general reader to feel that philologists
display extraordinary differences of opinion with regard to certain
first principles of their science. I may, therefore, begin the present
chapter by reminding such a reader that I have hitherto been concerned
more with the differences of opinion than with the agreements. If one
takes a general view of the progress of philological science since
philology—almost in our own generation—first became a science, I
think he must feel much more impressed by the amount of certainty
which has been attained than by the amount of uncertainty which still
remains. And the uncertainty which does remain is due rather to a
backwardness of study than to differences of interpretation. When more
is known about the structure and mutual relations of the polysynthetic
tongues, it is probable that a better agreement will be arrived at
touching the relation of their common type to that of isolating
tongues on the one hand, and agglutinating on the other. But, be this
as it may, even as matters stand at present, I think we have more
reason to be surprised at the certainty which already attaches to the
principles of philology, than at the uncertainty which occasionally
arises in their applications to the comparatively unstudied branches of
linguistic growth.

Furthermore, important as these still unsettled questions are from a
purely philological point of view, they are not of any great moment
from that of the evolutionist, as I have already observed. For, so
long as it is universally agreed that all the language-groups have
been products of a gradual development, it is, comparatively speaking,
immaterial whether the groups all stand to one another in a relation of
serial descent, or whether some of them stand to others in a relation
of collateral descent. That is to say, the evolutionist is under no
obligation to espouse either the monotypic or the polytypic theory of
the origin of language. Therefore, it will make no material difference
to the following discussion whether the reader feels disposed to
follow the doctrine, that all languages must have originated in such
monosyllabic isolations as we now meet with in a radical form of speech
like the Chinese; that they all originated in such polysynthetic
incapsulations as we now find in the numberless dialects of the
American Indians; or, lastly, and as I myself think much more probably,
that both these, and possibly other types of language-structure, are
all equally primitive. Be these things as they may, my discussion
will not be overshadowed by their uncertainty. For this uncertainty
has reference only to the _origin_ of the existing language-types as
independent or genetically allied: it in no way affects the certainty
of their subsequent _evolution_. Much as philologists may still
differ upon the mutual relations of these several language-types,
they all agree that “von der ersten Entstehung der Sprachwurzeln an
bis zur Bildung der volkommenen Flexionssprachen, wie des Sanskrit,
Griechischen, oder Deutschen, ist Alles in der Entwicklung der Sprache
verständlich.... Sobald nur die Wurzeln als die fertigen Bausteine der
Sprache einmal da sind, lässt sich Schritt für Schritt das Wachsthum
des Sprachgebäudes verfolgen.”[165]

Therefore, having now said all that seems necessary to say on the
question of language-types, I will pass on to consider the information
that we possess on the subject of language-roots.

First, let us consider the number of roots out of which languages
are developed—or, rather, let me say, the number of elementary
constituents into which the researches of philologists have been able
to reduce those languages which have been most closely studied. Of
course the probability—nay, the certainty—is that the actual number
of roots must in all cases be considerably less than philologists are
now able to prove.

Chinese is composed of about five hundred separate words, each
being a monosyllable. In actual use, these five hundred root-words
are multiplied to over fifteen hundred by significant variety of
intonation; but the entire structure of this still living language is
made up of five hundred monosyllabic words. In the opinion of most
philologists we have here a survival of the root stage of language; but
in the opinion of some we have the remnants of erosion, or “phonetic
decay.”[166] This difference of opinion, however, is not a matter of
importance to us; and therefore I will not discuss it, further than to
say that on account of it I will not hereafter draw upon the Chinese
language for illustrations of “radical” utterance, except in so far as
philologists of all schools would allow as legitimate.[167]

Hebrew has been reduced to about the same number of roots as
Chinese—Renan stating it in round numbers at five hundred.[168] But
without doubt this number would admit of being considerably reduced, if
inquiries were sufficiently extended to the whole Semitic family.

According to Professor Skeat, English is entirely made up of 461 Aryan
roots, in combination with about twenty modifying constants.[169] The
remote progenitor, Sanskrit, has been estimated to present as many as
850 roots, or, according to Benfey, just about twice that number.[170]
On the other hand, Max Müller, as a result of more recent researches,
professes to have reduced the total number of Sanskrit roots to
121.[171]

It is needless to give further instances. For these are enough to show
that, even if we were to regard the analytic powers of comparative
philology as adequate to resolve all the compounds of a language
into its primitive elements the estimate of Pott would probably be
high above the mark, when he states that on an average the roots of
a language may be taken at a thousand.[172] Seeing that Chinese only
contains in its whole vocabulary half that number of words, and that
both Hebrew and English have similarly yielded each about five hundred
radicals in the crucible of more modern research, I think we may safely
reduce the general estimate of Pott by one-half, and probably would
be nearer the truth if we were to do so by three-quarters, or more.
At all events, we may be satisfied that the total number of radicals
sufficient to feed the most luxuriant of languages is expressible in
three figures; and this, as we shall presently see, is enough for all
the purposes of my subsequent discussion.

Passing on now from the question of number to that of character, we
have first to meet the question—What _are_ these roots? Are they the
actually primitive words of pre-historic languages, or are they what
Max Müller has aptly termed “phonetic types”? Here again we encounter
a difference of opinion among philologists. Thus, for instance,
Professor Whitney tells us that the Indo-European languages are all
descended from an original monosyllabic tongue, and, therefore, that
“our ancestors talked with one another in simple syllables, indicative
of ideas of prime importance, but wanting all designation of their
relations.”[173] On the other hand, it is objected to this view that
“such a language is a sheer impossibility;”[174] that “there could be
no hope of any mutual understanding” with a language restricted to
such isolated and general terms, &c.[175] On this side of the question
it is represented that “roots are the phonetic and significant types
discovered by the analysis of the comparative philologist as common to
a group of allied words;”[176] that “a root is the _core_ of a group
of allied words,”[177] “the naked kernel of a family of words.”[178]
Or, to adopt a simile previously used in another connection, we may
say that a root as now presented by the philologist is a composite
photograph (or _phonogram_) of a number of words, all belonging to the
same pre-historic language, and all closely allied in meaning.

The difference of authoritative teaching thus exhibited is not a matter
of much importance for us. Nor, indeed, as we shall subsequently see,
is it a difference so great as may at first sight appear. For even
the phonetic-type theory does not doubt that all the aboriginal and
unknown words, out of the composition of which a root is now extracted,
must have been genetically allied with one another, and exhibited the
closeness of their kinship by a close similarity of sound. Therefore,
it does not make any practical difference whether we regard a root
as itself a primitive word, which was used in some such way as the
Chinese now use their monosyllabic terms; or whether we regard it as a
generalized expression of a group of cognate words, all closely allied
as to meaning. In fact, even so strong an adherent of the phonetic-type
theory as Professor Max Müller very clearly states this, where he says
that, although “the mere root, _quâ_ root, may be denied the dignity of
a word, as soon as a root is used for predication it becomes a word,
whether outwardly it is changed or not.”[179]

Seeing, then, that this difference of opinion among philologists is
not one of great importance for us, I will henceforth disregard it.
And, as it will be conducive to brevity, if not also to clearness, I
will speak of roots as archaic words, although by so doing I shall not
intend to assume that they are more than phonetic types, or the nearest
approach we can make to the words out of which they were generated.

We may next consider the kind of meanings which roots convey.
Antecedently we might form various anticipations on this head, such as
that they should be imitative of natural sounds, expressive of concrete
ideas, and so forth. As a matter of fact, we find that they are not
expressive of natural sounds; but, as far as we have now any means of
judging, quite arbitrary. Moreover, they are not expressive of concrete
or particular ideas; but always of abstract or general. Here, then,
to begin with, we have two facts of apparently great importance. And
they are both facts which, at first sight, seem to countenance the
view that, in its last resort, comparative philology fails to testify
to the natural origin of speech. But we must look into the matter
more closely, and, in order to do this most fairly, I will quote from
Professor Max Müller the 121 roots into which he analyzes the Sanskrit
language. This is the language which has been most carefully studied in
the present connection, and of all its students Professor Max Müller is
least open to any suspicion of inclining to the side of “Darwinism.”
The following is a list of what he calls “the 121 original concepts.”

  1. Dig.

  2. Plat, weave, sew, bind.

  3. Crush, pound, destroy, waste, rub, smooth.

  4. Sharpen.

  5. Smear, colour, knead, harden.

  6. Scratch.

  7. Bite, eat.

  8. Divide, share, eat.

  9. Cut.

  10. Gather, observe.

  11. Stretch, spread.

  12. Mix.

  13. Scatter, strew.

  14. Sprinkle, drip, wet.

  15a. Shake, tremble, quiver, flicker.

  15b. Shake, mentally, be angry, abashed, fearfully, etc.

  16. Throw down, fall.

  17. Fall to pieces.

  18. Shoot, throw at.

  19. Pierce, split.

  20. Join, fight, check.

  21. Tear.

  22. Break, smash.

  23. Measure.

  24. Blow.

  25. Kindle.

  26. Milk, yield.

  27. Pour, flow, rush.

  28. Separate, free, leave, lack.

  29. Glean.

  30. Choose.

  31. Cook, roast, boil.

  32. Clean.

  33. Wash.

  34. Bend, bow.

  35. Turn, roll.

  36. Press, fix.

  37. Squeeze.

  38. Drive, thrust.

  39. Push, stir, live.

  40. Burst, gush, laugh, beam.

  41. Dress.

  42. Adorn.

  43. Strip, remove.

  44. Steal.

  45. Check.

  46. Fill, thrive, swell, grow
  strong.

  47. Cross.

  48. Sweeten.

  49. Shorten.

  50. Thin, suffer.

  51. Fat, stick, love.

  52. Lick.

  53. Suck, nourish.

  54. Drink, swell.

  55. Swallow, sip.

  56. Vomit.

  57. Chew, eat.

  58. Open, extend.

  59. Reach, strive, rule, have.

  60. Conquer, take by violence, struggle.

  61. Perform, succeed.

  62. Attack, hurt.

  63. Hide, drive.

  64. Cover, embrace.

  65. Bear, carry.

  66. Can, be strong.

  67. Show.

  68. Touch.

  69. Strike.

  70. Ask.

  71. Watch, observe.

  72. Lead.

  73. Set.

  74. Hold, wield.

  75. Give, yield.

  76. Cough.

  77. Thirsty, dry.

  78. Hunger.

  79. Yawn.

  80. Spue.

  81. Fly.

  82. Sleep.

  83. Bristle, dare.

  84. Be angry, harsh.

  85. Breathe.

  86. Speak.

  87. Seek.

  88. Hear.

  89. Smell, sniff.

  90. Sweat.

  91. Seethe, boil.

  92. Dance.

  93. Leap.

  94. Creep.

  95. Stumble.

  96. Stick.

  97. Burn.

  98. Dwell.

  99. Stand.

  100. Sink, lie, fail.

  101. Swing.

  102. Hang down, lean.

  103. Rise up, grow.

  104. Sit.

  105. Toil.

  106. Weary, waste, slacken.

  107. Rejoice, please.

  108. Desire, love.

  109. Wake.

  110. Fear.

  111. Cool, refresh.

  112. Stink.

  113. Hate.

  114. Know.

  115. Think.

  116. Shine.

  117. Run.

  118. Move, go.

  119a. Noise, inarticulate.

  119b. Noise, musical.

  120. Do.

  121. Be.

“These 121 concepts constitute the stock-in-trade with which I maintain
that every thought that has ever passed through the mind of India, so
far as it is known to us in its literature, has been expressed. It
would have been easy to reduce that number still further, for there are
several among them which could be ranged together under more general
concepts. But I leave this further reduction to others, being satisfied
as a first attempt with having shown how small a number of seeds may
produce, and has produced, the enormous intellectual vegetation that
has covered the soil of India from the most distant antiquity to the
present day.”[180]

Now, the first thing which strikes one on reading this list is, that it
unquestionably justifies the inference of its compiler, namely, “if the
Science of Language has proved anything, it has proved that every term
which is applied to a particular idea or object (unless it be a proper
name) is already a general term.” But the next thing which immediately
strikes one is that the list, surprisingly short as it is, nevertheless
is much too long to admit of being interpreted as, in any intelligible
sense of the words, an inventory of “original concepts”—unless by
“original” we are to understand the ultimate results of philological
analysis. That all these concepts are not “original” in the sense of
representing the ideation of really primitive man, is abundantly proved
by two facts.

The first is that fully a third of the whole number might be dispensed
with, and yet leave no important blank in the already limited resources
of the list for the purposes either of communication or reflection. To
yawn, to spew, to vomit, to sweat, and so on, are not forms of activity
of any such vital importance to the needs of a primitive community,
as to demand priority of naming by any aboriginal framers of language.
Moreover, as Professor Max Müller himself elsewhere observes, “even
these 121 concepts might be reduced to a much smaller number, if we
cared to do so. Any one who examines them carefully, will see how easy
it would have been to express to dig by to cut or to strike; to bite
by to cut or to crush; to milk by to squeeze; to glean by to gather;
to steal by to lift.... If we see how many special purposes can be
served by one root, as _I_, to go, or _Pas_, to fasten, the idea that a
dozen of roots might have been made to supply the whole wealth of our
dictionary, appears in itself by no means so ridiculous as is often
supposed.”[181]

Again, in the second place, a large proportional number of the words
have reference to a grade of culture already far in advance of that
which has been attained by most existing savages. “Many concepts, such
as to cook, to roast, to measure, to dress, to adorn, belong clearly
to a later phase of civilized life.”[182] It might have been suitably
added that such “concepts” as to dig, to plant, to milk, &c., betoken
a condition of _pastoral_ life, which, as we know from abundant
evidence, is representative of a comparatively high level of social
evolution.[183] But if “many” of these concepts are thus unmistakably
referable to semi-civilized as distinguished from savage life, what
guarantee can we have that the remainder are “original”? Obviously we
can have no such guarantee; but, on the contrary, find the very best,
because _intrinsic_ evidence, that they belong to a more or less high
level of culture, far removed from that of primitive man. In other
words, we must conclude that these 121 concepts are “original” only in
the sense that they do not now admit of further analysis at the hands
of comparative philologists: they are not original in the sense of
bringing us within any measurable distance of the first beginnings of
articulate speech.[184]

Nevertheless, they are of the utmost value and significance, in that
they bring us down to a period of presumably restricted ideation,
as compared with the enormous development since attained by various
branches of this Indo-European stock—so far, at least, as the growth
of language can be taken as a fair expression of such development.
They are likewise of the highest importance as showing in how
presumably short a period of time (comparatively speaking) so immense
and divergent a growth may proceed from such a simple and germ-like
condition of thought.[185] Lastly, they serve to show in a most
striking manner that the ideas represented, although all of a general
character, are nevertheless of the lowest degree of generality.
Scarcely any of them present us with evidence of reflective thought,
as distinguished from the naming of objects of sense-perception, or
of the simplest forms of activity which are immediately cognizable
as such.[186] In other words, few of these “original concepts”
rise much higher in the scale of ideation than the level to which
I have previously assigned what I have called “named recepts” or
“pre-concepts.” A dumb animal, or an infant, presents a full receptual
appreciation of the majority of actions which the catalogue includes;
and, therefore, so that a society of human beings can speak at all
(_i.e._ presents the power of naming their recepts), it is difficult
to see how they could have avoided a denotation of the more important
recepts which are here concerned.

Another most interesting feature of a general kind which the list
presents is, that it is composed exclusively of verbs.[187] This
peculiarity of the ultimate known roots of all languages, which shows
them to have been expressive of actions and states as distinguished
from objects and qualities, is a peculiarity on which Professor Max
Müller lays much stress. But the inference which he draws from the
fact is clearly not justifiable. This inference is that, as every
root expresses “the consciousness of repeated acts, such as scraping,
digging, striking,” &c., the naming of actions, as distinguished from
objects, “must be considered as the first step in the formation of
concepts.” Now, in drawing this inference—and, indeed, throughout
all his works as far as I remember—Professor Max Müller has entirely
overlooked two most important considerations. First, as already
observed, that the roots in question are _demonstrably_ very far
from having been the original material of language as first coined
by primitive man; and, next, that whatever this original material
may have been, from the first there must have been a struggle for
existence among the really primitive roots—only those surviving which
were most fitted to survive as roots, _i.e._ as the parent stems of
subsequent word-formations. Now, it appears to me obvious enough that
archaic—though not necessarily aboriginal—words which were expressive
of actions, would have stood a better chance of surviving as roots than
those which may have been expressive of objects; first because they
were likely to have been more frequently employed, and next because
many of them must have lent themselves more readily to metaphorical
extension—_especially under a system of animistic thought_.[188]
And, if these things were so, there is nothing remarkable in words
significant of actions having alone survived as roots.[189]

The consideration that it is only those words which were successful
in the struggle for existence that can have become the progenitors
of subsequent language—and therefore the only words that have been
handed down to us as roots—has a still more important bearing upon
another of Professor Max Müller’s generalizations. From the fact
that all his 121 Sanskrit roots are expressive of “general” ideas
(by which term he of course includes what I call generic ideas), he
concludes that from its very earliest origin speech must have been
thus expressive of general ideas; or, in other words, that human
language could not have begun by the naming of particulars: from
the first it must have been concerned with the naming of “notions.”
Now, of course, if any vestige of real evidence could be adduced
to show that this “must have been” the case, most of the foregoing
chapters of the present work would not have been written. For the
whole object of these chapters has been to show, that on psychological
grounds it is abundantly intelligible how the conceptual stage of
ideation may have been gradually evolved from the receptual—the
power of forming general, or truly conceptual ideas, from the power
of forming particular and generic ideas. But if it could be shown—or
even rendered in any degree presumable—that this distinctly human
power of forming truly general ideas arose _de novo_ with the first
birth of articulate speech, assuredly my whole analysis would be
destroyed: the human mind would be shown to present a quality
different in origin—and, therefore, in kind—from all the lower
orders of intelligence: the law of continuity would be interrupted
at the terminal phase: an impassable gulf would be fixed between the
brute and the man. As a matter of fact, however, there is not only no
vestige of any such proof or even presumption; but, as we shall see in
our two following chapters, there is uniform and overwhelming proof
of precisely the opposite doctrine—proof, indeed, so uniform and
overwhelming that it has long ago induced all other philologists to
accept this opposite doctrine as one of the axioms of their science.
Leaving, however, this proof to be adduced in its proper place, I have
now merely to point out the futility of the evidence on which Professor
Max Müller relies.

This evidence consists merely in fact that the “121 original
concepts,” which are embodied in the roots of Aryan speech, are
expressive of “general ideas.” Now, this argument might be worth
considering if there were the smallest reason to suppose that in these
roots of Aryan speech we possess the aboriginal elements of language as
first spoken by man. But as we well know that this is immeasurably far
from being the case, the whole argument collapses. The mere fact that
many words which have survived as roots are words expressive of general
ideas, is no more than we might have antecedently expected. Remembering
that it is a favourable condition to a word surviving as a root that it
should prove itself a prolific parent of other words, obviously it is
those words which were expressive of ideas presenting some degree of
generality that would have had the best chance of thus coming down to
us, even from the comparatively high level of culture which, as we have
seen, is testified to by “the 121 original concepts.” Of course, as I
have already said, the case would have been different if any one were
free to suppose, even as a merely logical possibility, that this level
of culture represented that of primitive man when he first began to
employ articulate speech. But any such supposition is beyond the range
of rational discussion. The 121 concepts themselves yield overwhelming
evidence of belonging to a time _immeasurably remote_ from that of any
speechless progenitor of _Homo sapiens_; and in the enormous interval
(whatever it may have been) many successive generations of words must
_certainly_ have flourished and died.[190]

These remarks are directed to the comparatively few instances of
general ideas which, as a matter of fact, the list of “121 concepts”
presents. As already observed, the great majority of these “concepts”
exhibit no higher degree of “generality” than belongs to what I have
called a “pre-concept,” _i.e._ a “named recept.” But precisely the
same considerations apply to both. For, even supposing that a named
recept was originally a word used only to designate a “particular” as
distinguished from a “generic” idea, obviously it would have stood but
a poor chance of surviving as a root unless it had first undergone a
sufficient degree of extension to have become what I call receptually
connotative. A proper name, for instance, could not, as such, become
a root. Not until it had become extended to other persons or things
of a like class could it have secured a chance of surviving as a
root in the struggle for existence. As a matter of fact, I think it
most probable—not only from general considerations, but also from a
study of the spontaneous names first coined in “baby-language,”—that
aboriginal speech was concerned simultaneously with the naming both of
particular and of generic ideas—_i.e._ of individual percepts and of
recepts. It will be remembered that in Chapter III., while treating of
the Logic of Recepts, I dealt at some length with this subject. Here,
therefore, it will be sufficient to quote the conclusion to which my
analysis led.

“A generic idea is generic because the particular ideas of which
it is composed present such obvious points of resemblance that
they spontaneously fuse together in consciousness; but a general
idea is general for precisely the opposite reason—namely, because
the points of resemblance which it has seized are obscured from
immediate perception, and therefore could never have fused together in
consciousness but for the aid of intentional abstraction, or of the
power of a mind knowingly to deal with its own ideas as ideas. In other
words, the kind of classification with which recepts are concerned is
that which lies nearest to the kind of classification with which all
processes of so called perceptual inference depend—such as mistaking a
bowl for a sphere. But the kind of classification with which concepts
are concerned is that which lies furthest from this purely automatic
grouping of perceptions. Classification there doubtless is in both
cases; but in the one order it is due to the closeness of resemblances
in an act of perception, while in the other it is due to their
remoteness.”[191]

Of course it goes without saying that this “closeness of resemblances
in an act of perception” may be due either to similarities of
sense-perceptions themselves (as when the colour of a ruby is seen
to resemble that of “pigeon’s blood”), or to frequency of their
associations in experience (as when a sea-bird groups together in one
recept the sundry sensations which go to constitute its perception of
water, with its generic classification of water as a medium in which
it is safe to dive). Now, if we remember these things, can we possibly
wonder that the palæontology of speech should prove early roots to
have been chiefly expressive of “generic” as distinguished from
“general” ideas on the one hand, or “particular” ideas on the other?
By failing to observe this real distinction between classification
as receptual and conceptual—_i.e._ as given immediately in the act
of perception itself, or as elaborated of set purpose through the
agency of introspective thought, Professor Max Müller founds his whole
argument on another and an unreal distinction: he everywhere regards
the bestowing of a name as in itself a sufficient proof of conceptual
thought, and therefore constitutes the faculty of denotation,
equally with that of denomination, the distinctive criterion of a
self-conscious mind. But, as we have now so repeatedly seen, such
is certainly not the case. Actions and processes so habitual, or so
immediately apparent to perception, as those with which the great
majority of these “121 concepts” are concerned, do not betoken any
order of ideation higher than the pre-conceptual, in virtue of which
a young child is able to give expression to its higher receptual life
prior to the advent of self-consciousness. Or, as Geiger tersely
says:—“In enzelnen Fällen ist die Entstehung von Gattungsbegriffe aus
Mangel an Unterscheidung gleichwohl kaum zu bezweifeln.”[192]

Again, if we look to the still closer analogy furnished by savages, we
meet with a still further corroboration of this view. For instance,
Professor Sayce remarks that in “all savage and barbarous dialects,
while individual objects of sense have a superabundance of names,
general terms are correspondingly rare.” And he gives a number of
remarkable illustrations.[193]

In view of these considerations, my only wonder is that these 120
root-words do not present _better_ evidence of conceptual thought. I
have already given my reasons for refusing to suppose that we have here
to do with the “original” framers of spoken language; and looking to
the comparatively high level of culture which the people in question
must have reached, it seems remarkable that the root-words of their
language should only in so few instances have risen above the level
of pre-conceptual utterance.[194] This, however, only shows how
comparatively small a part self-conscious reflection need play in the
practical life of uncultured man: it does not show that the people
in question were remarkably deficient in this distinctively human
faculty. Archdeacon Farrar tells us that he has observed the whole
conversational vocabulary of certain English labourers not to exceed
a hundred words, and probably further observation would have shown
that the great majority of these were employed without conceptual
significance. Therefore, if these labourers had had to coin their own
words, it is probable that, without exception, their language would
have been destitute of any terms betokening more than a pre-conceptual
order of ideation. Nevertheless, these men must have been capable,
in however undeveloped a degree, of truly conceptual ideation: and
this proves how unsafe it would be to argue from the absence of
distinctively conceptual terms to the poverty of conceptual faculty
among any people whose root-words may have come down to us—although,
no doubt, in such a case we appear to be getting within a comparatively
short distance of the origin of this faculty.

The point, however, now is that really aboriginal, and therefore
purely denotative names, must certainly have been “generic” as well as
“particular”: they must have been the names of recepts as well as of
percepts, of actions as well as of objects and qualities. Moreover, it
is equally certain that among this aboriginal assemblage of denotative
names as particular and generic, only those belonging to the latter
class could have stood much chance of surviving as roots. In other
words, no aboriginal name could have survived as a root until it had
acquired some greater or less degree of receptual and, therefore, of
connotative value. Hence the fact that the ultimate result of the
philological analysis of any language is that of reducing the language
to a certain small number of roots, and the fact that all these roots
are expressive of general and generic ideas,—these facts in themselves
yield no support whatever to the doctrine, either that these roots
were themselves the aboriginal elements of language, or, _a fortiori_,
that the aboriginal elements of language were expressive of general
ideas.[195]

       *       *       *       *       *

And this conclusion involves another of scarcely less importance. A
great deal of discussion has been expended over the question as to
whether, or how far, aboriginal language was indebted to the principle
of onomatopœia, or the imitation by articulate names of sounds
obviously distinctive of the objects or actions named. Of course, on
evolutionary principles we should be strongly inclined to suppose that
aboriginal language must have been largely assisted in its formation
by such intentional imitation of natural sounds, seeing that of all
forms of vocal expression they admit of most readily conveying an idea
of the object or action named. And the same applies to the so-called
interjectional element in word-formation, or the utilization as names
of sounds which are naturally expressive of states of human feeling.
On the other hand, contempt has been poured upon this theory as an
adequate explanation of the first beginnings of articulate speech,
on the ground that it is not supported either by history[196] or by
the results of philogenetic inquiry.[197] It is, however, forgotten
by those who argue on this side that names of onomatopoetic origin
must always be, in the first instance, particular; that so long as
they remain particular (as, for example, is the case with our word
“cuckoo”), they cannot have much chance of surviving as roots; that
in proportion as they increase their chances of survival as roots by
becoming more general, they must do so by becoming more conventional;
and, therefore, that the vast majority of roots, even if aboriginally
they were of onomatopoetic origin, must necessarily have had that
origin obscured.

In order to illustrate each and all of these general considerations,
let us turn to the example of our own “baby-language.” The fact that
such language presents so large an element of onomatopœia in itself
furnishes a strong presumption that what is now seen to constitute so
important a principle in the infancy of the individual (notwithstanding
the hereditary tendency to speak), must have constituted at least as
important a principle in the infancy of the race. But the point now is,
that if we mark the connotative extension of any such nursery word,
we may find that just in proportion as it becomes general does its
onomatopoetic origin become obscure. For instance, the late Mr. Darwin
gave me the following particulars with regard to a grandchild of his
own, who was then living in his house. I quote the account from notes
taken at the time.

“The child, who was just beginning to speak, called a duck ‘quack’;
and, by special association, it also called water ‘quack.’ By an
appreciation of the resemblance of qualities, it next extended the term
‘quack’ to denote all birds and insects on the one hand, and all fluid
substances on the other. Lastly, by a still more delicate appreciation
of resemblance, the child eventually called all coins ‘quack,’ because
on the back of a French sou it had once seen the representation of an
eagle. Hence, to the child, the sign ‘quack,’ from having originally
had a very specialized meaning, became more and more extended in
its signification, until it now serves to designate such apparently
different objects as ‘fly,’ ‘wine,’ and ‘coin.’”

Now, if any such process of extending or generalizing aboriginally
onomatopoetic terms were to have taken place among the primitive
framers of human speech, how hopeless would be the task of the
philologist who should now attempt to find the onomatopoetic root!
Yet, as above observed, not only may we be perfectly certain that
such extensions of aboriginal onomatopoetic terms must have taken
place, if any such terms were ever in existence at all (and this
cannot be doubted), but also that it must have been almost a necessary
condition to the survival of an onomatopoetic term as a root that
such an extension of its meaning should have taken place. In other
words, we can see very good reason to conclude that, as a rule, only
those instances of primitive onomatopœia can have survived as roots,
which must long ago have had their onomatopoetic origin hopelessly
obscured. So that nowhere so much as in this case should we be prepared
to entertain the general principle of philological research, that, as
Goethe graphically states it, the original meanings of words become
gradually worn out, like the image and superscription of a coin.[198]

In view of such considerations, my only wonder is that this origin
admits of being traced so often as it does, even as far back as the
comparatively recent times when a pastoral people coined the terms
which afterwards constituted the roots of Sanskrit. _Kas_, to cough;
_kshu_, to sneeze; _proth_, to snort; _ma_, to bleat, and not a
few others, are conceded, even by Professor Max Müller, to be of
obviously imitative origin. In the present connection, however, it
is of interest to notice how this authority deals with such cases. He
says:—“Not one of them is of any importance in helping us to account
for real words in Sanskrit. Most of them have had no offspring at all,
others have had a few descendants, mostly sterile. Their history shows
clearly how far the influence of onomatopœia may go, and if once we
know its legitimate sphere, we shall be less likely to wish to extend
it beyond its proper limits.”[199]

Now, under our present point of view we can see a very good reason
why this element of sterility should have attached to these roots of
Sanskrit whose onomatopoetic origin still admits of being clearly
traced: it is just because they failed to be extended that their
imitative source continues to be apparent.[200] But suppose, for the
sake of illustration, that any one of them had been extended, and what
would have happened? If _ma_, to bleat, had been metaphorically applied
to the crying of a child, and had then become more and more habitually
used in this new signification, while the original meaning became
more and more obsolete, it might have taken the place of any such
root as _bhi_, to fear; _ish_, to love, &c.; and in all the progeny
of words which in this its conventional use it might subsequently
have generated, no trace of imitative origin could now have been
met with—any more than such an origin can be detected in the sound
“quack,” as used by the above-mentioned child to designate a shilling.

Several other considerations to the same general effect might be
adduced. But, to mention only some of the more important, Steinthal
points out that imitative utterance differs widely even among
different races of existing men, so that the onomatopoetic words
of one race do not convey any imitative suggestion to the minds of
another.[201] Similarly, Professor Sayce insists, “it is not necessary
that the imitation of natural sounds should be an exact one; indeed,
that it never can be: all that is wanted is that the imitation
should be recognizable by those addressed. The same natural sound,
consequently, may strike the ear of different persons very differently,
and so be represented in articulate speech in a strangely varying
manner.”[202] Another very good illustration of the same point is to
be found in the names for a grasshopper in different languages. After
giving a number, Archdeacon Farrar remarks that obviously they are “all
imitative: yet how immensely varied by the fantasies of imitation! How
is this to be explained? Simply by the fact to which it is so often
necessary to recur, that words are not mere imitations, but subjective
echoes and reproductions—repercussions which are modified both
organically and ideally—which have moreover been immensely blurred and
disintegrated by the lapse of ages.”[203]

But perhaps the best illustration that has been given of this point is
in the different words which obtain in different languages as names for
Thunder. Two independent treatises have been written on the subject,
one by Grimm,[204] and the other by Pott.[205] While in nearly all the
languages the principle of imitation is more or less clearly apparent,
the greatest diversities occur among the resulting sounds.[206] In this
connection, also, I may adduce yet one further consideration. In his
_Introduction to the Science of Language_, Professor Sayce argues on
several grounds that, when articulation first began, the articulate
sounds were probably in large part dependent for their meaning on the
gestures with which they were accompanied. Consequently, aboriginal
root-words, even supposing that any such had come down to us, and that
their origin were imitative, inasmuch as their imitative value may thus
have in large part depended on appropriately accompanying gestures,
their imitative source would long ago have become obscured.

In view of all these considerations, therefore, I cannot deem the
merely negative evidence against the onomatopoetic origin of articulate
sounds as of any value at all. Even if we had any reason to suppose
that philological analysis were in possession of the really aboriginal
commencements of spoken language, we should still be unable reasonably
to conclude against their imitative origin, merely on the ground that
in our greatly altered circumstances of life and of mind we are not now
able to trace the imitations.

As a matter of fact, however, the evidence which we have on the subject
is not all negative. On the contrary, there is an overwhelming body of
actual and unquestionable proof of the imitative origin of very many
words in all languages—especially those which are spoken by savages,
and are known from their general structure to be in a comparatively
undeveloped state. The evidence being much too copious for quotation, I
must content myself with referring to the excellent and most forcible
epitome which is given of it by Archdeacon Farrar in his works on the
_Origin of Language_ and _Chapters on Language_.[207] The foregoing
remarks, therefore, which I have made on the negative side of the
question, are merely intended to show that the element of onomatopœia
must have entered into the composition of aboriginal speech much more
largely than philologists are now able to prove, notwithstanding that
they have been able to prove how immensely important an element it has
been in this respect. The only wonder is, that when so many causes
have been at work in obscuring and corroding the originally imitative
significance of words, this significance should still admit of being
traced in all languages—even the most highly conventionalized—to the
very large extent in which it does.

The hostility which Professor Max Müller has displayed to the
onomatopoetic theory of the origin of language is the more remarkable,
because in his latest work he has enthusiastically embraced a special
branch of this theory, which has been put forward by M. Noiré.
This special branch of the onomatopoetic theory is that articulate
sign-making had its origin in sounds which are made by bodies of men
when engaged in some common occupation. When sailors row, soldiers
march, builders co-operate in pulling or in lifting, &c., there is
always a tendency to give vent to appropriate sounds, which the nature
of the occupation usually breaks up into rhythmic periods. “These
utterances, noises, shouts, hummings, or songs are a kind of natural
reaction against the inward disturbance caused by muscular effort.
They are the almost involuntary vibrations of the voice, corresponding
to the more or less regular movements of our whole bodily frame.” The
hypothesis, therefore, is that sounds thus naturally evolved, and
differing with different occupations, would sooner or later come to be
conventionally used as the names of these different occupations. And,
if thus used habitually, they would be virtually the same as words,
inasmuch as they would not merely admit of immediate understanding
on the part of others, but, what is even of more importance, they
would, by the mere fact of such conventional usage of names, elevate
what had previously been but a receptual appreciation of an act into a
pre-conceptual designation of it.

Now, I say that this hypothesis, whatever may be thought as to its
probability, is clearly but a special branch of the general theory
of onomatopœia. So that primitive names were intentionally imitative
of natural sounds, for all the purposes of onomatopoetic theory it
makes no difference whether such sounds were made by natural objects
or by man himself. Nor, of the natural sounds which were made by man
himself, does it in any way affect this theory whether the naturally
human sounds were “interjectional” only, “co-operative” only, or
sometimes one and sometimes the other. If, following the example set
by Professor Max Müller, I may be allowed to designate Noiré’s special
branch of the onomatopoetic theory as the Yeo-he-ho theory, it appears
to me impossible to distinguish it in any essential particular from
those other branches which are called by him the Bow-wow and Pooh-pooh
theories—_i.e._ the imitative and the interjectional. Yet he has
become as ardent a supporter of the one branch as he was a vehement
opponent of the others.[208]

For my own part, I think it highly probable that there is an element
of truth in the Yeo-he-ho theory, although I deem it in the last
degree improbable that imitative sounds of this kind constituted the
_only_ source of aboriginal speech. At the most, it seems to me, this
branch of onomatopœia can be accredited with supporting but a small
proportional part of aboriginal language-growth. Nevertheless, as
already observed, I can have no doubt at all that the principle of
onomatopœia in all its branches has been the most important of all
principles which were concerned in the first genesis of speech. That is
to say, I fully agree with the almost unanimous voice of philological
authority on this matter, which may be tersely expressed by allowing
Professor Whitney to act as spokesman.

“Beyond all reasonable question, there was a positively long period
of purely imitative signs, and a longer one of mixed imitative and
traditional ones, the latter gradually gaining upon the former, before
the present condition of things was reached, when the production of new
signs by imitation is only sporadic and of the utmost rarity, and all
language-signs besides are traditional, their increase in any community
being solely caused by variation and combination, and by borrowing from
other communities.”[209]

But now, having thus stated as emphatically as possible my acceptance
of the theory of onomatopœia, I have to express dissent from many of
its more earnest advocates where they represent that it is necessarily
the only theory to be entertained. In other words, I do not agree with
the dogma that articulate speech cannot possibly have had any source,
or sources, other than that which is supplied by vocal imitations.[210]
For, on merely antecedent grounds, I can see no adequate reason for
arbitrarily excluding the possibility of arbitrary invention. If even
civilized children, who are not under the discipline of the “mother
of invention,” will coin a language of their own in which the element
of onomatopœia is barely traceable;[211] and if uneducated deaf-mutes
will spontaneously devise articulate sounds which are necessarily
destitute of any imitative origin;[212] I do not see why it should be
held antecedently impossible that primitive man can have found any
other means of word-formation than that which is supplied by mimicry.
Therefore, while I fully agree with Professor Wundt in holding that
the question before us is one to be dealt with by psychology rather
than philology (seeing that language cannot record the conditions of
its own birth, and that so many causes have been at work to obliterate
aboriginal onomatopœia), I cannot follow him where he argues that
on grounds of psychology there is no room for any other inference
than that the principle of onomatopœia in its widest sense must have
constituted the sole origin of significant articulation.[213]

We have already seen that even the most imitative of vocalists, the
talking birds, will invent wholly arbitrary sounds as denotative
names,[214] and it would be psychologically absurd to suppose that
they are superior to what primitive man must have been in the matter
of finding expedients for semiotic utterance. Again, the clicks of
Hottentots and Bushmen, whatever we suppose their origin to have been,
certainly cannot have had that origin in onomatopœia; and no less
certainly, as Professor Sayce remarks, they still survive to show how
the utterances of speechless man could be made to embody and convey
ideas.[215] Lastly, on the general principle that the development of
the individual furnishes information touching the development of the
race, it is highly significant that the _hitherto speechless_ child
will spontaneously use arbitrary sounds (both articulate and otherwise)
whereby to denotate habitual recepts. And even after it has begun to
learn the use of actual words, arbitrary additions are frequently
made to its vocabulary which defy any explanation at the hands of
onomatopœia—not only, as in the cases above alluded to, where they are
left to themselves, but even in cases where they are in the closest
contact with language as spoken by their elders. I could quote many
instances of this fact; but it will be enough to refer to one already
given on page 144 (foot-note). When, however, these spontaneous efforts
are not controlled by constant association with elders, but fostered
by children of about the same age being left much together, the
remarkable consequence previously alluded to arises—namely, a newly
devised language which depends but in small part upon the principle
of onomatopœia, and is therefore wholly unintelligible to all but its
inventors.[216]

I have now briefly stated all the main facts and considerations
which appear to me worth stating, both for and against the theory of
onomatopœia. And, having done this, I wish in conclusion to make it
clear that the matter is not one which seriously affects the theory of
evolution. To the philologist, no doubt, the question as to how far
the element of onomatopœia entered into the formation of aboriginal
speech is a really important question, so that, as Geiger says,
“Diess ist die gemeinsame Frage, und die antwort wird auf der einen
Seite von einem inneren Zusammenhang zwischen je einem Laut und dem
entsprechenden Begriffe, auf der andern aus Willkür und Uebereinkunft
hergeleitet.”[217] But the question is one which the evolutionist may
view with indifference. Whether words were all originally dependent
on an inherent connection between every sound they made and the idea
thereby expressed, or whether they were all due to arbitrary invention,
in either case the evolutionist may see that they can equally well have
come into existence as the natural products of a natural psychogenesis.
And, _a fortiori_, as an evolutionist, he need not greatly concern
himself with any further question as to the relative degrees in which
imitation and invention may have entered into the composition of
primitive speech.




CHAPTER XIV.

THE WITNESS OF PHILOLOGY.


We are now in a position to consider certain matters which are of high
importance in relation to the subject of the present work. In earlier
chapters I have had occasion to show that the whole stress of the
psychological distinction between man and brute must be laid—and, in
point of fact, has been laid by all competent writers who are against
me—on the distinctively human faculty of judgment. Moreover, I have
shown that, by universal consent, this faculty is identical with that
of predication. Any mind that is able, in the strict psychological
signification of the term, to judge, is also able to predicate, and
_vice versâ_. I claim, indeed, to have conclusively shown that certain
writers have been curiously mistaken in their analysis of predication.
These mistakes on their part, however, do not relieve me of the
burden of explaining the rise of predication; and I have sought to
discharge the burden by showing how the faculty must have been given
in germ so soon as the denotative stage of sign-making passed into the
connotative, and thus furnished the condition to bringing into contact,
or _apposition_, the names of objects and the names of qualities or
actions. The discussion of this important matter, however, has so far
proceeded on grounds of psychological analysis alone. The point has
now arrived when we may turn upon the subject the independent light
of philological analysis. Whereas we have hitherto considered, on
grounds of mental science only, what _must have been_ the genesis of
predication—supposing predication to have had a genesis,—we have
next to ascertain whether our deduction admits of corroboration by any
inductive evidence supplied by the science of language, as to what this
genesis _actually was_.

And here I had better say at once that the results of philological
science will be found to carry us back to an even more primitive state
of matters than any which I have hitherto contemplated. For, so long
as I was restricted to psychological analysis, I was obliged to follow
my opponents where they take language as it now exists. In order to
argue with them at all upon these grounds, it was necessary for me to
consider what they had said on the philosophy of predication; and, in
order to do this, it was further necessary that I should postpone for
independent treatment those results of philological inquiry which they
have everywhere ignored. But now we have come to the place where we
can afford to abandon psychological analysis altogether, and take our
stand upon the still surer ground of what I have already termed the
palæontological record of mental evolution as this has actually been
preserved in the stratified deposits of language. Now, when we do this,
we shall find that hitherto we have not gone so far back in tracing the
genesis of conceptual out of receptual ideation as in point of fact we
are able to go on grounds of the most satisfactory evidence.

Up to this time, then, I have been meeting my opponents on their own
assumptions, and one of these assumptions has been that language must
always have existed as we now know it—at least to the extent of
comprising words which admit of being built up into propositions to
express the semiotic intention of the speaker. But this assumption
is well known by philologists to be false. As a matter of fact,
language did not begin with any of our later-day distinctions between
nouns, verbs, adjectives, prepositions, and the rest: it began as the
undifferentiated protoplasm of speech, out of which all these “parts
of speech” had afterwards to be developed by a prolonged course of
gradual evolution.” Die Sprache ist nicht stückweis order atomistisch;
sie ist gleich in allen ihren Theilen als Ganzes und demnach organisch
entstanden.”[218]

This highly general and most important fact is usually stated as it
was, I believe, first stated by the anthropologist Waitz, namely,
that “the unit of language is not the word, but the sentence;”[219]
and, therefore, that historically the sentence preceded the word.
Or, otherwise and less ambiguously expressed, every word was
originally itself a proposition, in the sense that of and by itself
it conveyed a statement. Of course the more that a single word thus
assumed the functions now discharged by several words when built
into a proposition, the more generalized—that is to say, the less
defined—must have been its meaning. The sentence or proposition as
we now have it represents what may be termed a psychological division
of labour as devolving upon its component parts: subject-words,
attributive-words, qualifying-words indicative of time, place, agent,
instrument, and so forth, are now all so many different organs
of language, which are set apart for the performance of as many
different functions of language. The life of language under this its
fully evolved form is, therefore, much more complex, and capable of
much more refined operations, than it was while still in the wholly
undifferentiated condition which we have now to contemplate.

In order to gain a clear conception of this protoplasmic condition of
language, we had better first take an example of it as it is presented
to our actual observation in the child which is just beginning to
speak. For instance, as Professor Max Müller points out, “if a child
says ‘Up,’ that _up_ is, to his mind, noun, verb, adjective, all in
one. If an English child says ‘Ta,’ that _ta_ is both noun (thanks),
and a verb (I thank you). Nay, even if a child learns to speak
grammatically, it does not yet think grammatically; it seems, in
speaking, to wear the garments of its parents, though it has not yet
grown into them.”[220]

Again, as Professor Friedrich Müller says, “the child’s word _Ba-ba_,
sleep, does not mean sleep only, as a particular kind of repose, but
rather also all the circumstances which appertain to sleep, such as
cot, bed, bolster, bed-clothes, &c.[221] It likewise and indifferently
means, sleeping, sleepy, sleeper, &c., and may stand for any variety of
propositions, such as “I am sleepy,” “I want to go to sleep,” “He is
asleep,” &c.

Of course innumerable other illustrations might be given; but these
are enough to show what is meant by a “sentence-word.” The next thing
we have to notice is the manner in which a young child particularizes
the meanings of its sentence-words, so as to limit their highly generic
significance _per se_, and thus to make them convey the special
significance intended. Briefly, the one and only means which the child
has of doing this is by the employment of tone and gesture. Here the
suiting of the action to the word is a necessary condition to semiotic
utterance; the more primitive forms of sign-making are the needful
supplements to these commencements of higher forms. And not only so;
they are likewise in large part the parents of these higher forms. It
is by pointing (_i.e._ falling back on what I have called the earliest
or “indicative stage” of language) that a child is able to signify the
place, agent, instrument, &c., to which it requires a sentence-word to
apply; and thus we catch our first glimpse of the highly important fact
that the earliest indications of grammar are given by the simultaneous
use of sentence-words and gesture-signs.

It will now be my object to prove, that in the history of the race
spoken language began in the form of sentence-words; that grammar
is the child of gesture; and, consequently, that predication is but
the adult form of the self-same faculty of sign-making, which in its
infancy we know as indication. Being myself destitute of authority in
matters philological, I will everywhere rely upon the agreement of
recognized leaders of the science.

Bunsen, I believe, was the first to point out that in Egyptian there
is no formal distinction between noun, adjective, verb, or particle;
such a word as _anh_, for instance, meaning indifferently, life, alive,
to live, lively, &c.[222] Similarly, in Chinese “the word can still be
used indifferently as a noun, a verb, an adverb, or the sign of a case,
much like such English words as silver, and picture, and its place in
the sentence alone determines in what sense it shall be construed.
This is an excellent illustration of the early days of speech, when
the sentence-words contained within themselves all the several parts
of speech at once—all that was needed for a complete sentence; and it
was only by bringing them into contact and contrast [i.e. _apposition_]
with other sentence-words, that they came to be restricted in their
meaning and use, and to be reduced to mere ‘words.’”[223]

Later on I will give abundant evidence of a similar state of matters
in the case of other existing languages presenting a low order of
development—especially those of savages. But perhaps it is even
of more importance to prove that the most highly developed of all
languages—namely, the Indo-European group—still bears unmistakable
evidence of having passed through this primitive phase. This is a
statement which it would be easy to substantiate by any number of
quotations; but I will only call the testimony of one witness in the
person of Professor Max Müller, whose evidence on this point may be
regarded as that of an opponent.

“Nothing, it is true, can exist in language except what is a sentence,
_i.e._ that conveys a meaning; but for that very reason it ought
to have been perceived that every word must originally have been a
sentence. The mere root, _quâ_ root, cannot be called a sentence, and
in that sense a mere root may be denied the dignity of a word. But as
soon as a root is used for predication, it becomes a word, whether
outwardly it is changed or not. What in Chinese is effected by position
or by tone, namely, the adaptation of a root to serve the purposes of
words, is in the Aryan languages achieved by means of suffixes and
terminations, though often also by change of tone. We saw that, in an
earlier stage, the Aryan languages, too, could raise a root into a
word, without the aid of suffixes, and that, for instance, _yudh_, to
fight, could be used in the five senses of the act of fighting, the
agent of fighting, the instrument of fighting, the place of fighting,
and the result of fighting. For the sake of distinction, however, as
soon as the necessity began to be felt, the Aryan language introduced
derivative elements, mostly demonstrative or pronominal.”

“The imperative may truly be called the most primitive sentence, and it
is important to observe how little in many languages it deviates from
what has been fixed upon as the true form of a root ... _va_, weave,
whether as a reminder or as a command, would have as much right to be
called a sentence as when we say, ‘Work,’ _i.e._ ‘Let us work.’ ...
From the use of a root in the imperative, or in the form of a general
assertion, there is a very easy transition to its employment in other
senses and for other purposes.... A master requiring his slaves to
labour, and promising them their food in the evening, would have no
more to say than ‘Dig—Feed,’ and this would be quite as intelligible
as ‘Dig, and you shall have food,’ or, as we now say, ‘If you dig, you
shall have food.’”[224]

Thus we may lay it down as a general doctrine or well-substantiated
principle of philological research, that “Language begins with
sentences; not with single words;”[225] or that originally every
word in and of itself required to convey a meaning, after the manner
of the early utterances of children. “The sentence is the only unit
which language can know, and the ultimate starting-point of all our
linguistic researches.... If the sentence is the unit of significant
speech, it is evident that all individual words must once have been
sentences; that is to say, when first used they must each have implied
or represented a sentence.”[226]

“The making of words as distinct from sentences was a long and
laborious process, and there are many languages, like those of North
America, in which the process has hardly yet begun. A dictionary is the
result of reflection, and ages must elapse before a language can enter
upon its reflective stage.”[227]

Or, to give only one more quotation, as Professor Max Müller says, “it
is difficult for us to think in Chinese, or in any radical language,
without transferring to it our categories of thought. But if we watch
the language of a child, which is really Chinese spoken in English,
we see that there is a form of thought, and of language, perfectly
rational and intelligible to those who have studied it, in which,
nevertheless, the distinction between noun and verb, nay, between
subject and predicate, is not yet realized.”[228]

Starting, then, from this undifferentiated condition of language, let
us next see how the “parts of speech” became evolved.

There appears to be no doubt that one of the earliest parts of
speech to become differentiated was the pronoun. Moreover, all the
pronouns (or “pronominal elements”) as originally differentiated
were indistinguishable from what we should now call adverbs; and
they were all concerned with denoting relations of place.[229] No
exception to this general statement can be made even as regards the
personal pronouns. “_Hic_, _iste_, _ille_, are notoriously a sort
of correlatives to _ego_, _tu_, _sui_, and, if the custom of the
languages had allowed it, might, on every occasion, be substituted
for them.”[230] Now, there is very good reason to conclude that these
pronominal adverbs, or adverbial pronouns, were in the first instance
what may be termed articulate translations of gesture-signs—_i.e._
of a pointing to place-relations. _I_ being equivalent to _this one_,
_he_ or _she_ or _it_ to _that one_, &c., we find it easy to supply
the indicative gestures out of which these denotative terms arose; and
although we are not now able to supply the phonetic source of these
highly ancient “pronominal” or “demonstrative elements,” it is easy to
imagine that they may have arisen in the same apparently spontaneous
way as very young children will now devise arbitrary sounds, both as
proper names and as adverbs of position. That we should not err in
thus comparing the grade of mental evolution exhibited by the earliest
framers of spoken language with that of a young child, is rendered
apparent by the additional and highly interesting fact, that, just as
a young child begins by speaking of the _Ego_ in the third person, so
it was with early man in his use of personal pronouns. “Man regarded
himself as an object before he learnt to regard himself as a subject;
and hence ‘the objective cases of the personal as well as of the other
pronouns are always older than the subjective;’ and the Sanskrit _mâm_,
_ma_ (Greek [Greek: me], Latin _me_) is earlier than _aham_ ([Greek:
egôn] and _ego_).”[231]

Lest it should be thought that I am assuming too much in thus referring
the origin of pronominal elements to gesture-signs, I will here quote
the opinion of Professor Max Müller, who of all philologists is least
open to suspicion of bias towards my side of the present argument.
Speaking of these “demonstrative elements, which point to an object in
space and time, and express what we now express by _then_, _this_ [=
I], _that_ [= there, he, she, it, &c.], near, far, above, below, &c.;”
he says, “in their primitive form and intention they are addressed
to the senses rather than to the intellect: they are sensuous, not
conceptual.”[232] And elsewhere he adds, “I see no reason why we should
not accept them as real survivals of a period of speech during which
pantomime, gesture, pointing with the fingers to actual things were
still indispensable ingredients of all conversation.”[233] Again, “it
was one of the characteristic features of Sanskrit, and the other Aryan
languages, that they tried to distinguish the various applications of a
root by means of what I have called demonstrative roots or elements. If
they wished to distinguish the mat as the product of their handiwork,
from the handiwork itself, they would say ‘Platting—there;’ if they
wished to encourage the work they would say, ‘Platting—they, or you,
or we.’ We found that what we call demonstrative roots or elements
must be considered as remnants of the earliest and almost pantomimic
phase of language, in which language was hardly as yet what we mean by
language, namely _logos_, a gathering, but only a pointing.”[234]

It is the opinion of some philologists, however, that these
demonstrative elements were probably “once full or predicative words,
and that if we could penetrate to an earlier stage of language, we
should meet with the original forms of which they are the maimed
half-obliterated representatives.”[235] But as even these philologists
do not question that all originally “predicative words” would be
found to have had their predicative value determined by gesture, “if
we could penetrate to an earlier stage of language,” the question
whether such demonstrative elements as have come down to us were or
were not themselves of originally predicative value, is not of vital
importance in the present connection. For there is no doubt that
pronominal elements which really were aboriginal as such, depended
on accompanying gesture-signs for a conveyance of their predicative
meaning; and although, as we might expect, there is a necessary absence
of proof in particular cases whether these elements have come down to
us in a practically aboriginal form, or whether they have done so as
the worn-out remnants of independently predicative words, the general
principles on which we are now engaged are not really affected by any
such philological uncertainties in matters of detail. For even the
authority just quoted as doubting whether we have evidence enough to
conclude that demonstrative elements which have come down to us were
never themselves predicative words, elsewhere says of early predicative
utterance in general,—“It is certain that there was a time in the
history of speech when the articulate, or semi-articulate, sounds
uttered by primitive man were made the significant representatives of
thought by the gestures with which they were accompanied; and this
complex of sound and gesture—a complex in which, be it remembered,
the sound had no meaning apart from the gesture—was the earliest
sentence.”[236] And, after giving examples from languages of Further
India, he adds,—“But an inflectional language does not permit us to
watch the word-making process so clearly as do those savage jargons,
in which a couple of sounds, like the Grebo _ni ne_, signify ‘I do
it,’ or ‘You do not,’ according to the context and the gestures of
the speaker. Here by degrees, with the growth of consciousness and
the analysis of thought, the external gesture is replaced by some
portion of the uttered sounds which agrees in a number of different
instances, and in this way the words by which the relations of grammar
are expressed came into being. A similar process has been at work in
producing those analogical terminations whereby our Indo-European
languages adapt a word to express a new grammatical relation.”

Therefore, not unduly to multiply quotations, we may take it as
the now established doctrine of philology that, as even this more
sceptical authority puts it, “Grammar has grown out of gesture and
gesticulation.”[237] Later on I will show in how interesting a manner
early forms of articulate utterance follow in their structure the
language of gesture already treated of in a previous chapter. It was
for the sake of displaying this resemblance that I there occupied so
much space with the syntax of gesture-language; and, therefore, it will
now be my object to trace the family likeness between the constructions
of primitive modes of utterance, and those of the parent gestures from
which these constructions have been directly inherited. But in order
to do this more completely, we must first consider the philology of
predicative words.

The parts of speech which are primarily concerned in predication, and
which, therefore, may be called _par excellence_ predicative words,
are substantives, adjectives, and verbs. I will, therefore, begin by
briefly stating what is known touching the evolution of these parts of
speech.

We have abundant evidence to show that originally there was no
distinction between substantives and adjectives, or object-words and
quality-words. Nor is this at all surprising when we remember that even
in fully developed forms of speech one and the same word may stand as
a substantive or an adjective according to its context. “Cannon” in
“cannonball,” or “pocket” in “pocket-book,” &c., are adjectives in
virtue of position—_i.e._ of _apposition_ with the substantives which
they thus serve to qualify.

Similarly as regards the genitive case. This, also, is of an
attributive quality, and, therefore, like the now independent
adjective, originally had no independent existence. When the force of
the genitive had to be conveyed, it was conveyed by this same device of
apposition. And, lastly, the same device was resorted to for purposes
of predication. Or, to quote these important facts from responsible
sources, Professor Sayce says:—“Even the genitive case, necessary
as it appears to us to be, once had no existence, as indeed it still
has none in groups of languages like the Taic or the Malay. Instead
of the genitive, we here have two nouns placed in apposition to one
another, two individuals, as it were, set side by side without any
effort being made to determine their exact relations beyond the mere
fact that one precedes the other, and is therefore thought of first....
Now, this apposition of two nouns, which still serves the purpose of
the genitive in many languages, might be regarded as attributive or
as predicative. If predicative, then the two contrasted nouns formed
a complete sentence, ‘Cup gold,’ for instance, being equivalent to
‘The cup is gold.’ If attributive, then one of the two nouns took the
place of an adjective, ‘gold cup’ being nothing more than ‘a golden
cup.’”[238] Then, after giving examples from different languages of
the artificial contrivances whereby in course of time these three
grammatical differentiations originated (namely, by conventional
changes of position between the words apposed, in some cases the form
of predication being A B, and that of attribution or possession B A,
while in other languages the reverse order has obtained), Professor
Sayce goes on to say:—“These primitive contrivances for distinguishing
between the predicate, the attribute, and the genitive, when the three
ideas had in the course of ages been evolved by the mind of the
speaker, gradually gave way to the later and more refined machinery of
suffixes, auxiliaries, and the like.”[239]

For the sake of putting this point beyond the reach of question, I will
quote another and independent authority to the same general effect.

“It is a curious fact hitherto overlooked by grammarians and logicians,
that the definition of a noun applies strictly only to the nominative
case. The oblique cases are really attribute-words, and the inflection
is practically nothing but a device for turning a noun into an
adjective or adverb. This is perfectly clear as regards the genitive,
and, indeed there is historical evidence to show that the genitive in
Aryan languages was originally identical with an adjective ending;
‘man’s life’ and ‘human life’ being expressed in the same way. It is
also clear that ‘noctem’ in ‘flet noctem’ is a pure adverb of time.
It is not so easy to see that the accusative in such sentences as ‘He
beats the boy’ is also a sort of adverb, because the connection between
verb and object is so intimate as almost to form one simple idea, as
in the case of noun-composition. But it is clear that if ‘boy’ in the
compound ‘boy-beating’ is an attribute-word, it can very well be so
also when ‘beating’ is thrown into the verbal form without any change
of meaning.”[240]

Lastly, upon this point Professor Max Müller says, while speaking of
Aryan adjectives:—“These were not used for the first time when people
said ‘The sun is bright,’ but when they predicated the quality of
brightness, or the act of shooting out light, and said, as it were,
‘Brightness-here.’ Adjectives, in fact, were formed, at first, exactly
like substantives, and many of them could be used in both characters.
There are languages in which adjectives are not distinguished from
substantives. But though outwardly alike, they are conceived as
different from substantives the moment they are used in a sentence for
the purpose of predicating or of qualifying a substantive.”[241]

So much, then, for substantives and adjectives: it cannot be said that
there is any evidence of historical priority of the one over the other;
but rather that so soon as the denotative meanings of substantives
became fixed, they admitted of having imparted to them the meanings
of adjectives, genitives, and predicates, by the simple expedient of
apposition—an expedient which, as we have seen in earlier chapters,
is rendered inevitable by the laws of association and “the logic of
events:” it is an expedient that must have been furnished _to_ the
mind, and therefore need never have been intentionally devised _by_ it.

Turning next to the case of verbs, or the class of words upon which
more especially devolves the office of predication, it is the opinion
of some philologists that these arose through the apposition of
substantives with the genitives of pronouns.[242] And there can be
no doubt that in many actually existing languages the functions of
predication are still discharged in this way, without the existence of
any verbs at all, as we shall see later on. But, on the other hand, it
is shown that a great many Aryan substantives were formed by joining
pronominal elements to previously existing verbal roots, in a manner
so strongly suggestive of pointing-gestures, that it is difficult to
doubt the highly primitive source of the construction. For example
“digging-he” = labourer, “digging-it” = spade, “digging-here” = labour,
“digging-there” = hole,[243] &c. Or again, “‘The hole is dark’ would
have been expressed originally (in Aryan) by ‘digging-it,’ ‘hiding
here,’ or, ‘hiding-somewhere.’ ‘Hiding-here’ might afterwards be
used in the sense of a hiding-place. But when it was used as a mere
qualifying predicate in a sentence in which there was but one subject,
it assumed at once the character of an adjective.”[244]

To me it appears evident that there is truth in both these views,
which, therefore, are in no way contradictory to one another. We have
evidence that many substantives were of later origin than many verbs,
and _vice versâ_; but this does not show which of these two parts of
speech preceded the other as a whole. Nor does it appear that we are
likely to obtain any definite evidence upon the point. On psychological
grounds, and from the analogy furnished by children, we might be
prepared to think it most probable that substantives preceded verbs;
and this view is no doubt corroborated by the remarkable paucity of
verbs in certain savage languages of low development. But as a matter
of pure philology “we cannot derive either the verb from the noun,
or the noun from the verb.”[245] This writer goes on to say, “they
are co-existent creations, belonging to the same epoch and impulse of
speech.” But whether or not this inference represents the truth is a
matter of no importance for us. With or without verbs, primitive man
would have been able to predicate—in the one case after the manner of
children who have just begun to learn the use of them, and in the other
case after the manner of those savages recently mentioned, who throw
upon their nouns, in conjunction with pronouns, the office of verbs.

Seeing that my psychological opponents have laid so much stress upon
the substantive verb as this is used by the Romance languages in
formal predication, I will here devote a paragraph to its special
consideration from a philological point of view. It will be remembered
that I have already pointed out the fallacy which these opponents have
followed in confounding the substantive verb, as thus used, with the
copula—it being a mere accident of the Romance languages that the two
are phonetically identified. Nevertheless, even after this fallacy
has been pointed out to them, my opponents may seek to take refuge
in the substantive verb itself: forced to acknowledge that it has
nothing especially to do with predication, they may still endeavour to
represent that elsewhere, or in itself, it represents a high order of
conceptual thought. This, of course, I allow; and if, as my opponents
assume, the substantive verb belonged to early, not to say primitive
modes of speech, I should further allow that it raises a formidable
difficulty in the otherwise even path of evolutionary explanation.
But, as a matter of fact, these writers are no less mistaken about the
primitive nature of the substantive verb itself, than they are upon the
function which it accidentally discharges in copulation.[246] In order
to prove this, or to show that the substantive verb is really very far
from primitive, I will furnish a few extracts from the writings of
philological authorities upon the subject.

“Whatever our _a priori_ estimate of the power of the verb-substantive
may be, its origin is traced by philology to very humble and material
sources. The Hebrew verbs חחמוה (_houa_) or הוה (_haia_) may very
probably be derived from an onomatopœia of respiration. The verb
_kama_, which has the same sense, means primitively ‘to stand out,’
and the verb _koum_, ‘to stand,’ passes into the sense of ‘being.’
In Sanskrit, _as-mi_ (from which all the verbs-substantives in the
Indo-European languages are derived, as [Greek: eimi], _sum_, am; Zend
_ahmi_; Lithuanic, _esmi_, Icelandic, _em_, &c.) is, properly speaking,
no verbal root, but ‘a formation on the demonstrative pronoun _sa_, the
idea meant to be conveyed being simply that of local presence.’ And of
the two other roots used for the same purpose, namely, _bhu_ ([Greek:
phuô], _fui_, &c.) and _sthâ_ (_stare_, &c.), the first is probably an
imitation of breathing, and the second notoriously a physical verb,
meaning ‘to stand up.’ May we not, then, ask with Bunsen, ‘What is
_to be_ in all languages but the spiritualization of _walking_ or
_standing_ or _eating_?’”[247]

Again, to quote only one other authority:—“In closing, for the
present, the discussion of this extensive subject, it is proposed to
make a few remarks upon the so-called verb-substantive, respecting
the nature and functions of which there has perhaps been more
misapprehension than about any other element of language. It is
well known that many grammarians have been accustomed to represent
this element as forming the basis of all verbal expression, and as
a necessary ingredient in every logical proposition. It would seem
to follow, from this statement, that nations so unfortunate as to
be without it, could neither employ verbal expression nor frame a
logical proposition. How far this is the case will be seen hereafter:
at present we shall make some brief remarks on this verb, and on the
substitutes usually employed in dialects where it is formally wanting.
It will be sufficient to produce a few prominent instances, as the
multiplying of examples from all known languages would be a mere
repetition of the same general phenomena.

“In the portion of the essay relating to the Coptic, it was observed:
‘What are called the auxiliary and substantive verbs in Coptic are
still more remote from all essential verbal character (than the
so-called verbal roots). On examination they will almost invariably be
found to be articles, pronouns, particles, or abstract nouns, and to
derive their supposed verbal functions entirely from their accessories,
or from what they imply.’ In fact any one who examines a good Coptic
grammar or dictionary will find that there is nothing formally
corresponding to our _am_, _art_, _is_, _was_, &c., though there is a
counterpart to Lat. _fieri_ (_sthopi_) and another to _poni_ (_chi_,
neuter passive of _che_); both occasionally rendered _to be_, which,
however, is not their radical import. The Egyptians were not, however,
quite destitute of resources in this matter, but had at least half a
dozen methods of rendering the Greek verb-substantive when they wished
to do so. The element most commonly employed is the demonstrative _pe_,
_te_, _ne_; used also in a slightly modified form for the definite
article; _pe_ = is, having reference to a subject in the singular
masculine; _te_, to a singular feminine; and _ne_ = are, to both
genders in the plural. The past tense is indicated by the addition of a
particle expressing remoteness. Here, then, we find as the counterpart
of the verb-substantive an element totally foreign to all the received
ideas of a verb; and that instead of its being deemed necessary to
say in formal terms ‘Petrus est,’ ‘Maria est,’ ‘Homines sunt,’ it is
quite sufficient, and perfectly intelligible, to say, ‘Petrus hic,’
‘Maria hæc,’ ‘Homines hi.’ The above forms, according to Champollion
and other investigators of ancient hieroglyphics, occur in the oldest
known monumental inscriptions, showing plainly that the ideas of the
ancient Egyptians as to the method of expressing the category _to be_,
did not exactly accord with those of some modern grammarians.... Every
Semitic scholar knows that personal pronouns are employed to represent
the verb-substantive in all the known dialects, exactly as in Coptic,
but with less variety of modification. In this construction it is not
necessary that the pronoun should be of the same person as the subject
of the proposition. It is optional in most dialects to say either _ego
ego, nos nos_, for _ego sum, nos sumus_, or _ego ille, nos illi_. The
phrase ‘Ye are the salt of the earth,’ is, in the Syriac version,
literally ‘You they (_i.e._ the persons constituting) the salt of the
earth.’ Nor is this employment of the personal pronoun confined to the
dialects above specified, it being equally found in Basque, in Galla,
in Turco-Tartarian, and various American languages.... It is true
that the Malayan, Javanese, and Malagassy grammarians talk of words
signifying _to be_; but an attentive comparison of the elements which
they profess to give as such, shows clearly that they are no verbs at
all, but simply pronouns or indeclinable particles, commonly indicating
the time, place, or manner of the specified action or relation. It is
not therefore easy to conceive how the mind of a Philippine islander,
or of any other person, can supply a word totally unknown to it, and
which there is not a particle of evidence to show that it was ever
thought of.... A verb-substantive, such as is commonly conceived,
vivifying all connected speech, and binding together the terms of every
logical proposition, is much upon a footing with the phlogiston of
the chemists of the last generation, regarded as a necessary pabulum
of combustion, that is to say, _vox et præterea nihil_.... If a given
subject be ‘I,’ ‘thou,’ ‘he,’ ‘this,’ ‘that,’ ‘one;’ if it be ‘here,’
‘there,’ ‘yonder,’ ‘thus,’ ‘in,’ ‘on,’ ‘at,’ ‘by;’ if it ‘sits,’
‘stands,’ ‘remains,’ or ‘appears,’ we need no ghost to tell us that it
_is_, nor any grammarian or metaphysician to proclaim that recondite
fact in formal terms.”[248]

       *       *       *       *       *

Having thus briefly considered the philology of predicative words, we
must next proceed to the not less important matter of the philology
of predication itself. And here we shall find that the evidence is
sufficiently definite. We have already seen good reason for concluding
that what Grimm has called the “antediluvian” pronominal roots were the
phonetic equivalents of gesture-signs—or rather, that they implied
accompanying gesture-signs for the conveyance of their meaning. Now, it
is on all hands allowed that these pronominal roots, or demonstrative
elements, afterwards became attached to nouns and verbs as affixes or
suffixes, and so in older languages constitute the machinery both of
declension and conjugation. Thus, we can trace back, stage by stage,
the form of predication as it occurs in the most highly developed, or
inflective, languages, to that earliest stage of language in general,
which I have called the indicative. In order to show this somewhat more
in detail, I will begin by sketching these several stages, and then
illustrate the earliest of them that still happen to survive by quoting
the modes of predication which they actually present.

As we thus trace language backwards, its structure is found to
undergo the following simplification. First of all, auxiliary words,
suffixes, affixes, prepositions, copulas, particles, and, in short,
all inflections, agglutinations, or other parts of speech which are
concerned in the indication of _relationship_ between the other
component parts of a sentence, progressively dwindle and disappear.
When these, which I will call relational words, are shed, language is
left with what may be termed object-words (including pronominal words),
attributive-words, action-words, and words expressive of states of mind
or body, which, therefore, may be designated condition-words. Roughly
speaking, this classification corresponds with the grammatical nouns,
pronouns, adjectives, active verbs, and passive verbs; but as our
regress through the history of language necessitates a total disregard
of all grammatical forms, it will conduce to clearness in my exposition
if we consent to use the terms suggested.

The next thing we notice is that the distinction between object-words
and attributive-words begins to grow indistinct, and eventually all but
disappears: substantives and adjectives are fused in one, and whether
the resulting word is to be understood as subject or predicate—as the
name of the object or the name of a quality—depends upon its position
in the sentence, upon the tone in which it is uttered, or, in still
earlier stages, upon the gestures by which it is accompanied. Thus,
as Professor Sayce remarks, “the apposition of two substantives [and,
_a fortiori_, of two such partly or wholly undifferentiated words as
we are now contemplating] is the germ out of which no less than three
grammatical conceptions have developed—those of the genitive, of the
predicate, and of the adjective.”[249]

While this process of fusion is being traced in the case of
substantives and adjectives, it becomes at the same time observable
that the definition of verbs is gradually growing more and more vague,
until it is difficult, and eventually impossible, to distinguish a verb
at all as a separate part of speech.

Thus we are led back by continuous stages, or through greater and
greater simplifications of language-structure, to a state of things
where words present what naturalists might term so generalized a
type as to include, each within itself, all the functions that
afterwards severally devolve upon different parts of speech. Like those
animalcules which are at the same time but single cells and entire
organisms, these are at the same time single words and independent
sentences. Moreover, as in the one case there is life, in the other
case there is meaning; but the meaning, like the life, is vague
and unevolved: the sentence is an organism without organs, and is
generalized only in the sense that it is protoplasmic. In view of
these facts (which, be it observed, are furnished by languages still
existing, as well as by the philological record of languages long since
extinct) it is impossible to withhold assent from the now universal
doctrine of philologists—“language diminishes the farther we look back
in such a way, that we cannot forbear concluding it must once have had
no existence at all.”[250]

       *       *       *       *       *

From all the evidence which has now been presented showing that
aboriginally words were sentences, it follows that aboriginally
there can have been no distinction between terms and propositions.
Nevertheless, although this follows deductively from the general truth
in question, it is desirable that we should study in more detail the
special application of the principle to the case of formal predication,
seeing that, as so often previously remarked, this is the place where
my opponents have taken their stand. The reader will remember that I
have already disposed of their assertions with regard to the copula.
It will now be my object to show that their analysis is equally
erroneous where it is concerned with both the other elements of which a
formal proposition consists. Not having taken the trouble to acquaint
themselves with the results of linguistic research, and therefore
relying only on what may be termed the accidents of language as these
happen to occur in the Aryan branch of the great language-tree, these
writers assume that a proposition must always and everywhere have been
thrown into the precisely finished form in which it was analyzed by
Aristotle. As a matter of fact, however, it is now well known that
such is not the case; that the form of predication as we have it in
our European languages has been the outcome of a prolonged course of
evolution; and that in its most primitive stage, or in the earliest
stage which happens to have been preserved in the palæontology of
language, predication can scarcely be said to have been differentiated
from what I have called indication. For the sake of placing this
important fact beyond the reach of doubt, I will begin by quoting the
statements of a few among the leading authorities upon the philology of
the subject.

“Primitive man would not trouble himself much with such propositions
as ‘Man is mortal,’ ‘Gold is heavy,’ which are a source of such
unfailing delight to the formal logician; but if he found it necessary
to employ permanent attribute-words, would naturally throw them into
what is called the attributive form, by placing them in immediate
proximity with the noun, whose inflections they would afterwards
assume. And so the verb gradually came to assume the purely formal
function of predication. The use of verbs denoting action necessitated
the formation of verbs to denote ‘rest,’ ‘continuance in state,’ and
when, in course of time, it became necessary in certain cases to
predicate permanent as well as changing attributes, these words were
naturally employed for the purpose, and such a sentence as ‘The sun
continues bright’ was simply ‘The bright sun’ in another form. By
degrees these verbs became so worn away in meaning, gradually coming
to signify simple existence, that at last they lost all vestiges of
meaning whatever, and came simply to be marks of predication. Such is
the history of the verb ‘to be,’ which in popular language has entirely
lost even the sense of ‘existence.’ Again, in a still more advanced
state, it was found necessary to speak, not only of things, but of
their attributes. Thus such a sentence as ‘Whiteness is an attribute of
snow,’ has identically the same meaning as ‘Snow is white’ and ‘White
snow;’ and the change of ‘white’ into ‘whiteness’ is a purely formal
device to enable us to place an attribute-word as the subject of a
proposition.”[251]

“Now comes a very important consideration, that not only is the order
of subject and predicate to a great extent conventional, but that the
very idea of the distinction between subject and predicate is purely
linguistic, and has no foundation in the mind itself. In the first
place, there is no necessity for a subject at all: in such a sentence
as ‘It rains,’ there is no subject whatever, the _it_ and the terminal
_s_ being merely formal signs of predication. ‘It rains: therefore I
will take my umbrella,’ is a perfectly legitimate train of reasoning,
but it would puzzle the cleverest logician to reduce it to any of his
figures. Again, the mental proposition is not formed by thinking first
of the subject, then of the copula, and then of the predicate; it is
formed by thinking of the three simultaneously. When we formulate in
our minds the proposition ‘All men are bipeds,’ we have two ideas, ‘all
men’ and ‘an equal number of bipeds,’ or, more tersely, ‘as many men,
as many bipeds,’ and we think of the two ideas simultaneously [_i.e._
in _apposition_] not one after the other, as we are forced to express
them in speech. The simultaneity of conception is what is expressed by
the copula in logic, and by the various forms of sentences in language.
It by no means follows that logic is entirely destitute of value, but
we shall not arrive at the real substratum of truth until we have
eliminated that part of the science which is really nothing more than
an imperfect analysis of language.”[252]

Again, as a result of his prolonged study of some of the most primitive
forms of language still extant among the Bushmen of South Africa, Dr.
Bleek entertains no doubt whatever that aboriginally the same word,
without alteration, implied a substantival or a verbal meaning, and
could be used indifferently also as an adjective, adverb, &c.[253] That
is to say, primitive words were sentence-words, and as such were used
by early man in just the same way as young children use their hitherto
undifferentiated signs, _Byby_ = _sleep_, _sleeping_, _to sleep_,
_sleeper_, _asleep_, _sleepy_, &c.; and, by connotative extension,
_bed_, _bolster_, _bed-clothes_, &c.

Lastly, as already indicated, we are not left to mere inference
touching the aboriginal state of matters with regard to predication.
For in many languages still existing we find the forms of predication
in such low phases of development, that they bring us within easy
distance of the time when there can have been no such forms at all.
Even Professor Max Müller allows that there are still existing
languages “in which there is as yet no outward difference between what
we call a root, and a noun or a verb. Remnants of that phase in the
growth of language we can detect even in so highly developed a language
as Sanskrit.” Elsewhere he remarks:—“A child says, ‘I am hungry,’
without an idea that _I_ is different from _hungry_, and that both are
united by an auxiliary verb.... A Chinese child would express exactly
the same idea by one word, ‘Shi,’ _to eat_, or _food_, &c. The only
difference would be that a Chinese child speaks the language of a
child, an English child the language of a man.”[254]

It is no doubt remarkable that the Chinese should so long have retained
so primitive a form; but, as we know, the functions of predication
have here been greatly assisted by devices of syntax combined with
conventionally significant intonation, which really constitute Chinese
a well-developed language of a particular type. Among peoples of a
much lower order of mental evolution, however, we are brought into
contact with still more rudimentary forms of predication, inasmuch
as these devices of syntax and intonation have not been evolved. As
previously stated, the most primitive of all actually existing forms
of predication where articulate language is concerned, is that wherein
the functions of a verb are undertaken by the apposition of a noun with
what is equivalent to the genitive case of a pronoun. Thus, in Dayak,
if it is desired to say, “Thy father is old,” “Thy father looks old,”
&c., in the absence of verbs it is needful to frame the predication by
mere apposition, thus:—“Father-of-thee, age-of-him.” Or, to be more
accurate, as the syntax follows that of gesture-language in placing the
predicate before the subject, we should translate the proposition into
its most exact equivalent by saying, “His age, thy father.” Similarly,
if it is required to make such a statement as that “He is wearing a
white jacket,” the form of the statement would be, “He-with-white
with-jacket,” or, as we might perhaps more tersely translate it, “He
jackety whitey.”[255]

Again, in Feejee language the functions of a verb may be discharged
by a noun in construction with an oblique pronominal suffix, _e.g._,
_loma-qu_ = heart or will-of-me, = I will.[256]

So likewise, “almost all philologists who have paid attention to the
Polynesian languages, concur in observing that the divisions of parts
of speech received by European grammarians are, as far as external form
is concerned, inapplicable, or nearly so, to this particular class. The
same element is admitted to be indifferently substantive, adjective,
verb, or particle.”[257] “I will eat the rice,” would require to be
rendered, “The-eating-of-me-the-rice = My eating will be of the rice.”
“The supposed verb is, in fact, an abstract noun, including in it the
notion of futurity of time in construction with an oblique pronominal
suffix; and the ostensible object of the action is not a regimen in the
accusative case, but an apposition. It is scarcely necessary to say
how irreconcilable this is with the ordinary grammatical definition of
a transitive verb; and that, too, in a construction where we should
expect that true verbs would be infallibly employed, if any existed
in the language.”[258] And, not to overburden the argument with
illustrations, it will be enough to add with this writer, “there can be
no question that nouns in conjunction with oblique cases of pronouns
may be, and, in fact, are employed as verbs. Some of the constructions
above specified admit of no other analysis; and they are no accidental
partial phenomena, but capable of being produced by thousands.”[259]

       *       *       *       *       *

It would be easy to multiply quotations from other authorities to the
same effect; but these, I think, are enough to show how completely the
philology of predication destroys the philosophy of predication, as
this has been presented by my opponents. Not only, as already shown,
have they been misled by the verbal accident of certain languages with
which they happen to be familiar identifying the copula with the verb
“to be” (which itself, as we have also seen, has no existence in many
languages); but, as we now see, their analysis is equally at fault
where it deals with the subject and predicate. Such a fully elaborated
form of proposition as “A negro is black,” far from presenting “the
simplest element of thought,” is the demonstrable outcome of an
enormously prolonged course of mental evolution; and I do not know a
more melancholy instance of ingenuity misapplied than is furnished
by the arguments previously quoted from such writers, who, ignoring
all that we now know touching the history of predication, seek to
show that an act of predication is at once “the simplest element of
thought,” and so hugely elaborate a process as they endeavour to
represent. The futility of such an argument may be compared with that
of a morphologist who should be foolish enough to represent that the
Vertebrata can never have descended from the Protozoa, and maintain
his thesis by ignoring all the intermediate animals which are known
actually to exist.

Take an instance from among the quotations previously given. It will be
remembered that the challenge which my opponents have thrown down upon
the grounds of logic and psychology, is to produce the brute which “can
furnish the blank form of a judgment—the ‘is’ in ‘A is B.’”[260]

Now, I cannot indeed produce a brute that is able to supply such a
form; but I have done what is very much more to the purpose: I have
produced many nations of still existing men, in multitudes that
cannot be numbered, who are as incapable as any brute of supplying
the blank form that is required. Where is the “is,” in “Age-of-him
Father-of-thee” = “His-age-thy-father” = “Thy-father-is-old”? Or, in
still more primitive stages of human utterance, how shall we extract
the blank form of predication from a “sentence-word,” where there
is not only an absence of any copula, but also an absence of any
differentiation between the subject and the predicate? The truth, in
short, is, as now so repeatedly shown, that not only the brute, but
likewise the young child—and not only the young child, but likewise
early man—and not only early man, but likewise savage man—are all and
equally unable to furnish the blank form of predication, as this has
been slowly elaborated in the highest ramifications of the human mind.

Of course all this futile (because erroneous) argument on the part
of my opponents, rests upon the analysis of the proposition as this
was given in the Aristotelian system of logic—an analysis which,
in turn, depends on the grammar of the Greek language. Now, it goes
without saying that the whole of this system is obsolete, so far as any
question of the _origin_ either of thought or of speech is concerned.
I do not doubt the value of this grammatical study, nor of the logic
which is founded upon it, provided that inferences from both are kept
within their legitimate sphere. But at this time of day to regard as
primitive the mode of predication which obtained in so highly evolved a
language as the Greek, or to represent the “categories” of Aristotle’s
system as expressive of the simplest elements of human thought, appears
to me so absurd that I can only wonder how intelligent men can have
committed themselves to such a line of argument.[261]

Quitting, then, all these old-world fallacies which were based on an
absence of information, we must accept the analysis of predication
as this has been supplied to us by the advance of science. And this
analysis has proved to demonstration, that “the division of the
sentence into two parts, the subject and the predicate, is a mere
accident; it is not known to the polysynthetic languages of America,
which herein reflect the condition of primeval speech.... So far as
the act of thought is concerned, subject and predicate are one and the
same, and there are many languages in which they are so treated.”[262]
Consequently, it appears to me that the only position which remains for
my opponents to adopt is that of arguing in some such way as follows.

Freely admitting, they may say, that the issue must be thrown back
from predication as it occurs in Greek to predication as it occurs in
savage languages of low development, still we are in the presence of
predication all the same. And even when you have driven us back to the
most primitive possible form of human speech, wherein as yet there are
no parts of speech, and predication therefore requires to be conducted
in a most inefficient manner, still most obviously it _is_ conducted,
inasmuch as it is only for the purpose of conducting it that speech can
have ever come into existence at all.

Now, in order to meet this sole remaining position, I must begin by
reminding the reader of some of the points which have already been
established in previous chapters.

First of all, when seeking to define “the simplest element of thought,”
I showed that this does not occur in the fully formed proposition,
but in the fully formed concept; and that it is only out of two such
concepts as elements that full or conceptual propositions can be
formed as compounds. Or, as this was stated in the chapter on Speech,
“conceptual names are the ingredients out of which is formed the
structure of propositions; and, in order that this formation should
take place, there must be in the ingredients that element of conceptual
ideation which is already present in every denominative term.” Or, yet
again, as the same thing was there quoted from Professor Sayce, “it is
a truism of psychology that the terms of a proposition, when closely
interrogated, turn out to be nothing but abbreviated judgments.”[263]

Having thus defined the simplest element of thought as a concept, I
went on to show from the psychogenesis of children, that before there
is any power of forming concepts—and therefore of bestowing names as
denominative terms, or, _a fortiori_, of combining such terms in the
form of conceptual propositions—there is the power of forming recepts,
of naming these recepts by denotative terms, and even of placing such
terms in apposition for the purpose of conveying information of a
pre-conceptual kind. The pre-conceptual, rudimentary, or unthinking
propositions thus formed occur in early childhood, prior to the advent
of self-consciousness, _and prior, therefore, to the very condition
which is required for any process of conceptual thought_. Moreover, it
was shown that this pre-conceptual kind of predication is itself the
product of a gradual development. Taking its origin from the ground of
gesture-signs, when it first begins to sprout into articulate utterance
there is absolutely no distinction to be observed between “parts of
speech.” Every word is what we now know as a “sentence-word,” any
special applications of which can only be defined by gesture. Next,
these sentence-words, or others that are afterwards acquired, begin
to be imperfectly differentiated into denotative names of objects,
qualities, actions, and states; and the greater the definition which
they thus acquire as parts of speech, the more do they severally
undergo that process of connotative extension as to meaning which is
everywhere the index of a growing appreciation of analogies. Lastly,
object-words and attributive-words (_i.e._ denotative names of things
and denotative names of qualities or actions), come to be used in
apposition. But the rudimentary or unthinking form of predication
which results from this is due to merely sensuous associations and
the external “logic of events;” like the elements of which it is
composed, it is not conceptual, but pre-conceptual. With the dawn
of self-consciousness, however, predication begins to become truly
conceptual; and thus enters upon its prolonged course of still gradual
development in the region of introspective thought.

All these general facts, it will be remembered, were established on
grounds of psychological observation alone; I nowhere invoked the
independent witness of philology. But the time having now come for
calling in this additional testimony, the corroborating force of it
appears to me overwhelming. For it everywhere proves the growth of
predication to have been the same in the race as we have found it to be
in the individual. Therefore, as in the latter case, so in the former,
I now ask—Will any opponent venture to affirm that pre-conceptual
ideation is indicative of judgment? Or, which is the same thing, will
he venture to deny that there is an all-important distinction between
predication as receptual and predication as conceptual? Will he still
seek to take refuge in the only position now remaining, and argue, as
above supposed, that not only in the childish appositions of denotative
names, but even in the earlier and hitherto undifferentiated protoplasm
of a “sentence-word,” we have that faculty of predication on which he
founds his distinction between man and brute? Obviously, if he will not
do this, his argument is at an end, seeing that in the race, as in the
individual, there is now no longer any question as to the continuity
between the predicative germ in a sentence-word, and the fully evolved
structure of a formal proposition. On the other hand, if he does elect
to argue thus, the following brief considerations will effectually
dislodge him.

If the term “predication” is extended from a conceptual proposition
to a sentence-word, it thereby becomes deprived of that distinctive
meaning upon which alone the whole argument of my opponents is reared.
For, when used by a young child (or primitive man), sentence-words
require to be supplemented by gesture-signs in order to particularize
their meaning, or to complete the “predication.” But, where such is
the case, there is no longer any psychological distinction between
_speaking_ and _pointing_: if this is called predication, then the
predicative “category of language” has become identified with the
indicative: man and brute are conceded to be “brothers.”

Take an example. At the present moment I happen to have an infant who
has not yet acquired the use of any one articulate word. Being just
able to toddle, he occasionally comes to grief in one way or another;
and when he does so he seeks to communicate the nature of his mishap
by means of gesture-signs. To-day, for instance, he knocked his head
against a table, and forthwith ran up to me for sympathy. On my asking
him where he was hurt, he immediately touched the part of his head in
question—_i.e._ _indicated_ the painful spot. Now, will it be said
that in doing this the child was _predicating_ the seat of injury? If
so, all the distinctive meaning which belongs to the term predicating,
or the only meaning on which my opponents have hitherto relied, is
discharged. The gesture-signs which are so abundantly employed by the
lower animals would then also require to be regarded as predicatory,
seeing that, as before shown at considerable length, they differ in no
respect from those of the still speechless infant.

Therefore, whether my opponents allow or disallow the quality of
predication to sentence-words, alike and equally this argument
collapses. Their only logical alternative is to vacate their argument
altogether; no longer to maintain that “Speech is the Rubicon of Mind,”
but to concede that, as between the indicative phase of language which
we share with the lower animals, and the truly predicative phase which
belongs only to man, there is no distinction of kind to be attributed;
seeing that, on the contrary, whether we look to the psychogenesis of
the individual or to that of the race, we alike find a demonstrable
continuity of evolution from the lowest to the highest level of the
sign-making faculty.




CHAPTER XV.

THE WITNESS OF PHILOLOGY (_continued_).


In the last chapter we have been concerned with the philology of
predication. In the present chapter I propose to consider the philology
of conception. Of course the distinction is not one that can be very
sharply drawn, because, as fully shown in my chapter on Speech, every
concept embodies a judgment, and therefore every denominative term is a
condensed proposition. Nevertheless, as my opponents have laid so much
stress on full or formal predication, as distinguished from conception,
I have thought it desirable, as much as possible, to keep these two
branches of our subject separate. Therefore, having now disposed of
all opposition that can possibly be raised on the ground of formal
predication, I will conclude by throwing the light of philology on the
origin of material predication, or the passage of receptual denotation
into conceptual denomination, as this is shown to have occurred in the
pre-historic evolution of the race.

It will be remembered that, under my analysis of the growth of
predication, much more stress has been laid in the last chapter than in
previous chapters on what I have called the protoplasm of predication
as this occurs in the hitherto undifferentiated “sentence-word.”
While treating of the psychology of predication in the chapter on
Speech, I did not go further back in my analysis than to point out
how the “nascent” or “pre-conceptual” propositions of young children
are brought about by the mere apposition of denotative terms—such
apposition having been shown to be due to sensuous association when
under the guidance of the “logic of events.” But when I came to deal
with the philology of predication, it became evident that there
was even an earlier phase of the faculty in question than that of
apposing denotative terms by sensuous association. For, as we have so
recently seen, philologists have proved that even before there were
any denotative terms respectively significant of objects, qualities,
actions, states, or relations, there were sentence-words which combined
in one vague mass the meanings afterwards apportioned to substantives,
adjectives, verbs, prepositions, &c., with the consequence that the
only kind of apposition which could be called into play for the purpose
of indicating the particular significance intended to belong to such
a word on particular occasions, was the apposition of gesture-signs.
Now, I had two reasons for thus postponing our consideration of what
is undoubtedly the earliest phase of articulate sign-making. In the
first place, it seemed to me that I might more easily lead the reader
to a clear understanding of the subject by beginning with a phase of
predication which he could most readily appreciate, than by suddenly
bringing him into the presence of a germ-like origin which is far
from being so readily intelligible. But over and above this desire
to proceed from the familiar to the unfamiliar, I had, in the second
place, a further and a better reason for not dealing with the ultimate
germ of articulate sign-making so long as I was dealing only with the
psychology of our subject. This reason was, that in the development of
speech as exhibited by the growing child—which, of course, furnishes
our only material for a study of the subject from a psychological point
of view—the original or germinal phase in question does not appear
to be either so marked, so important, or, comparatively speaking, of
such prolonged duration as it was in the development of speech in the
race. To use biological terms, this the earliest phase in the evolution
of speech has been greatly foreshortened in the ontogeny of mankind,
as compared with what it appears to have been in the phylogeny. The
result, of course, is that we should gain but an inadequate idea of
its importance, were we to estimate it by a merely psychological
analysis of what we now find in the life-history of the individual.

It is perfectly true, as Professor Max Müller says, that “if an
English child says ‘Up,’ that _up_ is, to his mind, noun, verb, and
adjective, all in one.” Nevertheless, in a young child, from the very
first, there is a marked tendency to observe the distinctions which
belong to the principal parts of speech. The earliest words uttered
by my own children have always been nouns and proper names, such as
“Star,” “Mamma,” “Papa,” “Ilda,” &c.; and although, later on, some of
these earliest words might assume the functions of adjectives by being
used in apposition with other nouns subsequently acquired (such as
“Mamma-ba,” for a sheep, and “Ilda-ba” for a lamb), neither the nouns
nor the adjectives came to be used as verbs. It has been previously
shown that the use of adjectives is acquired almost as soon as that
of substantives; and although the poverty of the child’s vocabulary
then often necessitates the adjectives being used as substantives,
the substantives as adjectives, and both as rudimentary propositions,
still there remains a distinction between them as object-words and
quality-words. Similarly, although action-words and condition-words
are often forced into the position of object-words and quality-words,
it is apparent that the primary idea attaching to them is that which
properly belongs to a verb. And, of course, the same remarks apply to
relation-words, such as “Up.”

Take, for instance, the cases of pre-conceptual predication which
were previously quoted from Mr. Sully, namely, “Bow-wow” = “That is
a dog;” “Ot” = “This milk is hot;” “Dow” = “My plaything is down;”
“Dit ki” = “Sister is crying;” “Dit naughty” = “Sister is naughty;”
“Dit dow ga” = “Sister is down on the grass.” In all these cases it is
evident that the child is displaying a true perception of the different
functions which severally belong to the different parts of speech; and
so far as psychological analysis alone could carry us, there would
be nothing to show that the forcing of one part of speech into the
office of another, which so frequently occurs at this age, is due to
anything more than the exigencies of expression where as yet there
are scarcely any words for the conveyance of meaning of any kind.
Therefore, on grounds of psychological analysis alone, I do not see
that we are justified in arguing from these facts that a young child
has no appreciation of the difference between the functions of the
different parts of speech—any more than we should were we to argue
that a grown man has no such appreciation when he extends the meaning
of a substantive (such as “pocket”) so as to embrace the function of
an adjective on the one hand (_e.g._ “pocket-book”), and of a verb on
the other (_e.g._ “he _cannoned_ off the white, and _pocketed_ the
red”). What may be termed this grammatical abuse of words becomes an
absolute necessity where the vocabulary is small, as we well know when
trying to express ourselves in a foreign language with which we are
but slightly acquainted. And, of course, the smaller the vocabulary,
the greater is such necessity; so that it is greatest of all when an
infant is only just emerging from its infancy. Therefore, as just
remarked, on grounds of psychological analysis alone, I do not think
we should be justified in concluding that the first-speaking child has
no appreciation of what we understand by parts of speech; and it is on
account of the uncertainty which here obtains as between necessity and
incapacity, that I reserved my consideration of “sentence-words” for
the independent light which has been thrown upon them by the science of
comparative philology.

Now, when investigated by this light, it appears, as already
observed, that the protoplasmic condition of language prior to its
differentiation into parts of speech was of much longer duration in
the race than, relatively speaking, it is in the individual. Moreover,
it appears to have been of relatively much greater importance to the
subsequent development of language. How, then, is this difference
to be explained? I think the explanation is sufficiently simple. An
infant of to-day is born into the medium of already-spoken language;
and long before it is itself able to imitate the words which it hears,
it is well able to understand a large number of them. Consequently,
while still literally an _infant_, the use of grammatical forms is
being constantly borne in upon its mind; and, therefore, it is not at
all surprising that, when it first begins to use articulate signs, it
should already be in possession of some amount of knowledge of their
distinctive meanings as names of objects, qualities, actions, states,
or relations. Indeed, it is only as such that the infant has acquired
its knowledge of these signs at all; and hence, if there is any wonder
in the matter, it is that the first-speaking child should exhibit so
much vagueness as it does in the matter of grammatical distinction.

But how vastly different must have been the case of primitive man!
The infant, as a child of to-day, finds a grammar already made to its
use, and one which it is bound to learn with the first learning of
denotative names. But the infant, as an adult in primeval time, was
under the necessity of slowly elaborating his grammar together with his
denotative names; and this, as we have previously seen, he only could
do by the aid of gesture and grimace. Therefore, while the acquisition
of names and forms of speech by infantile man must have been thus in
chief part dependent on gesture and grimace, the acquisition by the
infantile child is now not only independent of gesture and grimace, but
actively inimical to both. The already-constructed grammar of speech
is the evolutionary substitute of gesture, from which it originally
arose; and, hence, so soon as a child of to-day begins to speak,
gesture-signs begin at once to be starved out by grammatical forms. But
in the history of the race gesture-signs were the nursing-mothers of
grammatical forms; and the more that their progeny grew, the greater
must have been the variety of functions which the parents were called
upon to perform. In other words, during the infancy of our race the
growth of articulate language must not only have depended, but also
reacted upon that of gesture-signs—increasing their number, their
intricacy, and their refinement, up to the time when grammatical forms
were sufficiently far evolved to admit of the gesture-signs becoming
gradually dispensed with. Then, of course, Saturn-like, gesticulation
was devoured by its own offspring; the relations between signs
appealing to the eye and to the ear became gradually reversed; and,
as is now the case with every growing child, the language of formal
utterance sapped the life of its more informal progenitor.

We are now in a position to consider the exact psychological relation
of sentence-words to denotative and receptually connotative words. It
will be remembered that I have everywhere spoken of sentence-words as
representing an even more primitive order of ideation than denotative
words, and, _a fortiori_, than receptually connotative words. On the
other hand, in earlier parts of this treatise I showed that both the
last-mentioned kinds of words occur in children when they first begin
to speak, and may even be traced so low down in the psychological scale
as the talking birds. This apparent ambiguity, therefore, now requires
to be cleared up. Can anything, it may be reasonably asked, in the
shape of spoken language be more primitive than the very first words
which are spoken by a child, or even by a parrot? But, if not, how can
I agree with those philologists who conclude that there is an even
still more primitive stage of conceptual evolution to be recognized in
sentence-words?

Briefly, my answer to these questions is that in the young child
and the talking bird denotative-words, connotative-words, and
sentence-words are all equally primitive; or, if there is any priority
to be assigned, that it must be assigned to the first-named. But the
reason of this, I hold to be, is, that the child and the bird are
both living in an already-developed medium of spoken language, and,
therefore, as recently stated, have only to learn their denotative
names by special association, while primitive man had himself to
fashion his names out of the previously inarticulate materials of his
own psychology. Now this, as we have also seen, he only could do by
such associations of sounds and gestures as in the first instance
must have conveyed meanings of a pre-conceptually predicative kind.
In the absence of any sounds already given—and therefore already
_agreed upon_—as denotative names, there could be no possibility of
primitive man arbitrarily _assigning_ such names; and thus there could
have been no parallel to a young child who receptually _acquires_
them. In order that he should assign names, primitive man must first
have had occasion to make his pre-conceptual statements about the
objects, qualities, &c., the names of which afterwards grew out of
these statements, or sentence-words. Adam, indeed, gave names to
animals; but Adam was already in possession of conceptual thought, and
therefore in a psychological position to appreciate the importance of
what he was about. But the “pre-Adamite man” who is now before us could
not possibly have invented names for their own sakes, unless he were
already capable of thinking about names _as_ names, and, therefore,
already in possession of that very conceptual thought which, as we
have now so often seen, depends upon names for its origin. Even with
all our own fully developed powers of conceptual thought, we cannot
_name_ an object when in the society of men with whose language we
are totally unacquainted, without _predicating_ something about that
object by means of gestures or other signs. Therefore, without further
discussion, it must be obvious—not only, as already shown, that there
is here no exact parallel between ontogenesis and phylogenesis, and
that we have thus a full explanation why sentence-words were of so much
more importance to the infant man than they are to the infant child,
but further and consequently—that the question whether sentence-words
are more primitive than denotative words is not a question that
is properly stated, unless it be also stated whether the question
applies to the individual or to the race. As regards the individual
of to-day, it cannot be said that there is any priority, historical
or psychological, of sentence-words over denotative words, or even
over receptually connotative words of a low order of extension. Nay,
we have seen that the leading principles of grammatical form admit of
being acquired by the child together with his acquisition of words of
all kinds, and that even talking birds are able to distinguish between
names as severally names of objects, qualities, states, or actions.

Thus we find that to almost any order of intelligence which is already
surrounded by the medium of spoken language, the understanding—and, in
the presence of any power of imitative utterance, the acquisition—of
denotative names as signs or marks of corresponding objects, qualities,
&c., is, if anything, a more primitive act than that of using a
sentence-word; but that in the absence of such an already-existing
medium, sentence-words are more primitive than denotative names.
Nevertheless, it is of importance to note how low an order of
receptual ideation is capable of learning a denotative name by special
association, because this fact proves that as soon as mankind advanced
to the stage where they first began to coin their sentence-words, they
must already have been far above the psychological level required
for the acquisition of denotative words, _if only such words had
previously been in existence_. Consequently, we can well understand
how such words would soon have begun to come into existence through
the habitual employment of sentence-words in relation to particular
objects, qualities, states, actions, &c.; by such special associations,
sentence-words would readily degenerate into merely semiotic marks.
How long or how short a time this genesis of relatively “empty words”
out of the primordially “full words” may have occupied, it is now
impossible to say; but the important thing for us to notice is, that
during the whole of this time—whatever it may have been—the mind of
primitive man was already far above the psychological level which is
required for the apprehension of a denotative name.[264]

So much, then, for the first class of considerations which has been
opened up by throwing upon the results of our psychological analysis
the independent light of philological research. I will now pass on to a
second class, which is even of more importance.

From the fact that sentence-words played so all-important a part in the
origin of speech, and that in order to do so they essentially depended
on the co-operation of gestures with which they were accompanied, so
that in the resulting “complex of sound and gesture the sound had
no meaning apart from the gesture;” from these now well-established
facts, we may gain some additional light on a question previously
considered—namely, the extent to which primitive words were “abstract”
or “concrete,” “particular” or “general,” and, therefore, “receptual”
or “conceptual.” According to Professor Max Müller, “the science of
language has proved by irrefragable evidence that human thought, in the
true sense of that word—that is, human language—did not proceed from
the concrete to the abstract, but from the abstract to the concrete.
Roots, the elements out of which all language has been constructed,
are abstract, never concrete; and it is by predicating these abstract
concepts of this or that, by localizing them here or there, in fact by
applying the category of οὐϚία or substance, to the roots, that the
first foundation of our language and our thought were laid.”[265]

Here, to begin with, there is an inherent contradiction. When it is
said that the roots in question already presented abstract concepts, it
becomes a contradiction to add that “the first foundations of language
and thought were laid by applying the category of substance to the
roots.” For, if these roots already presented abstract concepts, they
already presented the distinctive feature of human “thought,” whose
“foundations,” therefore, must have been “laid” somewhere further back
in the history of mankind. But, besides this inherent contradiction, we
have here an emphatic re-statement of the two radical errors which I
previously mentioned, and which everywhere mar the philosophical value
of Professor Max Müller’s work. The first is his tacit assumption that
the roots of Aryan speech represent the original elements of articulate
language. The second is that, upon the basis of this assumption, the
science of language has proved, by irrefragable evidence, that human
thought proceeded from the abstract to the concrete—or, in other
words, that it sprang into being Minerva-like, already equipped with
the divine inheritance of conceptual wisdom. Now, in entertaining this
theory, Professor Max Müller is not only in direct conflict with all
his philological brethren, but likewise, as we have previously seen,
often compelled to be irreconcilably inconsistent with himself.[266]
Moreover, as we have likewise seen, his assumption as to the aboriginal
nature of Aryan roots, on which his transcendental doctrine rests, is
intrinsically absurd, and thus does not really require the united voice
of professed philologists for its condemnation. Therefore, what the
science of language _does_ prove “by irrefragable evidence” is, _not_
that these roots of the Aryan branch of language are the aboriginal
elements of human speech, or indices of the aboriginal condition of
human ideation; but that, being the survivals of incalculably more
primitive and immeasurably more remote phases of word-formation,
they come before us as the already-matured products of conceptual
thought—and, _a fortiori_, that on the basis of these roots alone _the
science of language has absolutely no evidence at all to furnish_ as
touching the matter which Professor Max Müller here alludes to in such
positive terms. In this connection there can be no possible escape from
the tersely expressed conclusion previously quoted from Geiger, and
unanimously entertained as an axiom by philologists in general:—“These
roots are not the primitive roots: we have perhaps in no one single
instance the first aboriginal articulate sound—just as little, of
course, the aboriginal signification.”[267]

But the point which I now wish to bring forward is this. We have
previously seen the source of these unfortunate utterances in Professor
Max Müller’s philology appears to reside in certain prepossessions
which he exhibits in the domain of psychology. For he adopts the
assumption that there can be no order of words which do not, by the
mere fact of their existence, imply concepts: he does not sufficiently
recognize that there may be a power of bestowing names as signs,
without the power of thinking these signs as names. Consequently, the
distinction which, on grounds of comparative psychology, appears to me
so obvious and so necessary—_i.e._ between names as merely denotative
marks due to pre-conceptual association, and denominative judgments due
to conceptual thought—has escaped his sufficient notice. Consequently,
also, he has failed to distinguish between ideas as “general” and what
I have called “generic;” or between an idea that is general because
it is born of an intentional synthesis of the results of a previous
analysis, and an idea that is _generalized_[268] because not yet
differentiated by any intentional analysis, and therefore representing
simply an absence of conceptual thought. My child on first beginning
to speak had a generalized idea of similarity between all kinds of
brightly shining objects, and therefore called them all by the one
denotative name of “star.” The astronomer has a general idea answering
to his denominative name of “star;” but this has been arrived at after
a prolonged course of mental evolution, wherein conceptual analysis
has been engaged in conceptual classification in many and various
directions: it therefore represents the psychological antithesis of the
generalized idea, which was due to the merely sensuous associations
of pre-conceptual thought. Ideas, then, as general and as generic
severally occupy the very antipodes of Mind.

All this we have previously seen. My object in here recurring to the
matter is to show that much additional light may be thrown upon it by
the philological doctrine of “sentence-words,” which Professor Max
Müller, in common with other philologists, fully accepts.

Of all the writers on primitive modes of speech as represented by
existing savages, no one is entitled to speak with so much authority
as Bleek. Now, as a result of his prolonged and first-hand study of
the subject, he is strongly of opinion that aboriginal words were
expressive “not at all of an abstract or general character, but
exclusively concrete or individual.” By this he means that primitive
ideas were what I have called generic. For he says that had a word been
formed from imitation of the sound of a cuckoo, for instance, it could
not possibly have had its meaning limited to the name of that bird;
but would have been extended so as to embrace “the whole situation
so far as it came within the consciousness of the speaker.” That is
to say, it would have become a generic name for the whole recept of
bird, cry, flying, &c., &c., just as to our own children the word
_Ba_=sheep, bleating, grazing, &c. Now, this process of comprising
under one denotative term the hitherto undifferentiated perceptions of
“a whole situation so far as it comes within the consciousness of the
speaker,” is the very opposite of the process whereby a denominative
term is brought to unify, by an act of “generalization,” the previously
well-differentiated concepts between which some analogy is afterwards
discovered. Therefore the absence of any parts of speech in primitive
language is due to a generic order of ideation, whereas the unions
of parts of speech in any languages which present them is due to the
generalizing order of ideation. Or, as Bleek puts it while speaking
of the comparatively undifferentiated condition of South African
languages, “this differs entirely from the principle which prevails in
modern English, where a word, without undergoing any change of form,
may nevertheless belong to different parts of speech. For in English
the parts of speech, though not always differing in sound, are always
accurately distinguished in concept; while in the other case there
was as yet no consciousness of any difference, inasmuch as neither
form nor position had hitherto called attention to anything of the
kind. For forms had not yet made their appearance, and determinate
position [_i.e._ significance expressed by syntax], as, for example,
in Chinese, could only arise in a language of highly advanced internal
formation.”[269]

Indeed, if we consider the matter, it is not conceivable that the case
could be otherwise. No one will maintain that the sentence-words of
young children exhibit the highest elaborations of conceptual thought,
on the ground that they present the highest degree of “generality”
which it is possible for articulate sounds to express. But if this
is not to be suggested as regards the infant child, what possible
ground can there be for suggesting it as regards the infant man, or
for inferring that aboriginal speech must have been expressive of
“general” and “abstract” ideas, merely because the further backwards
that we trace the growth of language the less organized do we find its
structure to be? Clearly, the contradiction arises from a confusion
between ideas as generic and general, or between the extension which
is due to original vagueness and that which is laboriously acquired by
subsequent precision. An Amœba is morphologically more “generalized”
than a Vertebrate; but for this very reason it is the less highly
evolved as an organism. The philology of sentence-words, therefore,
leads us back to a state of ideation wherein as yet the powers of
conceptual thought were in that nascent condition which betokens what
I have called their pre-conceptual stage—or a stage which may be
observed in a comparatively foreshortened state among children before
the dawn of self-consciousness.

There can be no reasonable doubt that during this stage of mental
evolution sentence-words arose in the race as they now do in the
individual, the only difference being that then they had to be invented
instead of learnt. This difference would probably have given a larger
importance to the principle of onomatopœia,[270] and certainly a
much larger importance to the co-operation of gesture, than now
obtains in the otherwise analogous case of young children. But in
the one case as in the other, I think there can be no reasonable
question that sentence-words must have owed their origin to receptual
and pre-conceptual apprehensions of all kinds, whether of objects,
qualities, actions, states, relations, or of any two or more of these
“categories” as they may happen to have been blended in the hitherto
undifferentiating perceptions of aboriginal man.

       *       *       *       *       *

I must now allude to the results of our previous inquiry touching
“the syntax of gesture-language.” For comparison will show that in
all essential particulars the semiotic construction of this the
most original and immediately graphic mode of communication, bears
a striking resemblance to that which is presented by the earliest
forms of articulate language, both as revealed by philology and in
“baby-talk.”[271] Thus, as we saw, “gesture-language has no grammar
properly so called. The same sign stands for ‘walk,’ ‘walkest,’
‘walking,’ ‘walked,’ ‘walker.’ Adjectives and verbs are not easily
distinguished by the deaf and dumb. Indeed, our elaborate system of
parts of speech is but little applicable to the gesture-language.”
Next, to quote again only one of the numerous examples previously
given to show the primitive order of apposition, whereby the language
of gesture serves to convey a predication, “I should be punished if
I were lazy and naughty” would be put, “I lazy, naughty, no!—lazy,
naughty, I punished; yes!” Again, “to make is too abstract for the
deaf-mute; to show that the tailor makes the coat, or that the
carpenter makes the table, he would represent the tailor sewing the
coat and the carpenter sawing and planing the table. Such a proposition
as ‘Rain makes the land fruitful’ would not come into his way of
thinking: ‘Rain, fall; plants, grow,’ would be his pictorial (_i.e._
receptual) expression.” Elsewhere this writer remarks that the absence
of any distinction between substantive, adjective, and verb, which
is universal in gesture-language, is customary in Chinese, and not
unknown even in English. “To _butter_ bread, to _cudgel_ a man, to
_oil_ machinery, to _pepper_ a dish, and scores of such expressions,
involve action and instrument in one word, and that word a substantive
treated as the root or crude form of a verb. Such expressions are
concretisms, picture-words, gesture-words, as much as the deaf-and-dumb
man’s one sign for ‘butter’ and ‘buttering.’” And similarly as to the
substantive-adjective, in such words as _iron-stone_, _feather-grass_,
_chesnut-horse_, &c.; here the mere apposition of the words
constitutes the one an attribution of the other, as is the case in
gesture-language. And not only in Chinese, but as shown in the last
chapter, in a great number and variety of savage tongues this mode of
construction is habitual. In all these cases distinctions between parts
of speech can be rendered only by syntax; and this syntax is the syntax
of gesture.

I will ask the reader to refer to the whole passage in which
I previously treated of the syntax of gesture,[272] giving
special attention to the points just noted, and also to the
following:—invariable absence of the copula, and frequent absence
of the verb (as “Apple-father-I” = “My father gave me an apple”);
resemblance of sentences to the polysynthetic or unanalyzing type (as
“I-Tom-struck-a-stick” = “Tom struck me with a stick”); the device
whereby syntax, or order of apposition, is made to distinguish between
predicative, attributive, and possessive meanings, and therefore
also between substantives and adjectives; the importance of grimace
in association with gesture (as when a look of inquiry converts an
assertion into a question); the highly instructive means whereby
relational words, and especially pronouns, are rendered in the gestures
of pointing; the no less instructive manner whereby a general idea is
rendered in a summation of particular ideas (as “Did you have soup? did
you have porridge?” &c. = “What did you have for dinner?”); and the
receptual or sensuous source of all gesture-signs which are concerned
in expressing ideas presenting any degree of abstraction (as striking
the hand to signify “hard,” &c.).

Hence, we may everywhere trace a fundamental similarity between the
comparatively undeveloped form of conceptual thought as displayed
in gesture, and that which philology has revealed as distinctive of
early speech. Of course in both cases conceptual thought is there:
the ideation is human, though, comparatively speaking, immature.
But the important point to notice is the curiously close similarity
between the forms of language-structure as revealed in gesture and in
early speech. For no one, I should suppose, can avoid perceiving the
idiographic character of gesture-language, whereby it is more nearly
allied to the purely receptual modes of communication which we have
studied in the lower animals, than is the case with our fully evolved
forms of predication. It therefore seems to me highly suggestive that
the earliest forms and records of spoken language that we possess
(notwithstanding that they are still far from aboriginal), follow so
closely the model which is still supplied to us in the idiographic
gestures of deaf-mutes. Such syntax as there is—_i.e._ such _a putting
in order_ as is expressive of the mode of ideational grouping—so
nearly resembles the syntax of gesture-language, that we can at once
perceive their common psychological source. It is on account of this
structural resemblance between gesture and early speech that I have
devoted so much space to our consideration of the former; and if I do
not now dwell at greater length upon the significance of the analogy,
it is only because this significance appears too obvious to require
further treatment.

There is, however, one point with reference to this analogy on which
a few words must here be said. If there is any truth at all in the
theory of evolution with reference to the human mind, we may be quite
sure, from what has been said in earlier chapters, that tone, gesture,
and grimace preceded articulation as the medium of pre-conceptual
utterance. Therefore, the structural similarity between existing
gesture-language and the earliest records of articulate language now
under consideration, is presumably due, not only to a similarity of
psychological conditions, but also to direct continuity of descent.
Or, as Colonel Mallery well puts it, while speaking of the presumable
origin of spoken language, “as the action was then the essential, and
the consequent or concomitant sound the accident, it would be expected
that a representation, or feigned reproduction of the action, would
have been used to express the idea before the sound associated with
that action could have been separated from it. The visual onomatopœia
of gestures, which even yet have been subjected to but slight
artificial corruption, would therefore serve as a key to the audible.
It is also contended that in the pristine days, when the sounds of
the only words yet formed had close connection with objects and the
ideas directly derived from them, signs were as much more copious for
communication than speech as the sight embraces more and more distinct
characteristics of objects than does the sense of hearing.”[273]

       *       *       *       *       *

All the foregoing and general conclusions thus reached, touching the
genesis of conceptual from pre-conceptual ideation, admit of being
strikingly corroborated through another line of philological research.
On antecedent grounds the evolutionist would suppose that “the first
language-signs must have denoted those physical acts and qualities
which were directly apprehensible by the senses; both because these
alone are directly significable, and because it was only they that
untrained human beings had the power to deal with or the occasion to
use.”[274] In other words, if, as we suppose, language had its origin
in merely denotative sign-making, which gradually became more and more
connotative and thus gradually more and more predicative; obviously the
original denotations must have referred only to objects (or actions,
states, and qualities) of merely receptual significance—_i.e._
“those physical acts and qualities which are directly apprehensible
by the senses.” And, no less obviously, the connotative extension
of such denotative names must, for an enormously long period, have
been confined to a pre-conceptual cognizance of the most obvious
analogies—_i.e._ such analogies as would necessarily thrust themselves
upon the merely sensuous perception by the force of direct association.

Now, if this were the case, what would the evolutionist expect to find
in language as it now exists? Clearly, he would expect to find more
or less well-marked traces, in the fundamental constitution of all
languages, of what has been called “fundamental metaphor”—by which
is meant an intellectual extension of terms that originally were of
no more than sensuous signification. And this is precisely what we do
find. “The whole history of language, down to our own day, is full
of examples of the reduction of physical terms and phrases to the
expression of non-physical conceptions and relations; we can hardly
write a line without giving illustrations of this kind of linguistic
growth. So pervading is it, that we never regard ourselves as having
read the history of any intellectual or moral term till we have traced
it back to its physical origin.”[275]

Now, I hold that this receptual nucleus of all our conceptual terms
furnishes the strongest possible evidence, not only of the historical
priority of the former, but also of what Professor Max Müller
calls their “dire necessity” to the growth of the latter.[276] In
other words, the facts appear conclusively to show that conceptual
connotation (denomination) has always had—_and can only have had_—a
receptual core (denotation) around which to develop. Psychological
analysis has already shown us the psychological priority of the recept;
and now philological research most strikingly corroborates this
analysis by _actually finding the recept in the body of every concept_.

How this large and general fact is to be met by my antagonists I know
not. It certainly does not satisfy the case to say, with Professor Max
Müller,[277] Noiré,[278] and those who think with them, that in no
other way could the growth of conceptual thought have been possible;
for this is merely to reiterate on _a priori_ grounds the conclusion
which I have reached _a posteriori_. And the more that this historical
priority of denotation can thus be shown an _a priori_ necessity to the
subsequent genesis of denomination, the greater becomes the cogency
of our evidence _a posteriori_ that, as a matter of fact, such has
been invariably the order of historical succession. For, if conceptual
ideation differs from receptual in kind, why this necessity for the
historical priority of the latter? Why should denotation thus always
require to precede denomination—or receptual connotation thus always
require to precede conceptual predication—unless it be that the one
is a further and a continuous development of the other? Surely as well
might the botanist institute a specific distinction between the root
and the flower of the self-same plant, as the psychologist, with these
results of philological research before him, still persist in drawing
a distinction of kind between the receptual denotation of “radical
elements,” and the full efflorescence of conceptual thought.

A single illustration may serve to convey the force of this argument
more fully than any abstract discussion of it. But I will introduce the
illustration with an analogous case. The following well-established
fact I quote from Geiger:—

“Man had language before he had tools.... On considering a word
denoting an activity carried on with a tool, we shall invariably
find that this was not its original meaning, but that it previously
implied a similar activity requiring only the natural organs....
This fact of the activity with implements deriving its name from
one more simple, ancient, and brute-like, is quite universal, and
I do not know how otherwise to account for it but that the name is
older than the activity with tools which it denotes at the present
time—that, in fact, the word was already extant before men used any
other organs but the native and natural ones.... The vestiges of his
earliest conceptions still preserved in language proclaim it loudly and
distinctly that man has developed from a state in which he had solely
to rely on the aid of his organs—a state, therefore, in which he
differed little in his habits from the brute creation, and with respect
to the enjoyment of his existence, nay, to his preservation, depended
almost entirely on whatever lucky chance presented to him.”[279]

Now, to this special illustration on the general principle of
“fundamental metaphor” it will doubtless be said—Very interesting in
itself; but, after all, it merely amounts to a philological proof that
tools are younger than words; that men did not always possess tools;
that tools were gradually invented; and that, when invented, they were
named by a metaphorical application of words previously in use.—Well,
if we are all agreed so far, I will proceed to adduce my illustration.

Judging from the now extensive literature which is opposed to
evolutionary teaching in the case of man, I gather that the great
majority of writers are quite as much impressed by the moral and
religious aspects of human psychology as they are by the intellectual.
Now, as already stated in the Preface, I reserve for a future volume
a full consideration of these distinctively human faculties. In the
present part of my work I am concerned exclusively with the question
as to the origin of those powers of conceptual thought which, under
any point of view, must be regarded as the necessary and antecedent
condition to the possibility both of conscience and religion.
Nevertheless, merely for the sake of supplying an illustration touching
the point now before us, I may here forestall a little of what I shall
hereafter have to present in detail touching the evidence that we have
of the genesis of conscience. And this I will do by another quotation
from the same philologist, seeing that he is an authority whom none of
my opponents can afford to ignore.

“If we examine the words, those oldest pre-historic testimonies,
we shall find that all moral notions contain something morally
indifferent.” That is to say, they all contain what I have termed
a “receptual core,” expressive of some simple physical process,
or condition, the name of which has been afterwards transferred,
by “fundamental metaphor,” to the moral “concept.” Omitting the
illustrations, the passage continues as follows:—“But why have not
the morally good and bad their own names in language? Why do we know
them from something else that previously had its appellation? Evidently
because language dates from a period when a moral judgment, a knowledge
of good and evil, had not yet dawned in the human mind.”[280]

Now, at present I am not concerned with this conclusion, further than
to remark that I do not see how it is to be obviated, if our previous
agreement is to stand with regard to the precisely analogous case
of the names of tools. That is to say, if any one allows that the
philological evidence is sufficient to prove the priority of words
to the tools which they designate, consistency must constrain him
also to allow that the fundamental concepts of morality are of later
origin than the names by which they have been baptized, and in virtue
of which they must be regarded as having become concepts at all.
These names—just like the names of tools—were all originally of
nothing more than pre-conceptual significance, serving to denote such
obvious physical states or activities as were immediately cognizable
by the powers of sensuous perception and direct association. Then,
as the moral sense began to dawn, and the utilitarian significance
of conduct as ethical began to be appreciated, the principles of
“fundamental metaphor” were applied to the naming of these newly found
concepts—presumably at about the same time as these same principles
were applied to the naming of newly found tools.

Now, this is only one illustration out of a practically infinite number
of others which it would be easy to quote—seeing, indeed, as Whitney
observes, that “we can hardly write a line without giving illustrations
of this kind of linguistic growth.” And whatever may be thought (at
this premature stage of our inquiry) concerning the application of
the general principle before us to the special case of conscience,
it appears to me there can be no question at all that this general
principle of “fundamental metaphor” reveals the fact of an intellectual
growth from what I have called the pre-conceptual to the conceptual
phase; and, moreover, that it proves such a growth to have been the
universal characteristic of human faculty in those pre-historic times
of which language preserves to us the only record.[281]

       *       *       *       *       *

There still remains one other department of philological inquiry to
be considered, and its consideration will tend yet further and most
forcibly to corroborate all the general conclusions already attained.
Hitherto we have been engaged for the most part on what I have already
called the palæontology of human thought as revealed, fossil-like, in
the linguistic petrifactions of pre-historic man. But the science of
comparative philology is not confined in its researches upon early
forms of speech to the bygone remnants of a distant age. On the
contrary, just like the science of comparative anatomy, it is furnished
with still existing materials for study, which are of the nature of
living organisms, and which present so many grades of evolution that
the lowest members of the series bring us within easy distance of
those aboriginal forms which can only be studied in the fossil state.
Hitherto I have considered these lowest existing languages only with
reference to their forms of predication. Here I desire to consider them
with reference to the quality of ideation that they betoken.

In the next instalment of my work I shall have to treat of the
psychology of savages, and then it will become apparent that there
is no very precise relation to be constantly traced between grades
of mental evolution in general, and of language-development in
particular. Nevertheless there is a general relation: and therefore
it is among the lowest savages that we meet with the lowest types
of language-structure.[282] In the present connection I shall have
to treat of these languages only in so far as they throw light upon
the quality of ideation with which they are concerned, or so far as
they are related to the general principles with which we have already
been occupied. And, even as thus limited, I will endeavour to make my
exposition as brief as possible.

I will begin by supplying a few quotations from the more competent
authorities who have written upon the subject from a linguistic point
of view.

“It requires but the feeblest power of abstraction—a power even
possessed by idiots—to use a name as the sign of a conception,
_e.g._ to say ‘sun’;[283]—to say ‘sheen,’ as the description of a
phenomenon common to all shining objects, is a higher effort, and to
say ‘to shine’ as expressive of the state or act is higher still. Now,
familiar as such efforts may be to us, there is ample proof that they
could not have been so to the inventors of language, because they are
not so, even now, to some nations of mankind after all their long
millenniums of existence. Instances of this fact have been repeatedly
adduced.”[284] Thus, for example, the Society Islanders have separate
words for dog’s-tail, bird’s-tail, sheep’s-tail, &c., but no word for
tail itself—_i.e._ tail in general.[285] The Mohicans have words to
signify different kinds of cutting, but no verb “to cut;” and forms
for “I love him,” “I love you,” &c., but no verb “to love;” while the
Choctanis have names for different species of oak, but no word for the
genus oak.[286] Again, the Australians have no word for tree, or even
for bird, fish, &c.;[287] and the Eskimo, although he has verbs which
signify to fish-seal, to fish-whale, &c., has not any verb “to fish.”
“Ces langues,” Du Ponceau remarks, “généralisent rarement;” and he
shows that they have not even any verb to imply “I will,” or “I wish,”
although they have separate verbal forms for “I wish to eat meat,”
“I wish to eat soup;” neither have they any general noun-substantive
which means “a blow,” although they have a variety which severally mean
blows with as many different kinds of instruments.[288] Similarly, Mr.
Crawford tells us, “the Malay is very deficient in abstract words; and
the usual train of ideas of the people who speak it does not lead them
to make a frequent use even of the few they possess. With this poverty
of the abstract is united a redundancy of the concrete,”—and he gives
many instances of the same kind as those above rendered from other
languages.[289] So, likewise, we are told, “the dialect of the Zulus is
rich in nouns denoting different objects of the same genus, according
to some variety of colour, or deficiency of members, or some other
peculiarity,” such as “white-cow,” “red-cow,” “brown-cow;”[290] and the
Sechuâna has no fewer than ten words all meaning “horned cattle.”[291]
Cheroki presents thirteen different verbs to signify different kinds
of washing, without any to indicate “washing” itself;[292] and Milligan
says that the aborigines of Tasmania had “no words representing
abstract ideas; for each variety of gum-tree, wattle-tree, &c., they
had a name, but they had no equivalent for the expression of ‘a tree;’
neither could they express abstract qualities, such as hard, soft,
warm, cold, long, short, round.”[293]

Lastly, to give only one other example, Dr. Latham states that a Kurd
of the Zaza tribe, who furnished Dr. Sandwith with a list of native
words, was not “able to conceive a hand or father, except so far as
they were related to himself, or something else; and so essentially
concrete rather than abstract were his notions, that he combined the
pronoun with the substantive whenever he had a part of the human body
or a degree of consanguinity to name,” saying _sere-min_, “my head,”
and _pie-min_, “my father.”

Thus, as Professor Sayce remarks, after alluding to some of the above
facts, “we may be sure that it was not “the ‘ideas of prime importance’
which primitive man struggled to represent, but those individual
objects of which his senses were cognisant.”[294] And, without further
multiplying testimony, we may now be prepared to accept from him the
general statement that, “all over the world, indeed, wherever we come
across a savage race, or an individual who has been unaffected by the
civilization around him, we find this primitive inability to separate
the particular from the universal by isolating the individual word,
and extracting it, as it were, from the ideas habitually associated
with it.”[295] Or, in my own phraseology, among all primitive races
still existing, we meet with what must seem to my opponents a wholly
unintelligible incapacity to evolve a concept from any number of
recepts, notwithstanding that the latter may all be most nearly
related together, and severally named by as many denotative signs:
even with their numberless already-formed words for different kinds
of trees, the aborigines of Tasmania could not designate “a tree.”
Of course they must have had a recept of a tree, or a generic image
formed out of innumerable perceptions of particular trees—so that,
for instance, it would doubtless have surprised a Tasmanian could he
have seen a tree (even though it were a new species for which he had no
name) standing inverted with its roots in the air and its branches in
the ground. In just the same way a dog is surprised when it first sees
a man walking on his hands: the dog will bark at such an object because
it conflicts with the generic image which has been automatically formed
by numberless perceptions of individual men walking on their feet.
But, in the absence of any name for trees in general, there is nothing
to show that the savage has a concept answering to “tree,” any more
than that the dog has a concept answering to “man.” Indeed, unless my
opponents vacate the basis of Nominalism on which their opposition is
founded, they must acknowledge that in the absence of any _name_ for
tree there _can be no conception_ of tree.

So much, then, for what Archdeacon Farrar has called “_the hopeless
poverty of the power of abstraction_” in savages. Their various
languages unite, in verbal testimony, to assure us that human thought
does _not_ “proceed from the abstract to the concrete;” but, on the
contrary, that in the race, as in the individual, receptual ideation is
the precursor of conceptual—denotation the antecedent of denomination,
as in still earlier stages it was itself preceded by gesticulation.
Such being the case with regard to names, it is no wonder, as we
previously found, that low savages are so extraordinarily deficient in
their forms of predication.

       *       *       *       *       *

The palæontology of human thought, then, as recorded in language,
incontestibly proves that the origin and progress of ideation in
the race was psychologically identical with what we now observe in
the individual. All the stages of ideation which we have seen to be
characteristic of psychogenesis in a child, are thus revealed to us as
having been characteristic of psychogenesis in mankind.

First there was the indicative stage. This is proved in two ways. On
the one hand, all philologists will now agree with Geiger—“But, what
says more than anything, language diminishes the further we look back,
in such a way that we cannot forbear concluding it must once have had
no existence at all.”[296] On the other hand, even if we tap the tree
of language as high up in its stem as the pronominal roots of Sanskrit,
what is the kind of ideational sap which flows therefrom? It is, as we
have already seen, so strongly suggestive of gesture and grimace that
even Professor Max Müller allows that in it we have “remnants of the
earliest and almost pantomimic phase of language, in which language was
hardly as yet what we mean by language, namely _logos_, a gathering,
but only a pointing.”[297]

Secondly, we have clear evidence of sentence-words, as well as of
what I have called the denotative phase, or the naming of simple
recepts—whether only of actions, or, as we may safely assume,
likewise also of objects and qualities; and whether arbitrarily,
or, as seems virtually certain, in chief part by onomatopœia. Both
these subordinate points, however—which are rendered more doubtful
on account of the struggle for existence among words having proved
favourable to denotative terms expressive of actions, and unfavourable
to the survival of onomatopœia—are of comparatively little moment to
us; the important fact is the one which is most clearly testified to by
the philological record, namely, that the lowest strata of this record
yield fossils of the lowest order of development: the “121 concepts,”
appear to be, for the most part, denotations of simple recepts.

Thirdly, higher up in the stratified deposits, we meet with
overwhelming evidence of the connotative extension of these denotative
terms. Indeed, many of these terms have probably undergone a certain
amount of connotative extension as the condition to their having
survived as roots; and, therefore, in these lowest deposits it is
difficult to be sure that an apparently denotative term is not really a
term which has undergone the earlier stages of connotative extension.
If such were the case, we can understand the loss of any onomatopoetic
significance which it may originally have presented. But, however this
may be, there is an endless mass of evidence to prove the subsequent
and continuous growth of connotative extension throughout the whole
range of philological time.

Lastly, as regards the predicative phase, we have seen that philology
shows the same order and method to have been followed in the race as
in the child. In the growing child, as we have seen, pre-conceptual
predication is contemporary with—or occupies the same psychological
level as—the connotative extension of denotative terms. Indeed, the
very act of connotation is in itself an act of predication—if in the
conceptual sphere, of conceptual predication (denomination); if in
the pre-conceptual, of pre-conceptual. Again, in the psychogenesis of
the child we noted how important a part is played in the development
of pre-conceptual predication by the mere apposition of connotative
terms—such apposition being rendered inevitable by the laws of
association. If A is the connotative name for _A_, B the connotative
name for _B_, when the young child sees that _A_ and _B_ occur
together, the statement A B is rendered inevitable by “the logic of
events;” and this statement is a pre-conceptual proposition. Now,
in both these respects philology yields abundant parallels. The
quotations which I have given conclusively prove that “every word
must originally have been a sentence;” or, in my own terminology, a
pre-conceptual proposition of precisely the same kind as that which is
employed by a young child. If it be replied that the young child is
without self-consciousness, while the primitive man was not without
self-consciousness, this would merely be to beg the whole question on
which we are engaged, and, moreover, to beg it in the teeth of every
antecedent probability, as well as of every actual analogy, to which
appeal can possibly be made. If it be true—and who will venture to
doubt it?—that “language diminishes the further we look back, in
such a way that we cannot forbear concluding it must once have had no
existence at all,” will it be maintained that the man-like being who
was then unable to communicate with his fellows by means of any words
at all was gifted with self-consciousness? Should so absurd a statement
be ventured, it would be fatal to the argument of my adversaries; for
the statement would imply, either that concepts may exist without
names, or that self-consciousness may exist without concepts. The
truth of the matter is that philology has proved, in a singularly
complete manner, the origin and gradual development in time, first of
pre-conceptual communication, and next of the self-consciousness which
supplied the basis of conceptual predication. No wonder, therefore,
as Professor Max Müller somewhat naively observes, “it may be said
that the first step in the formation of names and concepts is very
imperfect. So it is.” Truly “to name the act of carrying by a root
formed from sounds which accompany the act of carrying a heavy load,
is a far more primitive act than to fix an attribute by a name”
conceptually applied. So primitive, indeed, is nomination of this
kind, that I defy any one to show wherein it differs psychologically
from what I have called the denotation of a young child, or even of a
talking bird.

And, having reduced the matter to this issue so far as the results of
philology are concerned, I may fitly conclude by briefly indicating
the principal point which appears to divide my opinions from those
of the eminent philologist just alluded to—if not also from those
of the majority of my psychological opponents. Briefly, the point is
that on the other side an unwarrantable assumption is made—to wit,
that conceptual thought is an antecedent condition, _sine quâ non_, to
any and every act of bestowing a name; and, _a fortiori_, to any and
every act of predication. This is the fundamental assumption, which,
whether openly expressed or covertly implied, serves as the basis of
the whole superstructure of my opponents’ argument. Now, I claim to
have shown, by a complete inductive proof, that this assumption is not
only unwarrantable in theory, but false in fact. There are names and
names. Not every name that is bestowed betokens conceptual thought on
the part of the namer. Alike from the case of the talking bird, of the
young child, and of early man (so far as he has left any traces of his
psychology in the structure of language), I have demonstrated that
prior to the stage of denomination there are the stages of indication,
denotation, and receptual connotation. These are the psychological
stepping-stones across that “Rubicon of Mind,” which, owing to their
neglect, has seemed to be impassable. The Concept (and, _a fortiori_,
the Proposition) is not a structure of ideation which is presented to
us without a developmental history. Although it has been uniformly
assumed by all my opponents “that the simplest element of thought” can
have had no such history, the assumption is, as I have said, directly
contradicted by observable fact. Had the case been otherwise—had
the concept really been without father and without mother, without
beginning of days or end of life—then truly a case might have been
shown for regarding it as an entity _sui generis_, destitute of kith
or kin among all the other faculties of mind. But, as we have now so
fully seen, no such unique exception to the otherwise uniform process
of evolution can here be maintained: the phases of development which
have gradually led up to conceptual thought admit of being as clearly
traced as those which have led to any other product, whether of life or
of mind.

Here, then, I bring to a close this brief and imperfect rendering of
the “Witness of Philology.” But, brief and imperfect as the rendering
is, I am honestly unable to see how it is conceivable that the witness
itself could have been more uniform as to its testimony, or more
multifarious as to its facts—more consistent, more complete, or more
altogether overwhelming than we have found it to be. In almost every
single respect it has corroborated the results of our psychological
analysis. It has come forward like a living thing, which, in the very
voice of Language itself, directly and circumstantially narrates to
us the actual history of a process the constituent phases of which we
had previously inferred. It has told us of a time when as yet mankind
were altogether speechless, and able to communicate with one another
only by means of gesticulation and grimace. It has described to us
the first articulate sounds in the form of sentence-words, without
significance apart from the pointings by which they were accompanied.
It has revealed the gradual differentiation of such a protoplasmic form
of language into “parts of speech;” and declared that these grammatical
structures were originally the offspring of gesture-signs. More
particularly, it has shown that in the earliest stages of articulate
utterance pronominal elements, and even predicative words, were used
in the impersonal manner which belongs to a hitherto undeveloped form
of self-consciousness—primitive man, like a young child, having
therefore spoken of his own personality in objective terminology.
It has taught us to find in the body of every conceptual term a
pre-conceptual core; so that, as the learned and thoughtful Garnett
says, “_nihil in oratione quod non prius in sensu_ may now be regarded
as an incontrovertible axiom.”[298] It has minutely described the whole
of that wonderful aftergrowth of articulate utterance through many
lines of divergent evolution, in virtue of which all nations of the
earth are now in possession, in one degree or another, of the god-like
attributes of reason and of speech. Truly, as Archdeacon Farrar says,
“to the flippant and the ignorant, how ridiculous is the apparent
inadequacy of the origin to produce such a result.”[299] But here, as
elsewhere, it is the method of evolution to bring to nought the things
that are mighty by the things that are of no reputation; and when we
feel disposed to boast ourselves in that we alone may claim the Logos,
should we not do well to pause and remember in what it was that this
our high prerogative arose? “So hat auch keine Sprache ein abstractum,
zu dem sie nicht durch Ton und Gefühl gelangt wäre.”[300] To my mind
it is simply inconceivable that any stronger proof of mental evolution
could be furnished, than is furnished in this one great fact by the
whole warp and woof of the thousand dialects of every pattern which
are now spread over the surface of the globe. We cannot speak to each
other in any tongue without declaring the pre-conceptual derivation of
our speech; we cannot so much as discuss the “origin of human faculty”
itself, without announcing, in the very medium of our discussion,
what that origin has been. It is to Language that my opponents have
appealed: by Language they are hopelessly condemned.




CHAPTER XVI.

THE TRANSITION IN THE RACE.


At this point I shall doubtless be expected to offer some remarks
on the probable mode of transition between the brute and the human
being. Having so fully considered both the psychology and philology of
ideation, it may be thought that I am now in a position to indicate
what I suppose to have been the actual stepping-stones whereby an
intelligent species of ape can be conceived to have crossed “the
Rubicon of Mind.” But, if I am expected to do this, I might reasonably
decline, for two reasons.

In the first place, the attempt, even if it could be successful, would
be superfluous. The only objection I have had to meet is one which has
been raised on grounds of psychology. This objection I have met, and
met upon its own grounds. If I have been successful, for the purposes
of argument nothing more remains to be said. If I have not been
successful, it is obviously impossible to strengthen my case by going
beyond the known facts of mind, as they actually exist before us, to
any hypothetical possibilities of mind in the dim ages of an unrecorded
past.

In the second place, any remarks which I have to offer upon this
subject must needs be of a wholly speculative or unverifiable
character. As well might the historian spend his time in suggesting
hypothetical histories of events known to have occurred in a
pre-historic age: his evidence that such and such events must have
occurred may be conclusive, and yet he may be quite in the dark as
to the precise conditions which led up to them, the time which was
occupied by them, and the particular method of their occurrence. In
such cases it often happens that the more certain an historian may be
that such and such an event did take place, the greater is the number
of ways in which he sees that it might have taken place. Merely for the
sake of showing that this is likewise the case in the matter now before
us, I will devote the present chapter to a consideration of three
alternative—and equally hypothetical—histories of the transition.
But, from what has just been said, I hope it will be understood that I
attach no argumentative importance to any of these hypotheses.

       *       *       *       *       *

Sundry German philologists have endeavoured to show that speech
originated in wholly meaningless sounds, which in the first instance
were due to merely physiological conditions. In their opinion the
purely reflex mechanisms connected with vocalization would have
been sufficient to yield not only many differences of tone under
different states as to suffering, pleasure, effort, &c., but even the
germ of articulation in the meaningless utterance of vowel sounds
and consonants. Thus, for example, Lazarus says:—“Der Process der
eigenthümlich menschlichen Laut-Erzeugung, die Articulation der Tone,
die Hervorbringung von Vocalen und Consonanten, ist demnach auf
rein physiologischem Boden gegeben—in der urprünglichen Natur des
menschlichen physischen bewegten Organismus begründet, und wird vor
aller Willkür und Absicht also ohne Einwirkung des Geistes obwohl auf
Veranlassung von Gefühlen und Empfindungen vollzogen.”[301]

This, it will be observed, is the largest possible extension of
the interjectional theory of the origin of speech. It assumes that
not only inarticulate, but also articulate sounds were given forth
by the “sprachlosen Urmenschen,” in the way of instinctive cries,
wholly destitute of any semiotic intention. By repeated association,
however, they are supposed to have acquired, as it were automatically,
a semiotic value. For, to quote Professor Friedrich Müller, “Sie
sind zwar Anfangs bedeutungslos: sie können aber bedeutungsvoll
werden. Alles, was in unserem Inneren vorgeht, wird von der Seele
wahrgenommen. Sobald durch gewisse aüssere Einflüsse in Folge einer
Combination mehrerer Empfindungen eine Anschauung entsteht, nimmt die
Seele dieselbe an, Diese Anschauung hat—in Folge der durch eine der
Empfindungen hervorgebrachten Reflexbewegung in den Stimmorganen—einen
Laut zum Begleiter, welcher in gleicher Weise wie die Anschauung von
der Seele wahrgenommen wird, diese beiden Wahrnehmungen, nämlich jene
der Anschauung und jene des Lautes, _verbinden_ sich miteinander
vermöge ihrer _Gleichzeitigkeit_ im menschlichen Bewusstsein, es
findet also eine _Association_ der Laut-Anschauung mit jener der
_Sach_-Anschauung statt, die Elemente der Sach-Anschauung bekommen
an der Laute-Anschauung einen _festen_ Mittelpunkt, durch den die
_Anschauung_ zur _Vorstellung_ sich entwickelt. Wir sind damit bei
der menschlichen Sprache angelangt, welche also ihrem Wesen nach auf
der _Substituirung_ eines _Klang_-oder _Ton_bildes für das Bild einer
Anschauung beruht.”[302]

Now, without at all doubting the important part which originally
meaningless sounds may have played in furnishing material for vocal
sign-making, and still less disputing the agency of association in
the matter, I must nevertheless refuse to accept the above hypothesis
as anything like a full explanation of the origin of speech. For it
manifestly ignores the whole problem which stands to be solved—namely,
the genesis of those powers of ideation which first put a soul of
meaning into the previously insignificant sounds. Nearly all the
warm-blooded animals so far share with mankind the same physiological
nature as to give forth a variety of vocal sounds under as great a
variety of mental states. Therefore, if in accordance with the above
hypothesis we regard all such sounds as meaningless (or arising from
the “purely physiological basis” of reflex movement), the question
obviously presents itself, Why have not the lower animals developed
speech? According to the above doctrine, aboriginal and hitherto
speechless man started without any superiority in respect of the
sign-making faculty, and thus far precisely resembled what is taken
to be the present psychological condition of the lower animals.[303]
Why, then, out of the same original conditions has there arisen
so enormous a difference of result? If, in the case of mankind,
associations of meaningless sounds with particular states, objects,
&c., led to a substitution of the former for the latter, and thus
gave to them the significance of names, how are we to account for the
total absence of any such development in brutes? To me it appears
that this is clearly an unanswerable difficulty; and therefore I do
not wonder that the so-called interjectional theory of the origin of
speech has brought discredit on the whole philosophy of the subject.
But, as so often happens in philosophical writings, we have here a
case where an important truth is damaged by imperfect or erroneous
presentation. All the principles set forth in the above hypothesis
are sound in themselves, but the premiss from which they start is
untrue. This premiss is, that aboriginal man presented no rudiments
of the sign-making faculty—that this faculty itself required to be
originated _de novo_ by accidental associations of sounds with things.
But, as we now well know from all the facts previously given, even
the lower animals present the sign-making faculty in no mean degree
of development; and, therefore, it is perfectly certain that the
“Urmenschen,” at the time when they were “sprachlosen,” were not on
this account _zeichenlosen_. The psychological germ of communication,
which probably could not have been created by merely accidental
associations between sounds and things, must already have been given in
those psychological conditions of receptual ideation which are common
to all intelligent animals.

But to this all-essential germ, as thus given, I doubt not that the
soil of such associations as the interjectional theory has in view
must have been of no small importance; for this would naturally help
to nourish its semiotic nature. And the reason why the similar germ of
sign-making which occurs in the brute creation has not been similarly
nurtured, I have already considered in Chapter VIII. For, it is
needless to add, on every ground I disagree with the above quotations
where they represent articulate sounds as having been aboriginally
uttered by “Urmenschen” in the way of instinctive cries, without any
vestige of semiotic intention.[304]

       *       *       *       *       *

I will now pass on to consider the two other hypotheses; and by way of
introduction to both we must remember that our materials of study on
the side of the apes is very limited. I do not mean only that no single
representative of any of the anthropoid apes has ever been made the
object of even so much observation with respect to its intelligence
as I bestowed upon a cebus. Yet this, no doubt, is an important
point, because we know that of all quadrumana—and, therefore, of all
existing animals—the anthropoid apes are the most intelligent, and,
therefore, if specially trained would probably display greater aptitude
in the matter of sign-making than is to be met with in any other kind
of brute. But I do not press this point. What I now refer to is the
fact that the existing species of anthropoid apes are very few in
number, and appear to be all on the high-road to extinction. Moreover,
it is certain that none of these existing species can have been the
progenitor of man; and, lastly, it is equally certain that the extinct
species (or genus) which did give origin to man must have differed in
several important respects from any of its existing allies. In the
first place, it must have been more social in habits; and, in the next
place, it was probably more vociferous than the orang, the gorilla,
or the chimpanzee. That there is no improbability in either of these
suppositions will be at once apparent if we remember that both are
amply sustained by analogies among existing and allied species of
the monkey tribe. Or, to state these preliminary considerations in a
converse form, when it is assumed[305] that because the few existing
and expiring species of anthropoid apes are unsocial and comparatively
silent, therefore the simian ancestors of man must have been so, it is
enough to point to the variability of both these habits among certain
allied genera of monkeys and baboons, in order at the same time to
dispose of the assumption, and to indicate the probable reasons why one
genus of ape gradually became evolved into _Homo_, while all the allied
genera became, or are still becoming, extinct.

Again, and still by way of preliminary consideration, we must remember
that the analogy of the growing child, although most valuable up to a
certain point, is not to be unreservedly followed where we have to deal
with the genesis of speech. For, as previously noted, to the infancy of
the individual language is supplied from without, and has only to be
learnt; while to the infancy of the race language was not supplied, but
had to be made. Therefore, even apart from any question of heredity,
we have here an immense difference in the psychological conditions
between the case of a growing child and that of aboriginal man. Only
in so far as the growing child displays the tendency on which I have
dwelt of spontaneously extending the significance of denotative words,
or of spontaneously using such words in apposition for the purpose of
pre-conceptual predication—only to this extent may we hope to find any
true analogy between the individual and the race in respect of that
“transition” from receptual to conceptual ideation with which we are
now concerned.[306]

There is another preliminary consideration which I think is well worth
mentioning. The philologist Geiger is led by his study of language
to entertain, and somewhat elaborately to sustain, the following
doctrine. First he points out that man, much more than any other
animal, uses the sense of sight for the purposes of perceptual life.
By this he does not mean that man possesses a keener vision than
any other animal, but merely that of all his special senses that of
sight is most habitually used for taking cognizance of the external
world. And this, I think, must certainly be admitted. Even a hitherto
speechless infant may be seen to observe objects at great distances,
carefully to investigate objects which it holds in its hands, and
generally to employ its eyes much more effectively than any of the
lower animals at a comparable stage of development. Now, from this
relative superiority of the sense of sight in man, Geiger argues that
before the origin of articulate speech he, more than any other animal,
must have been accustomed to communicate with his fellows by means of
signs which appealed to that sense—_i.e._ by gesture and grimace. But,
if this be admitted, it follows that from the time when a particular
species of the order Primates began to use its eyesight more than the
allied species, a condition was given favourable to the subsequent
and gradual development of a gesticulating form of ape-like creature.
Here grimace also would have played an important part, and where
attention was particularly directed towards movements of the mouth for
semiotic purposes, articulate sounds would begin to acquire more or
less conventional significations. In this way Geiger supposes that the
conditions required for the origin of articulate signs were laid down;
and, in view of all that he says, it certainly is suggestive that the
animal which relies most upon the sense of sight is also the animal
which has made so prodigious an advance in the faculty of sign-making.
In this greater reliance on the sense of sight, therefore, we probably
have another among the many and complex conditions which determined the
difference in respect of sign-making between the remote progenitors
of man and their nearest zoological allies—a difference which would
naturally become more and more pronounced the more that vision and
gesticulation acted and reacted on one another.

It appears to me that this suggestion of Geiger admits of being
strikingly supported by certain facts which are known to obtain in the
case of deaf-mutes. Even when wholly uneducated, the born mute, as
we have previously seen, habitually invents articulate sounds as his
own names of things. These sounds are, of course, unheard by the mute
himself, and their use must be ascribed—as I have already ascribed
it—to the hereditary transmission of an acquired propensity. But the
point now is that, although the majority of these articulate sounds
appear to be wholly arbitrary (_e.g._ _ga_ for “one,” _schuppatter_
for “two,” _riecke_ for “I will not”), a certain proportion are often
clearly traceable to vocalizations incidental to movements of the
mouth in performing the actions signified (_e.g._ _mumm_ for “eating,”
_schipp_ for “drinking”).[307] Similarly, observation of a dog’s
mouth, while in the act of barking, leads to an imitative action on
the part of a mute as his sign for a dog, and this in turn may lead
to the utterance of such an articulate sound as _be-yer_, which the
mute afterwards uses as his name for a dog.[308] Now, if words may
thus be coined even by deaf-mutes as a result of observing movements
of the mouth, much more is this likely to have been the case among the
“Urmenschen,” who were able not only to see the movements, but also to
hear the sounds.

       *       *       *       *       *

I will now adduce the two hypotheses above alluded to as conceivable
suggestions touching the mode of transition. First, let us try to
imagine an anthropoid ape, social in habits, using its voice somewhat
extensively as an organ of sign-making after the manner of all other
species of social quadrumana, and possibly somewhat more sagacious
than the orang-outang mentioned in my previous work,[309] or the
remarkable chimpanzee now in the Zoological Gardens, which, in respect
of intelligence as well as comparative hairlessness and carnivorous
propensities, appears to be the most human-like of animals hitherto
discovered in the living state.[310] It does not seem to me difficult
further to imagine that such an animal should extend the vocal signs
which it habitually employs in the expression of its emotions and the
logic of its recepts, to an association with gesture-signs, so as to
constitute sentence-words indicative of such simple and often-repeated
ideas as the presence of danger, discovery of food, &c. Nay, I do not
think it is too much to suppose that such an animal may even have gone
so far as to make sounds which were denotative of a few of the most
familiar objects, such as food, child, enemy, &c., and also, possibly,
of frequently repeated forms of activity; for this, as I have shown at
considerable length, is no more than we actually observe to be done by
animals which are lower in the scale of intelligence; and although it
is not done by articulate signs (except in the psychologically poor
instance of talking birds), this, as I have also shown, is a matter
of no psychological import. Whether the denotative stage of language
in the ape was first reached by articulation, or (as I think is very
much more probable) by vocal sounds of other kinds assisted by gestures
and grimace, is similarly immaterial. In either case the advance of
intelligence which would thus have been secured would in time have
reacted upon the sign-making faculty, and so have led to the extension
of the vocabulary, both as to sounds and gestures. Sooner or later the
vocal signs—assisted out by gestures and ever leading to a gradual
advance of intelligence—would have become more or less conventional,
and so, in the presence of suitable anatomical and social conditions,
articulate. Thus far I cannot see anything to stumble over, when we
remember all that has been said upon the conventional signs which are
used by the more intelligent of our domesticated animals, and even by
talking birds.[311]

This is the hypothesis which is countenanced by Mr. Darwin in his
_Descent of Man_. He says:—“I cannot doubt that language owes its
origin to the imitation and modification of various natural sounds,
the voices of other animals, and man’s own instinctive cries, aided by
signs and gestures.... Since monkeys certainly understand much that
is said to them by man, and, when wild, utter signal-cries of danger
to their fellows; and since fowls give distinct warnings for danger
on the ground, or in the sky from hawks (both, as well as a third
cry, intelligible to dogs),[312] may not some unusually wise ape-like
animal have imitated the growl of a beast of prey, and thus told his
fellow-monkeys the nature of the expected danger? This would have been
a first step in the formation of a language.”[313]

But Mr. Darwin adds another feature to the hypothesis now under
consideration, as follows:—

“When we treat of sexual selection we shall see that primæval man, or
rather some early progenitor of man, probably first used his voice in
producing true musical cadences, that is in singing, as do some of the
gibbon-apes at the present day; and we may conclude, from a widely
spread analogy, that this power would have been especially exerted
during the courtship of the sexes,—would have expressed various
emotions, such as love, jealousy, triumph,—and would have served as
a challenge to rivals. It is, therefore, probable that the imitation
of musical cries by articulate sounds may have given rise to words
expressive of various complex emotional states.”[314]

       *       *       *       *       *

Such, then, is one way in which it appears to me quite conceivable
that the faculty of articulate sign-making might have taken the first
step towards the formation of speech. But, not to go further than this
first step, I can see another possibility as to the precise method of
attainment, and one which I think is still more probable. It is the
opinion of some authorities in anthropology that speech was probably,
and comparatively speaking, late in making its appearance; so that
our ancestors in whom it did first appear were already more human
than simian, and as such deserving of the name _Homo alalus_.[315]
Now, if this were the case, the course of our hypothetical history
would be even more easy to imagine than it was under the supposition
previously considered. For, under the present supposition, we start
with an already man-like creature, erect in attitude, much more
intelligent than any other animal, shaping flints to serve as tools and
weapons, living in tribes or societies, and able in no small degree
to communicate the logic of his recepts by means of gesture-signs,
facial expressions, and vocal tones. Clearly, from such an origin, the
subsequent evolution of sign-making in the direction of articulate
sounds would be an even more easy matter to imagine than under the
previous hypothesis. For, let us try to imagine a community of _Homo
alalus_, considerably more intelligent than the existing anthropoid
apes, although still considerably below the intellectual level of
existing savages. It is certain that in such a community natural signs
of voice, gesture, and grimace would be in vogue to a greater or
less extent.[316] As their numbers increased (and, consequently, as
natural selection laid a greater and greater premium on intelligent
co-operation, as in the case of social insects),[317] such signs
would require to become more and more conventional, or acquire more
and more the character of sentence-words and denotative signs.[318]
Now, where the signs were vocal, the only ways in which they could
be developed so as to meet this need would be, (1) conventional
modulations of intensity, (2) of pitch, and (3) of time-intervals. But
clearly, neither modulations of intensity nor of pitch could carry
improvement very far, seeing that the human voice does not admit of
any great range of either. Consequently, if any improvement at all
were to be effected—and it was bound to be effected, if possible, by
natural selection,—it could only be so in the direction of modulating
time-intervals between vocal sounds. Now, such a modulation of
time-intervals is the beginning of _articulation_.

That is to say, the first articulation probably consisted in nothing
further than a semiotic breaking of vocal tones, in a manner resembling
that which still occurs in the so-called “chattering” of monkeys—the
natural language for the expression of their mental states. The
great difference would be that the semiotic value of such incipient
articulation must have been more largely intellectual, or less purely
emotional: it must have partaken less of the nature of cries, and more
of the nature of names. It seems probable that, as all natural cries
are given forth by the throat and larynx, with little or no assistance
from the tongue and lips, these first efforts at articulation would
have been mainly restricted to vowel sounds, sparsely supplemented by
guttural and labial consonants. This state of matters might have lasted
for an enormous length of time, during which the liquid, and lastly
the lingual consonants would perhaps have begun to be used. This is
the order in which we might expect the consonants to arise, in view of
the consideration that the gutturals and labials would probably have
admitted of more easy pronunciation than the liquids and linguals by an
almost speechless _Homo_.[319] From this point onwards, the further
development of articulation would only be a matter of time and mental
growth; but I think it is highly probable that the initial stages thus
sketched probably occupied a lapse of time out of all proportion to
that which was afterwards required for the higher developments.

Moreover, in this connection we must not neglect to notice the “clicks”
of the African Bushmen and Hottentots, which appear to furnish us with
direct evidence of the survival among these low races of a primordially
inarticulate system of sign-making.[320] No one has studied the
languages of these peoples with so much labour or so much result as the
philosophically minded Dr. Bleek, and he says that the clicks which
occur in the great majority of their words, “must be made an object of
special attention if we would arrive at even an approximate idea of the
original vocal elements from which human language sprang.”[321]

The clicks in question are four in number, or, according to Bleek,
“at least six.” They are called the dental, palatal, cerebral, and
lateral. The lateral click is the same as that which is employed by our
own grooms when urging a horse. The dental is also used by European
races as a sound expressive of disappointment, unspeakable contempt,
&c. In books it is usually written “tut, tut,” which serves to show
how hopeless is any attempt at translating a click into any articulate
equivalent. The other two clicks are formed by the tongue operating
upon the roof of the mouth. Some remote idea of the difficulty of
rendering a language of this kind into any alphabetical form, may be
gained by trying to pronounce one of the words which are printed in
our European treatises upon them. For example, the Hottentot word for
“moon” is printed ║ _khãp_, where ║ stands for the lateral click, _kha_
for a guttural consonant, and ˜ for a nasal twang.

With reference to this inarticulate kind of sign-making, which thus
so largely prevails among the languages of low races in close organic
connection with articulate, it seems worth while to record the
following observation which was communicated by Professor Haeckel to
Dr. Bleek, and published by the latter in his work already quoted:—

“The language of apes has not hitherto received from zoologists the
attention which it deserves, and there are no accurate descriptions
of the sounds uttered by them. They are sometimes called ‘howls,’
sometimes ‘cries,’ ‘clicks,’ ‘roars,’ &c. Now, I have myself frequently
heard in zoological gardens, from apes of very different species,
remarkable clicking sounds, which are produced with the lips, and also,
though not so often, with the tongue; but I have nowhere been able to
find any account of them.”

Upon the whole, then, it appears to me extremely probable that in these
clicks we have survivals, in lowly developed languages, of a formerly
inarticulate condition of mankind; or, as Professor Sayce remarks from
a philological point of view, “the clicks of the Bushmen still survive
to show us how the utterances of speechless man could be made to embody
and convey thought.”[322]

       *       *       *       *       *

In its main outlines the hypothetical sketch which I have given follows
that which Mr. Darwin has drawn in his _Descent of Man_. As we have
already seen, however, there is this important difference. Mr. Darwin
entertains only the second of the three alternative hypotheses here
presented, or the hypothesis which assumes that the rudiments of
articulate speech began in the “ape-like,” or “early progenitors” of
man. He does not seem to have entertained the idea of _Homo alalus_ as
a connecting link between these early progenitors and _Homo sapiens_. I
may, therefore, here briefly give my reasons for thinking it probable
that this connecting link had an actual existence.

Let it be observed, in the first place, that there is no antagonism
between the two hypotheses in question—the latter, indeed, being
merely an extension of the former. For the latter adopts all
Mr. Darwin’s views as to the importance of instinctive cries,
danger-signals, &c., for the higher development of sign-making in
that “ape-like animal” which was the brutal progenitor of _Homo
alalus_.[323] Moreover, our hypothesis is entitled to assume, with
Mr. Darwin’s, that this anthropoid ape was presumably not only more
intelligent than any of the few surviving species, but also much
more social. And this is an important point to insist upon, because
it is obvious that the conditions of social life are also the prime
conditions to any considerable advance upon the sign-making faculty as
this occurs in existing apes. The only respect, therefore, in which
the two hypotheses differ is in the one supposing that the faculty of
articulate sign-making was a much later product of evolution than it
is taken to have been by the other. That is to say, while Mr. Darwin’s
hypothesis regards the commencement of articulation as a necessary
condition to any considerable advance upon the receptual intelligence
of our brutal ancestry, the present hypothesis regards it as more
probable that this receptual intelligence was largely developed by
gesture and vocal signs, before the latter can be said to have become
properly articulate—the result being that a creature rather more human
than “ape-like” was evolved, who, nevertheless, was still able to
communicate with his fellows only by means of gesture-signs and vocal
tones.

My reasons for regarding this hypothesis as more probable than the
other are these.

First of all, on grounds of psychology, I see no reason to doubt that
the receptual intelligence of an already intelligent and highly social
species of anthropoid ape would admit of considerable advance upon that
of any existing species without the aid of articulation—social habits
making all the difference as to the development of sign-making with
its consequent reaction upon mental development. Next, for these early
stages of advance, I do not see that articulate sign-making would have
conferred any considerable advantage over a further development of the
more natural systems. For, so long as the only co-operation required
had reference to comparatively simple actions, the language of tone
and gesture would have admitted of sufficient development to have met
all requirements. Lastly, if we take the growing child as an index of
psychogenesis in the race, there can be no doubt that it points to a
comparatively late origin of the faculty of articulation. Remembering
the general tendency of ontogenesis to foreshorten the history of
phylogenesis, it is, I think, most suggestive that—notwithstanding
its readiness to imitate, and notwithstanding its being surrounded by
spoken language—the infant does not begin to use articulate signs
until long after it has been able to express many of its receptual
ideas in the language of tone and gesture. It will be remembered that
I have already laid stress upon the astonishing degree of elaboration
which this form of language undergoes in the case of children who are
late in beginning to speak (see pp. 220). And although it might be
scarcely justifiable to take these cases as possibly representative
of the semiotic language of _Homo alalus_ (seeing that the child of
to-day inherits the cerebrum of _Homo sapiens_); still I think it is
no less certain that we should err on the opposite side, if we were
to take the case of a child who is precocious in the matter of speech
as a fair index of the grade of mental evolution at the time when
articulation first began in the race (seeing that the history of the
latter is probably foreshortened in that of the former). Yet, even if
we were to do this, for the sake of argument, the result would still
be most strongly to indicate that long before our remote ancestors
were able to use articulate speech, they were immeasurably in advance
of all existing brutes in their semiotic use of tone and gesture. For
even a precocious child does not begin to make any considerable use of
words as signs until it is well on into its second year, while usually
this stage is not reached until the third. And, at whatever age it
is reached, the general intelligence of the child is not only much
in advance of that of any existing brute, but the direction in which
this advance is most conspicuous is just the direction where, in the
present connection, it is most suggestive—namely, in that of natural
sign-making by tone and gesture.

       *       *       *       *       *

In view, then, of these several considerations, I am disposed to think
that the progress of mental evolution from the brute to the man most
probably took place by some such stages as the following.

Starting from the highly intelligent and social species of anthropoid
ape as pictured by Darwin, we can imagine that this animal was
accustomed to use its voice freely for the expression of its emotions,
uttering of danger-signals, and singing.[324] Possibly enough, also, it
may have been sufficiently intelligent to use a few imitative sounds
in the arbitrary way that Mr. Darwin suggests; and certainly sooner
or later the receptual life of this social animal must have advanced
far enough to have become comparable with that of an infant at about
two years of age. That is to say, this animal, although not yet having
begun to use articulate signs, must have advanced far enough in the
conventional use of natural signs (or signs with a natural origin in
tone and gesture, whether spontaneous only or intentionally imitative),
to have admitted of a tolerably free exchange of receptual ideas,
such as would be concerned in animal wants, and even, perhaps, in the
simplest forms of co-operative action.[325] Next, I think it probable
that the advance of receptual intelligence which would have been
occasioned by this advance in sign-making, would in turn have led to a
further development of the latter—the two thus acting and re-acting on
one another, until the language of tone and gesture became gradually
raised to the level of imperfect pantomime, as in children before they
begin to use words. At this stage, however, or even before it, I think
very probably vowel-sounds must have been employed in tone-language,
if not also a few of the consonants. And I think this not only on
account of the analogy furnished by an infant already alluded to, but
also because in the case of a “singing” animal, intelligent enough to
be constantly using its voice for semiotic purposes, and therefore
employing a variety of more or less conventional tones, including
clicks, it seems almost necessary that some of the vowel sounds—and
possibly also some of the consonants—should have been brought into
use. But, be this as it may, eventually the action and reaction of
receptual intelligence and conventional sign-making must have ended
in so far developing the former as to have admitted of the breaking
up (or articulation) of vocal sounds, as the only direction in which
any further improvement of vocal sign-making was possible. I think it
not improbable that this important stage in the development of speech
was greatly assisted by the already-existing habit of articulating
musical notes, supposing our progenitors to have resembled the gibbons
or the chimpanzees in this respect. But long after this first rude
beginning of articulate speech, the language of tone and gesture would
have continued as much the most important machinery of communication:
the half-human creature now before our imagination would probably
have struck us as a wonderful adept at making significant sounds and
movements both as to number and variety; but in all probability we
should scarcely have been able to notice the already-developing germ
of articulation. Nor do I believe that, if we were able to strike in
again upon the history thousands of years later, we should find that
pantomime had been superseded by speech. On the contrary, I believe we
should find that although considerable progress had been made in the
former, so that the object then before us might appear deserving of
being classed as _Homo_, we should also feel that he must needs still
be distinguished by the addition _alalus_. Lastly, I believe that this
most interesting creature probably lived for an inconceivably long time
before his faculty of articulate sign-making had developed sufficiently
far to begin to starve out the more primitive and more natural systems;
and I believe that, even after this starving-out process did begin,
another inconceivable lapse of time must have been required for such
progress to have eventually transformed _Homo alalus_ into _Homo
sapiens_.

       *       *       *       *       *

It is now time to consider a branch of this hypothesis which has been
suggested by the philologist Professor Noiré, to which allusion has
already been made in an earlier chapter.[326]

Before Mr. Darwin had published his views, Professor Noiré had
elaborated a theory of the origin of speech which was substantially
the same as that which I have already quoted from the _Descent of
Man_.[327] The only difference between the two was that, while Darwin
referred the origin of articulate speech from instinctive cries, &c.,
to the anthropoid apes, Noiré referred it to a being already human.
In other words, Noiré adopted what I have here called the third
hypothesis, which assumes a speechless form of man as anterior to the
existing form.[328] But, as a result of further deliberation, Noiré
came to the conclusion that “the objects of fear and trembling and
dismay are even now the least appropriate to enter into the pure,
clear, and tranquil sphere of speech-thought, or to supply the first
germs of it.” Accordingly, he discarded the view that these germs
were to be sought in instinctive cries and danger calls, in favour
of the hypothesis that articulation had its origin in sounds which
are made by bodies of men when engaged in common occupations. Having
already explained the elements of this Yo-he-ho theory, it will here
be enough to repeat that I think there is probably some measure of
truth in it; although I likewise think it self-evident that this cannot
have been the only source of aboriginal speech. In what proportion
this branch of onomatopœia was concerned in the genesis of aboriginal
words—supposing it to have been concerned at all—we have now no means
of even conjecturing. But seeing that there are so many other sources
of onomatopœia supplied by Nature, and that these other sources are
so apparent in all existing languages, while the one suggested by
Noiré has not left a record of its occurrence in any language,—seeing
these things, I conclude, as before stated, that at best the Yo-he-ho
principle can be accredited with but a small proportional part in the
aboriginal genesis of language.[329] Therefore, with respect to this
hypothesis I have only three remarks to make: (1) that it is plainly
but a special branch of the general onomatopoetic theory; (2) that,
as such, it not improbably presents some measure of truth; and (3)
that, consequently, it ought to be regarded—not as it is regarded
by its author Noiré and its advocate Max Müller, namely, as the sole
explanation of the origin of speech, but—as representing only one
among many other ways in which, during many ages, many communities of
vociferous though hitherto speechless men may have slowly evolved the
art of making articulate signs.

       *       *       *       *       *

Probably it will be objected to this third hypothesis, in all its
branches, that it amounts to a _petetio principii_: _Homo alalus_,
it may be said, is _Homo postulatus_. To this I answer, Not so. The
question raised has been raised expressly and exclusively on the
faculty of conceptual speech, and it is conceded that of this faculty
there can have been no earlier phase than that of articulation.
Consequently, if my opponents assume that prior to the appearance of
this earliest phase it is impossible that any hitherto speechless
animal should have been erect in attitude, intelligent enough to chip
flints, or greatly in advance of other animals in the matter of making
indicative gesture-signs, assisted by vocal tones,—if my opponents
assume all this, it is _they_ who are endeavouring to beg the question.
For they are merely assuming, in the most arbitrary way, that the
faculty of conceptual thought is necessary in order that an animal
already semi-erect, should become more erect; in order that an animal
already intelligent enough to use stones for cracking nuts and opening
oysters, should not only (as at present) choose the most appropriate
stones for the purpose, but begin to fashion them for these or other
purposes; in order that an animal already more apt than any other
in the use of gesture and vocal signs, should advance considerably
along the same line of psychical improvement.[330] The hypothesis
that such a considerable advance might have gradually taken place,
up to the psychological level supposed, may or may not be true; but,
at least, it does not beg the question. The question is whether the
distinctively human faculty of conceptual ideation differs in kind or
in degree from the lower faculty of receptual ideation; and my present
suggestion amounts to nothing more than a supposition that receptual
ideation may have been developed in the animal kingdom to some such
level as it reaches in a child who is late in beginning to speak.[331]
If any opponent should object to this suggestion on the score of its
appearing to beg the question, he must remember that this question
only arises—in accordance with his own argument—at the place where
the faculty of sign-making ministers to that of introspective thought.
The question as to how far the lower faculties of mind admit of being
developed apart from (or, as I believe, antecedent to) the occurrence
of introspective thought, is obviously quite a distinct question. And
it is a question that can only be answered by observation. Now, I
have already shown that in the case of intelligent animals—and still
more in that of a growing child—the faculties of receptual ideation
do admit of being wrought up to an astonishing degree of adaptive
efficiency, without the possibility of their having been in any way
indebted to the distinctively human faculty of conceptual thought.

       *       *       *       *       *

On the whole, then, it seems to me probable, on grounds of psychology
alone, that the developmental history of intelligence in our race
so far resembled this history in the growing child that, prior to
the advent of speech, receptual ideation had attained a much higher
level of perfection than it now presents in any animal—so much
so, indeed, that the adult creature presenting it might well have
merited the name of _Homo alalus_. And, as we shall see in my next
volume, this inference on psychological grounds is corroborated by
certain inferences which may reasonably be drawn from some other
classes of facts. But in now for the present taking leave of this
question, I desire again to repeat, that it has nothing to do with
my main argument. For it makes no essential difference to my case
whether the faculty of speech was early or late in making its first
appearance. Under either alternative, so soon as the denotative
stage of articulation had been reached by our progenitors in the way
already sketched on its psychological side, the next stage would
have consisted in an extension of denotative signs into connotative
signs. As we have now seen, by a large accumulation of evidence, this
extension of denotative into connotative signs is rendered inevitable
through the principle of sensuous association. In other words, I have
adduced what can only be deemed a superabundance of facts to prove
that, in the first-talking child and even in the parrot, originally
denotative names of particular objects are spontaneously extended
to other objects sensuously perceived to be like in kind. And no
less superabundantly have I proved that this process of connotative
extension is antecedent to the rise of conceptual thought, and,
therefore, to that of true denomination. The limits to which such
purely receptual connotation may extend, I have shown to be determined
by the degree of development which has been reached by the faculties of
purely receptual apprehension. In the parrot this degree of development
is but low; in the dog and monkey considerably higher (though,
unfortunately, these animals are not able to give any articulate
expression to their receptual apprehensions); in the child of two years
it is higher still. But, as before shown, no antagonist can afford
to allege that in any of these cases there is a difference of kind
between the mental faculties that are respectively involved; because
his argument on psychological grounds can only stand upon the basis of
conceptual cognition, which, in turn, can only stand upon the basis of
self-consciousness; and this is demonstrably absent in the child until
long after the time when denotative names are connotatively extended by
the receptual intelligence of the child itself.

Thus, there can be no reasonable question that it is psychologically
possible for _Homo sapiens_ to have had an ancestry, which—whether
already partly human or still simian—was able to carry denotation
to a high level of connotation, without the need of cognition
belonging to the order conceptual. Whether the signs were then made
by tone and gesture alone, or likewise by articulate sounds, is also,
psychologically considered, immaterial. In either case connotation
would have followed denotation up to whatever point the higher
receptual (“pre-conceptual”) intelligence of such an ancestry was
able to take cognizance of simple analogies. And this psychological
possibility becomes on other grounds a probability of the highest
order, so soon as we know of any independent evidence touching the
corporeal evolution of man from a simian ancestry.

Now, we have already seen that pre-conceptual connotation amounts to
what I have termed pre-conceptual judgment. The qualities or relations
thus connotated are not indeed contemplated _as_ qualities or _as_
relations; but in the mere act of such a connotative classification
the higher receptual intelligence is virtually judging a resemblance,
and virtually predicating its judgment. Therefore I think it probable
that the earliest forms of such virtual predication were those which
would have been conveyed in single words. And, as we have seen in the
foregoing chapters, there is abundant and wholly independent evidence
to show, that this form of nascent predication continued to hold an
important place until so late in the intellectual history of our race
as to leave a permanent record of its occurrence in the structure of
all languages now extant.

The epoch during which these sentence-words prevailed was probably
immense; and, as we have before seen, far from having been inimical
to gesticulation, must have greatly encouraged it—raising, in
fact, the indicative phase of language to the level of elaborate
pantomime. Out of the complex of sentence-words and gesture-signs thus
inaugurated, grammatical forms became slowly evolved, as we know from
the independent witness of philology. But long before grammatical forms
of any sort began to be evolved, a kind of uncertain differentiation
must have taken place in this protoplasmic material of speech, in such
wise that some sentence-words would have tended to become specially
denotative of particular objects, others of particular actions, states,
qualities, and relations. This “primitive streak,” as it were, of
what was afterwards to constitute the vertebral column of articulated
language in the independent yet mutually related “parts of speech,”
must in large measure have owed its development to gesture. Now, by
this time, gesture itself must already have acquired an elementary kind
of syntax, such as belongs even to semiotic movements of an infant who
happens to be late in beginning to speak.[332] This elementary kind
of syntax would necessarily be taken over by, or impressed upon, the
growing structure of speech, at all events so far as the principles
and the order of apposition were concerned. Moreover, this sign-making
value of apposition would at the same time have been promoted
within the sphere of articulate signs themselves. For, as we have
previously seen, as soon as words become in any measure denotative,
they immediately begin to undergo a connotative extension;[333] and
with this progressive widening of signification, words require to be
more and more frequently used in apposition. Quite independently of
any as yet non-existing powers of introspective thought, the external
“logic of events” must have constantly determined such apposition
of receptually connotative terms, as we have already so fully seen
in the case of the growing child. Thus the conditions were laid for
the tripartite division—the genitive case, the adjective, and the
verb. Not till long subsequent ages, however, would this division
have taken place in its fulness. During the time which we are now
contemplating, there could have been no distinction at all between the
genitive case and the adjective; neither could there have been any
verbs as independent parts of speech. Nevertheless, already some of the
denotative signs would have been used as names of particular objects,
others of particular qualities, and yet others of particular actions,
states, and relations. Not yet deserving to be regarded as fully
differentiated parts of speech, these object-words, quality-words,
&c., would have resembled those with which we are all well acquainted
in nursery language, and which still survive, in a remarkably large
measure, among many dialects of a low order of development. Now, as
soon as these denotative names became at all fixed in meaning within
the limits of the same community, those which respectively signified
objects, qualities, actions, states, and relations, must necessarily
have been often used in apposition; and, as often as they were thus
used, would have constituted nascent or pre-conceptual propositions.

The probability certainly is that immense intervals of time would
have been consumed in the passage through these various grades of
mental evolution; but when we remember the great importance of this
kind of evolution to the species which had once begun to travel in
that direction, we cannot wonder that survival of the fittest should
have placed a high premium upon the instrument of its attainment—or,
in other words, that the faculty of sign-making, when once happily
started, should have been successively pushed onwards through ascending
grades of efficiency, so that it should soon become as unique in the
mammalian series as, for analogous reasons, are the flying powers of
the Chiroptera. But however long or however short the time may have
been that was required for our early progenitors to pass from one of
these stages of sign-making to another, so soon as the denotative name
of an object was brought into apposition with the denotative name of a
quality or an action, so soon was there uttered the virtual statement
of a virtual judgment, even though the mind which formed it was very
far indeed from being able either to think about its judgment as a
judgment, or to state a truth as true.

Thus we perceive that two different principles were presumably
concerned in the genesis of what I have called pre-conceptual
predication. The first consists in the natural and inevitable
extension of denotative into connotative terms, through the force
of merely receptual association. The second consists in the no less
natural and inevitable apposition of denotative terms themselves,
whereby a receptually perceived relation is virtually—though not
conceptually—predicated as subsisting between the objects, qualities,
states, actions, or relations which are denoted. Of course it is
evident that these two modes of development must have mutually assisted
one another: the more that denotative signs underwent connotative
extension, the greater must have been their predicative value when
used in apposition; and the more frequently denotative signs were used
in apposition, the greater must have become the extension of their
connotative value.

Lastly, it is desirable throughout all this hypothetical discussion
to remember that we have the positive evidence of philology touching
two points of considerable importance. The first point is that, as
in the aboriginal sentence-words there was no differentiation of, or
distinction between, subject and predicate; so, until very late in
the evolution of predicative utterance, there was—and in very many
languages still continues to be—an absence of the copula. Nay, even
the substantive verb, which has been unwittingly confounded with the
copula by some of my opponents, was also very late in making its
appearance.

The second point is that, although “pronominal elements”—or verbal
equivalents of gesture-signs indicative of space-relations—were among
the earliest of verbal differentiations, it was not until after æons
of ages had elapsed that any pronouns arose as specially indicative
of the first person.[334] Now, this point I consider one of prime
importance. For it furnishes us with direct evidence of the fact that,
long after mankind had begun to speak, and even long after they had
gained considerable proficiency in the art of articulate language,
the speakers still continued to refer to themselves in that same
kind of objective phraseology as is employed by a child before the
dawn of self-consciousness. This, of course, is what on antecedent
or theoretical grounds we should infer _must have been_ the case;
but it is surely a matter of great moment that our inference on this
point should admit of such full and independent verification at the
hands of philological research. As we have now so repeatedly seen, the
distinction between ideas as receptual and conceptual turns upon the
presence or absence of self-consciousness, in the full or introspective
signification of that term. And, as we have likewise seen, the outward
and visible sign of this inward and spiritual grace is given in the
subjective use of pronominal words. But if these things admit of no
question in the case of an individual human mind—if in the case of
the growing child the rise of self-consciousness is demonstrably
the condition to that of conceptual thought,—by what feat of logic
can it be possible to insinuate that in the growing psychology of
the race there may have been conceptual thought before there was
any true self-consciousness? Obviously this cannot be insinuated
without denying those identical principles of psychology on which my
opponents themselves rely. Will it, then, be said that the criterion
of self-consciousness which is valid for a child is not valid for
the race—that although in the former the rise of self-consciousness
is marked by the change from objective to subjective phraseology, in
the latter a precisely similar change is not to be accredited with a
similar meaning? If this were to be suggested, it would not merely be
quite gratuitous as a suggestion, but directly opposed to the whole of
an otherwise perfectly parallel analogy. In point of fact, then, there
is obviously no escape from the conclusion that in the race, as in the
individual, the development of true, or “inward,” from receptual, or
“outward,” self-consciousness was a gradual process; that its birth in
the former is not merely a matter of inference—overpowering though
this inference be,—but a matter of actual fact which is recorded in
the archives of Language itself; and, therefore, that the central
question upon which the whole of the present treatise has been engaged
cannot any longer be regarded as an open question. It has been closed,
part by part, as the witness of philology has verified, stage by stage,
the results of our psychological analysis; and now, eventually, the
verification has extended to the central core of the matter, revealing
in all its naked simplicity the one decisive fact, that in the
childhood of the world, no less than in that of the man, we may see the
fundamental change from sense to thought: in the one as in the other do
we behold that—

      “As he grows he gathers much,
    And learns the use of ‘I,’ and ‘me,’
    And finds ‘I am not what I see,
  And other than the things I touch.’

  “So rounds he to a separate mind
    From whence clear memory may begin,
    As thro’ the frame that binds him in
  His isolation grows defined.”




CHAPTER XVII.

GENERAL SUMMARY AND CONCLUDING REMARKS.


In the present treatise I take as granted the general theory of
evolution, so far as it is now accepted by the vast majority of
naturalists. That is to say, I assume the doctrine of descent as
regards the whole of organic nature, morphological and psychological,
with the one exception of man. Moreover, I assume this doctrine even
in the case of man, so far as his bodily organization is concerned; it
being thus only with reference to the human mind that the exception to
which I have alluded is made. And I make this exception in deference to
the opinion of that small minority of evolutionists who still maintain
that, notwithstanding their acceptance of the theory of descent as
regards the corporeal constitution of man, they are able to adduce
cogent evidence to prove that the theory fails to account for his
mental constitution.

Such being my basis of assumption, we began by considering the state
of the question _a priori_. If, in accordance with our assumption,
the process of organic and of mental evolution has been continuous
throughout the whole region of life and of mind, with the one exception
of the mind of man, on grounds of an immensely large analogy we
must deem it antecedently improbable that the process of evolution,
elsewhere so uniform and ubiquitous, should have been interrupted at
its terminal phase. And this antecedent presumption is still further
strengthened by the undeniable fact that, in the case of every
individual human being, the human mind presents to actual observation
a process of gradual development, extending from infancy to manhood.
For it is thus shown to be a matter of observable fact that, whatever
may have been the origin or the history of human intelligence in the
past, as it now exists—or, rather, as in every individual case it
now comes into existence—it proves itself to be no exception to the
general law of evolution: it unquestionably does admit of gradual
growth from a zero level, and without such a gradual growth we have no
evidence of its becoming. Furthermore, so long as it is passing through
the lower stages of this growth, the human mind ascends through a
scale of faculties which are parallel with those that are permanently
presented by what I have termed the psychological species of the animal
kingdom—a general fact which tends most strongly to prove that, at
all events up to the time when the distinctively human qualities of
ideation are attained, no difference of kind is apparent between human
and brute psychology. Lastly, not only in the individual, but also in
the race, the phenomena of mental evolution are conspicuous—so far,
at least, as the records of the human race extend. Whether we have
regard to actual history, to tradition, to antiquarian remains, or
flint implements, we obtain uniform evidence of a continuous process of
upward development, which is thus seen to be as characteristic of those
additional attributes wherein the human mind now surpasses that of any
other species as it is of those attributes which it shares with other
species. Therefore, if the process of mental evolution was interrupted
between the anthropoid apes and primitive man during the pre-historic
period of which we have no record, it must again have been resumed with
primitive man, after which it must have continued as uninterruptedly in
the human species as it previously did in the animal species. This, to
say the least, is a most improbable supposition. The law of continuity
is proved to apply on both sides of a psychological interval, where
there happens to be a necessary absence of historical information.
Yet we are asked to believe that, in curious coincidence with this
interval, the law of continuity was violated—notwithstanding that in
the case of every individual human mind such is known never to be the
case.

In order to overturn so immense a presumption as is thus raised against
the contention of my opponents on merely _a priori_ grounds, it appears
to me that they must be fairly called upon to supply some very powerful
considerations of an _a posteriori_ kind, tending to show that there
is something in the constitution of the human mind which renders it
virtually impossible to suppose that such an order of mental existence
can have proceeded by way of genetic descent from mind of lower orders.
I therefore next proceeded to consider the arguments which have been
adduced in support of this thesis.

In order that the points of difference on which these arguments are
founded might be brought out into clear relief, I began by briefly
considering the points of resemblance between the human mind and
mind of lower orders. Here we saw that so far as the Emotions are
concerned no difference of kind has been, or can be, alleged. The
whole series of human emotions have been proved to obtain among the
lower animals, except those which depend on the higher intellectual
powers of man—_i.e._ those appertaining to religion and perception
of the sublime. But all the others—which in my list amount to over
twenty—occur in the brute creation; and although many of them do
not occur in so highly developed a degree, this is immaterial where
the question is one of kind. Indeed, so remarkable is the general
similarity of emotional life in both cases—especially when we have
regard to the young child and savage man—that it ought fairly to be
taken as direct evidence of a genetic continuity between them.

And so, likewise, it is with Instinct. For although this occurs in a
greater proportion among the lower animals than it does in ourselves,
no one can venture to question the identity of all the instincts which
are common to both. And this is the only point that here requires to be
established.

Again, with respect to the Will, no argument can arise touching the
identity of animal and human volition up to the point where the latter
is alleged to take on the attribute of freedom—which, as we saw, under
any view depends on the intellectual powers of introspective thought.

There remain, then, only these intellectual powers of introspective
Thought, _plus_ the faculties of Morality and Religion. Now, it is
evident that, whatever we may severally conclude as touching the
distinctive value of the two latter, we must all agree that a prime
condition to the possibility of either resides in the former: without
the powers of intellect which are competent to frame the abstract
ideation that is concerned both in morals and religion, it is manifest
that neither could exist. Therefore, in logical order, it is these
powers of intellect that first fall to be considered. In subsequent
parts of this work I shall fully deal both with morals and religion: in
the present part I am concerned only with the intellect.

And here it is, as I have acknowledged, that the great psychological
distinction is to be found. Nevertheless, even here it must be conceded
that up to a certain point, as between the brute and the man, there is
not merely a similarity of kind, but an identity of correspondence. The
distinction only arises with reference to those superadded faculties of
ideation which occur above the level marked 28 in my diagram—_i.e._
where the upward growth of animal intelligence ends, and the
development of distinctively human faculty begins. So that in the case
of intellect, no less than in that of emotion, instinct, and volition,
there can be no doubt that the human mind runs exactly parallel with
the animal, up to the place where these superadded powers of intellect
begin to supervene. Therefore, upon the face of them, the facts of
comparative psychology thus far, to say the least, are strongly
suggestive of these superadded powers having been due to a process of
continued evolution.

So much, then, for the points of agreement between animal and human
psychology. Turning next to the points of difference, we had first to
dispose of certain allegations which were either erroneous in fact or
plainly unsound in theory. This involved a rejection _in toto_ of the
following distinctions—namely, that brutes are non-sentient machines;
that they present no rudiments of reason in the sense of perceiving
analogies and drawing inferences therefrom; that they are destitute
of any immortal principle; that they show no signs of progress from
generation to generation; that they never employ barter, make fire,
wear clothes, use tools, and so forth. Among these sundry alleged
distinctions, those which are not demonstrably false in fact are
demonstrably false in logic. Whether or not brutes are destitute of
any immortal principle, and whether or not human beings present such
a principle, the science of comparative psychology has no means of
ascertaining; and, therefore, any arguments touching these questions
are irrelevant to the subject-matter on which we are engaged. Again,
the fact that brutes do not resemble ourselves in wearing clothes,
making fire, &c., clearly depends on an absence in them of those powers
of higher ideation which alone are adequate to yield such products
in the way of intelligent action. All such differences in matters of
detail, therefore, really belong to, or are absorbed by, the more
general question as to the nature of the distinction between the two
orders of _ideation_. To this, therefore, as to the real question
before us, we next addressed ourselves. And here it was pointed out,
_in limine_, that the three living naturalists of highest authority
who still argue for a difference of kind between the brute and the
man, although they agree in holding that only on grounds of psychology
can any such difference be maintained, nevertheless upon these grounds
all mutually contradict one another. For while Mr. Mivart argues that
there must be a distinction of kind, because the psychological interval
between the highest ape and the lowest man is so great; Mr. Wallace
argues for the same conclusion on the ground that this interval is
not so great as the theory of a natural evolution would lead us to
expect: the brain of a savage, he says, is so much more efficient an
instrument than the mind to which it ministers, that its presence can
only be explained as a preparation for the higher efficiency of mental
life as afterwards exhibited by civilized man. Lastly, Professor De
Quatrefages contradicts both the English naturalists by vehemently
insisting that, so far as the powers of intellect are concerned, there
is a demonstrable identity of kind between animal intelligence and
human, whether in the savage or civilized condition: he argues that the
distinction only arises in the domain of morals and religion. So that,
if our opinion on the issue before us were to be in any way influenced
by the voice of authority, I might represent the judgments of these my
most representative opponents as mutually cancelling one another—thus
yielding a zero quantity as against the enormous and self-consistent
weight of authority on the other side.

But, quitting all considerations of authority, I proceeded to
investigate the question _de novo_, or exclusively on its own merits.
To do this it was necessary to begin with a somewhat tedious analysis
of ideation. The general result was to yield the following as my
classification of ideas.

1. Mere memories of perceptions, or the abiding mental images of past
sensuous impressions. These are the ideas which, in the terminology of
Locke, we may designate Simple, Particular, or Concrete. Nowadays no
one questions that such ideas are common to animals and men.

2. A higher class of ideas, which by universal consent are also
common to animals and men; namely, those which Locke called Complex,
Compound, or Mixed. These are something more than the simple memories
of particular perceptions; they are generated by the mixture of such
memories, and therefore represent a compound, of which “particular
ideas” are the elements or ingredients. By the laws of association,
particular ideas which either resemble one another in themselves, or
frequently occur together in experience, tend to coalesce and blend
into one: as in a “composite photograph” the sensitive plate is able
to unite many more or less similar images into a single picture, so
the sensitive tablet of the mind is able to make of many simple or
particular ideas, a complex, a compound, or, as I have called it, a
_generic_ idea. Now, a generic idea of this kind differs from what
is ordinarily called a general idea (which we will consider in the
next paragraph), in that, although both are generated out of simpler
elementary constituents, the former are thus generated as it were
spontaneously or anatomically by the principles of merely perceptual
association, while the latter can only be produced by a consciously
intentional operation of the mind upon the materials of its own
ideation, known as such. This operation is what psychologists term
conception, and the product of it they term a concept. Hence we see
that between the region of percepts and those of concepts there lies a
large intermediate territory, which is occupied by what I have called
generic ideas, or _recepts_. A recept, then, differs from a percept in
that it is a compound of mental representations, involving an orderly
grouping of simpler images in accordance with past experience; while
it differs from a concept in that this orderly grouping is due to an
unintentional or automatic activity on the part of the percipient mind.
A recept, or generic idea, is _imparted to_ the mind by the external
“logic of events;” while a general idea, or concept, is _framed by_ the
mind consciously working to a higher elaboration of its own ideas. In
short, a recept is _received_, while a concept is _conceived_.

3. The highest class of ideas, which psychologists are unanimous in
denying to brutes, and which, therefore, we are justified in regarding
as the unique prerogative of man. These are the General, Abstract,
and Notional ideas of Locke, or the Concepts just mentioned in the
last paragraph. As we have there seen, they differ from recepts—and,
_a fortiori_, from percepts, in that they are themselves the objects
of thought. In other words, it is a peculiarity of the human mind
that it is able to think about its own ideas as such, consciously
to combine and elaborate them, intentionally to develop higher
products out of less highly developed constituents. This remarkable
power we found—also by common consent—to depend on the faculty of
self-consciousness, whereby the mind is able, as it were, to stand
apart from itself, to render one of its states objective to others, and
thus to contemplate its own ideas as such. Now, we are not concerned
with the philosophy of this fact, but only with its history. How it
is that such a faculty as self-consciousness is possible; what it
is that can thus be simultaneously the subject and the object of
thought; whether or not it is conceivable that the great abyss of
personality can ever be fathomed; these and all such questions are
quite alien to the scope of the present work. All that we have here
to do is to analyze the psychological conditions out of which, as a
matter of observable fact, this unique peculiarity emerges—to trace
the history of the process, and tabulate the results. Well, we have
seen that here, again, every one agrees in regarding the possibility
of self-consciousness to be given in the faculty of language. Whether
or not we suppose that these two faculties are one—that neither
could exist without the other, and, therefore, that we may follow the
Greeks in assigning to them the single name of Logos,—at least it is
as certain as the science of psychology can make it, that within the
four corners of human experience a self-conscious personality cannot
be led up to in any other way than through the medium of language. For
it is by language alone that, so far as we have any means of knowing,
a mind is rendered capable of so far fixing—or rendering definite to
itself—its own ideas, as to admit of any subsequent contemplation
of them as ideas. It is only by means of marking ideas by names that
the faculty of conceptual thought is rendered possible, as we saw at
considerable length in Chapter IV.

Such, then, was my classification of ideas. And it is a classification
over which no dispute is likely to arise, seeing that it merely sets in
some kind of systematic order a body of observable facts with regard to
which writers of every school are nowadays in substantial agreement.
Now, if this classification be accepted, it follows that the question
before us is thrown back upon the faculty of language. This faculty,
therefore, I considered in a series of chapters. First it was pointed
out that, in its widest signification, “language” means the faculty of
making signs. Next, I adopted Mr. Mivart’s “Categories of Language,”
which, when slightly added to, serve to give at once an accurate and
exhaustive classification of every bodily or mental act with reference
to which the term can possibly be applied. In all there were found to
be seven of these categories, of which the first six are admittedly
common to animals and mankind. The seventh, however, is alleged by my
opponents to be wholly peculiar to the human species. In other words,
it is conceded that animals do present what may be termed the germ of
the sign-making faculty; but it is denied that they be able, even in
the lowest degree, to make signs of an intellectual kind—_i.e._ of a
kind which consists in the bestowing of names as marks of ideas. Brutes
are admittedly able to make signs to one another—and also to man—with
the intentional purpose of conveying such ideas as they possess;
but, it is alleged, no brute is able to name these ideas, either by
gestures, tones, or words. Now, in order to test this allegation, I
began by giving a number of illustrations which were intended to show
the level that is reached by the sign-making faculty in brutes; next
I considered the language of tone and gesture as this is exhibited by
man; then I proceeded to investigate the phenomena of articulation, the
relation of tone and gesture to words; and, lastly, the psychology of
speech. Not to overburden the present summary, I will neglect all the
subordinate results of this analysis. The main results, however, were
that the natural language of tone and gesture is identical wherever it
occurs; but that even when it becomes conventional (as it may up to a
certain point in brutes), it is much less efficient than articulate
language as an agency in the construction of ideas; and, therefore,
that the psychological line between brute and man must be drawn, not
at language, or sign-making in general, but at that particular kind of
sign-making which we understand by “speech.” Nevertheless, the real
distinction resides in the intellectual powers; not in the symbols
thereof. So that a man means, it matters not by what system of signs
he expresses his meaning. In other words, although I endeavoured to
prove that articulation must have been of unique service in developing
these intellectual powers, I was emphatic in representing that, when
once these powers are present, it is psychologically immaterial
whether they find expression in gesture or in speech. In any case the
psychological distinction between a brute and a man consists in the
latter being able to _mean a proposition_; and the kind of mental act
which this involves is technically termed a “judgment.” Predication,
or the making of a proposition—whether by gesture, tone, speech, or
writing,—is nothing more nor less than the expression of a judgment;
and a judgment is nothing more nor less than the apprehension of
whatever meaning it may be that a proposition serves to set forth.

Now, this is admitted by all my opponents who understand the psychology
of the subject. Moreover, they allow that if once this chasm of
predication were bridged, there would be no further chasm to cross. For
it is universally acknowledged that, from the simplest judgment which
it is possible to make—and, therefore, from the simplest proposition
which it is possible to construct—human intelligence displays an
otherwise uninterrupted ascent through all the grades of excellence
which it afterwards presents. Here, therefore, we had carefully to
consider the psychology of predication. And the result of our analysis
was to show that the distinctively human faculty in question really
occurs further back than at the place where a mind is first able to
construct the formal proposition “A is B.” It occurs at the place
where a mind is first able to bestow a name, known as such,—to call
A _A_, and B _B_, with a cognizance that in so doing it is performing
an act of conceptual classification. Therefore, unless we extend the
term “judgment” so as to embrace such an act of conceptual naming (as
well as the act of expressing a relation between things conceptually
named), we must conclude that “the simplest element of thought” is not
a judgment, but a concept. It is needless again to go over the ground
of this proof; for, although in the course of it I had to point out
certain inexcusable errors in psychological analysis on the part of
some of my opponents, the proof itself is too complete to admit of any
question.

Thus, then, we were brought back to our original distinction between
a concept and a recept. But now we were in a position to show that,
just as in the matter of conducting “inferences,” so in the matter
of making signs, there is an order of ideation that is receptual as
well as one that is conceptual. And, more particularly, even in that
kind of sign-making which consists in the bestowing of names, ideation
of the receptual order may be concerned without any assistance at
all from ideation of the conceptual order. In other words, there are
names and names. Not every name that is bestowed need necessarily
be expressive of a concept, any more than every “inference” that is
conducted need necessarily be the result of self-conscious thought. Not
only young children before they attain to self-conscious thought, but
even talking birds habitually name objects, qualities, actions, and
states. Nevertheless, while giving abundant evidence of this fact, I
was careful to point out that thus far no argumentative implications
of any importance were involved. That a young child and a talking
bird should be able thus to learn the names of objects, qualities,
&c., by imitation—or even to invent arbitrary names of their own—is
psychologically of no more significance than the fact that both the
child and the bird will similarly employ gesture-signs or vocal tones
whereby to express the simple logic of their recepts. Nevertheless,
it is needful in some way to distinguish this non-conceptual kind of
naming from that kind which is peculiar to man after he has attained
self-consciousness, and thus is able, not only to name, but to _know
that he names_—not only to call A _A_, but to _think A as his symbol
of_ A. Now, in order to mark this distinction, I have assigned the term
_denotation_ to naming of the receptual kind, and applied the term
_denomination_ to naming of the conceptual kind. When a parrot calls a
dog “Bow-wow” (as a parrot, like a child, can easily be taught to do),
it may be said in a sense to be naming the dog; but obviously it is not
_predicating_ any characters as belonging to a dog, or performing any
act of _judgment_ with regard to a dog—as is the case, for example,
with a naturalist who, by means of his name _Canis_, conceptually
assigns that animal to a particular zoological genus. Although the
parrot may never utter the name “Bow-wow” save when it sees a dog,
this fact is attributable to the laws of association acting only in
the receptual sphere: it furnishes no shadow of a reason for supposing
that the bird ever thinks about the dog as a dog, or sets the concept
Dog before its mind as a separate object of thought. Therefore, none
of my opponents can afford to deny that in one sense of the word there
may be names without concepts: whether as gestures or as words (“vocal
gestures”), there may be signs of things without these signs presenting
any vestige of predicative value. Now, it is in order not to prejudice
the case of my opponents, and thus clearly to mark out the field of
discussion, that I have instituted the distinction between names as
receptual and conceptual, or denotative and denominative.

This distinction having been clearly understood, the next point was
that both kinds of names admit of connotative extension—denotative
names within the receptual sphere, and denominative within the
conceptual. That is to say, when a name has been applied to one thing,
its use may be extended to another thing, which is seen to belong to
the same class or kind. The degree to which such connotative extension
of a name may take place depends, of course, on the degree in which the
mind is able to take cognizance of resemblances or analogies. Hence
the process can go much further in the conceptual sphere than it does
in the receptual. But the important point is that it unquestionably
takes place in the latter within certain limits. Nor is this anything
more than we should antecedently expect. For in the lengthy account
and from the numerous facts which I gave of the receptual intelligence
of brutes, it was abundantly proved that long before the differential
engine of conception has come to the assistance of mind, mind is
able to reach a high level in the distinguishing of resemblances or
analogies by means of receptual discrimination alone. Consequently, it
is inevitable that non-conceptual or denotative names should undergo a
connotative extension, within whatever limits these powers of merely
receptual discrimination impose. And, as a matter of fact, we found
that such is the case. A talking bird will extend its denotative name
from one dog in particular to any other dog which it may happen to see;
and a young child, after having done this, will extend the denotative
name still further, so as to include images, and eventually pictures,
of dogs. Hence, if the receptual intelligence of a parrot were somewhat
more advanced than it happens to be, we can have no doubt that it would
do the same: the only reason why in this matter it parts company with
a child so soon as it does, is because its receptual intelligence is
not sufficiently developed to perceive the resemblance of images and
pictures to the objects which they are intended to represent. But the
receptual intelligence of a dog is higher than that of a parrot, and
some dogs are able to perceive resemblances of this kind. Therefore
if dogs, like parrots, had happened to be able to articulate, and so
to learn the use of denotative names, there can be no doubt that they
would have accompanied the growing child through a somewhat further
reach of connotative utterance than is the case with the only animals
which present the anatomical conditions required for the imitation of
articulate sounds. Both dogs and monkeys are able, in an extraordinary
degree, to _understand_ these sounds: that is to say, they can learn
the meanings of an astonishing number of denotative names, and also
be taught to apprehend a surprisingly large extension of connotative
significance. Consequently, if they could but _imitate_ these sounds,
after the manner of a parrot, it is certain that they would greatly
distance the parrot in this matter of receptual connotation.

But, lastly, we are not shut up to any such hypothetical case. For
the growing child itself furnishes us with evidence upon the point,
which is no less cogent than would be the case if dogs and monkeys were
able to talk. For, without argumentative suicide, none of my opponents
can afford to suggest that, up to the age when self-consciousness
dawns, the young child is capable of conceptual connotation; yet it is
unquestionable that up to that age a continuous growth of connotation
has been taking place, which, beginning with the level that it shares
with a parrot, is eventually able to construct what I have called
“receptual propositions,” the precise nature of which I will summarise
in a subsequent paragraph. The evidence which I have given of this
connotative extension of denotative names by children before the age
at which self-consciousness supervenes—and, therefore, _prior to
the very condition which is required for conceptual ideation_—is, I
think, overwhelming. And I do not see how its place in my argument
can be gainsaid by any opponent, except at the cost of ignoring my
distinction between connotation as receptual and conceptual. Yet
to do this would be to surrender his whole case. Either there is a
distinction, or else there is not a distinction, between connotation
that is receptual, and connotation that is conceptual. If there is no
distinction, all argument is at an end: the brute and the man are one
in kind. But I allow that there is a distinction, and I acknowledge
that the distinction resides where it is alleged to reside by my
opponents—namely, in the presence or absence of self-consciousness
on the part of a mind which bestows a name. Or, to revert to my own
terminology, it is the distinction between denotation and denomination.

Now, in order to analyze this distinction, it became needful further
to distinguish between the highest level of receptual ideation that
is attained by any existing brute, and those further developments of
receptual ideation which are presented by the growing child, after it
parts company with all existing brutes, but before it assumes even
the lowest stage of conceptual ideation—_i.e._ prior to the dawn of
self-consciousness. This subordinate distinction I characterized by the
terms “lower recepts” and “higher recepts.” Already I had instituted
a distinction between “lower concepts” and “higher concepts,” meaning
by the former the conceptual naming of recepts, and by the latter a
similar naming of other concepts. So that altogether four large and
consecutive territories were thus marked out: (1) Lower Recepts, which
are co-extensive with the psychology of existing animals, including
a very young child; (2) Higher Recepts, which occupy a psychological
area between the recepts of animals and the first appearance of
self-consciousness in man; (3) Lower Concepts, which are concerned only
with the self-conscious naming of recepts; (4) Higher Concepts, which
have to do with the self-conscious classification of other concepts
known as such, and the self-conscious naming of such ideal integrations
as may result therefrom.

Now, if all this is true of naming, clearly it must also be true of
judging. If there is a stage of pre-conceptual naming (denotation),
there must also be a stage of pre-conceptual judgment, of which such
naming is the expression. No doubt, in strictness, the term judgment
should be reserved for conceptual thought (denomination); but, in order
to avoid an undue multiplication of terms, I prefer thus to qualify the
existing word “judgment.” Such, indeed, has already been the practice
among psychologists, who speak of “intuitive judgments” as occurring
even in acts of perception. All, therefore, that I propose to do is to
institute two additional classes of non-conceptual judgment—namely,
lower receptual and higher receptual, or, more briefly, receptual and
pre-conceptual. If one may speak of an “intuitive,” “unconscious,” or
“perceptual” judgment (as when we mistake a hollow bowl for a sphere),
much more may we speak of a receptual judgment (as when a sea-bird
dives from a height into water, but will not do so upon land), or a
pre-conceptual judgment (as when a young child will extend the use of
a denotative name without any denominative conception). In all, then,
we have four phases of ideation to which the term judgment may be thus
either literally or metaphorically applied—namely, the perceptual,
receptual, pre-conceptual, and conceptual. Of these the last only
is judgment, properly so called. Therefore I do not say that a brute
really judges when, without any self-conscious thought, it brings
together certain reminiscences of its past experience in the form
of recepts, and translates for us the result of its ideation by the
performance of what Mr. Mivart calls “practical inferences.” Neither
do I say that a brute really judges when, still without self-conscious
thought, it learns correctly to employ denotative names. Nay, I should
deny that a brute really judges even if, after it is able to denotate
separately two different recepts (as is done by a talking bird), it
were to name these two recepts simultaneously when thus combined in an
act of “practical inference.” Although there would then be the outward
semblance of a proposition, we should not be strictly right in calling
it a proposition. It would, indeed, be the _statement of a truth
perceived_; but not the statement of a truth perceived _as true_.

Now, if all this be admitted in the case of a brute—as it must be
by any one who takes his stand on the faculty of true or conceptual
judgment,—obviously it must also be admitted in the case of the
growing child. In other words, if it can be proved that a child is able
to state a truth before it is able to state a truth as true, it is
thereby proved that in the psychological history of every human being
there is first the kind of predication which is required for dealing
with receptual knowledge, or for the stating of truths perceived;
and next the completed judgment which is required for dealing with
conceptual knowledge, or of stating truths perceived as true. Of course
the condition required for the raising of this lower kind of judgment
and this lower kind of predication (if, for the sake of convenience, we
agree to use these terms) into the higher or only true kind of judgment
and predication, is the advent of self-consciousness. Or, in other
words, the place where a mere statement of truth first passes into a
real predication of truth, is determined by the place at which there
first supervenes the faculty of introspective reflection. The whole
issue is thus reduced to an analysis of self-consciousness. To this
analysis, therefore, we next addressed ourselves.

Seeing that the faculty in question only occurs in man, obviously
it is only in the case of man that any material is supplied for the
analysis of it. Moreover, as previously remarked, so far as this our
analysis is concerned, we have only to deal with the psychology of
self-consciousness: we are not concerned with its philosophy. Now,
as a matter of psychology, no one can possibly dispute that the
faculty in question is one of gradual development; that during the
first two or three years of the growing intelligence of man there is
no vestige of any such faculty at all; that when it does begin to
dawn, the human mind is already much in advance of the mind of any
brute; but that, even so, it is much less highly developed than it
is afterwards destined to become; and that the same remark applies
to the faculty of self-consciousness itself. Furthermore, it will be
granted that self-consciousness consists in paying the same kind of
attention to internal, or psychical processes, as is habitually paid
to external, or physical processes—although, of course, the degrees
in which such attention may be yielded are as various in the one case
as in the other. Lastly, it will be further granted that in the minds
of brutes, as in the minds of men, there is a world of images, or
recepts; and that the only reason why in the former case these images
are not attended to unless called up by the sensuous association of
their corresponding objects, is because the mind of a brute is not
able to leave the ground of such merely sensuous association, so as
to move through the higher and more tenuous region of introspective
thought. Nevertheless, I have proved that this image-world, even in
brutes, displays a certain amount of internal activity, which is not
wholly dependent on sensuous associations supplied from without.
For the phenomena of “home-sickness,” pining for absent friends,
dreaming, hallucination, &c., amply demonstrate the fact that in our
more intelligent domesticated animals there may be an internal (though
unintentional) play of ideation, wherein one image suggests another,
this another, and so on, without the need of any immediate associations
supplied from present objects of sense. Furthermore, I have pointed
out that receptual ideation of this kind is not restricted to the
images of sense-perception; but is largely concerned with the mental
states of other animals. That is to say, the logic of recepts, even in
brutes, is sufficient to enable the mind to establish true analogies
between subjective states and the corresponding states of other
intelligences: animals habitually and accurately interpret the mental
states of other animals, while also well knowing that other animals
are able similarly to interpret theirs. Hence, it must be further
conceded that intelligent animals recognize a world of ejects, as well
as a world of objects: mental existence is known to them ejectively,
though, as I allow, never thought upon subjectively. At this stage of
mental evolution the individual—whether an animal or an infant—so
far realizes its own individuality as to be informed by the logic
of recepts that it is one of a kind, although of course it does not
recognize either its own or any other individuality as such.

Nevertheless, there is thus given a rudimentary or nascent form
of self-consciousness, which up to the stage of development
that it attains in a brute or an infant may be termed receptual
self-consciousness; while in the more advanced stages which it presents
in young children it may be termed pre-conceptual self-consciousness.
Pre-conceptual self-consciousness is exhibited by all children after
they have begun to talk, but before they begin to speak of themselves
in the first person, or otherwise to give any evidence of realizing
their own existence as such. Later on, when true self-consciousness
does arise, the child, of course, is able to do this; and then only
is supplied the condition _sine quâ non_ to a reflection upon its own
ideas—hence to a knowledge of names as names, and so to a statement
of truths as true. But long before this stage of true or conceptual
self-consciousness is reached—whereby alone is rendered possible true
or conceptual predication—the child, in virtue of its pre-conceptual
self-consciousness, is able to make known its wants, and otherwise to
communicate its ideas, by way of pre-conceptual predication. I gave
many instances of this pre-conceptual predication, which abundantly
proved that the pre-conceptual self-consciousness of which it is the
expression amounts to nothing more than a practical recognition of self
as an active and feeling agent, without any introspective recognition
of that self as an object of knowledge.

Given, then, this stage of mental evolution, and what follows? The
child, like the animal, is supplied by its logic of recepts with a
world of images, standing as signs of outward objects; with an ejective
knowledge of other minds, and with that kind of recognition of self
as an active, suffering, and accountable agent to which allusion
has just been made. But, over and above the animal, the child has
now at its command a much more improved machinery of sign-making,
which, as we have before seen, is due to the higher evolution of its
receptual ideation. Now among the contents of this ideation is a better
apprehension of the mental states of other human beings, together with
a greatly increased power of denotative utterance, whereby the child is
able to name receptually such ejective states as it thus receptually
apprehends. These, therefore, severally receive their appropriate
denotations, and so gain clearness and precision as ejective images
of the corresponding states experienced by the child itself. “Mamma
pleased to Dodo” would have no meaning as spoken by a child, unless
the child knew from his own feelings what is the state of mind which
he thus ejectively attributes to his mother. Hence, we find that at
the same age the child will also say “Dodo pleased to mamma.” Now it
is evident that we are here approaching the very borders of true or
conceptual self-consciousness. The child, no doubt, is still speaking
of himself in objective phraseology; but he has advanced so far in the
interpretation of his own states of mind as clearly to name them, in
the same way as he would name any external objects of sense-perception.
Thus is he enabled to fix these states before his mental vision as
things which admit of being denoted by verbal signs, although as yet
he has never thought about either the states of mind or his names for
them _as such_, and, therefore, has not yet attained to the faculty
of denomination. But the interval between denotation and denomination
has now become so narrow that the step from recognizing “Dodo” as not
only the object, but also the subject of mental changes, is rendered at
once easy and inevitable. The mere fact of attaching verbal signs to
mental states has the effect of focussing attention upon those states;
and when attention is thus focussed habitually, there is supplied the
only further condition which is required to enable a mind, through its
memory of previous states, to compare its past with its present, and so
to reach that apprehension of continuity among its own states wherein
the full introspective, or conceptual consciousness of self consists.

Several subordinate features in the evolution of this conceptual from
pre-conceptual self-consciousness were described; but it is needless
again to mention them. Enough has been here said to show ample grounds
for the conclusions which my chapter on “Self-consciousness” was
mainly concerned in establishing—namely, that language is quite as
much the antecedent as it is the consequent of self-consciousness;
that pre-conceptual predication is indicative of a pre-conceptual
self-consciousness; and that from these there naturally and inevitably
arise those higher powers of conceptual predication and conceptual
self-consciousness on which my opponents (disregarding the phases that
lead up to them) have sought to rear their alleged distinction of kind
between the brute and the man.

Thus, as a general result of the whole inquiry so far, we may say
that throughout the entire range of mental phenomena we have found
one and the same distinction to obtain between the faculties of
mind as perceptual, receptual, and conceptual. Percept, Recept, and
Concept; Perceptual Judgment, Receptual Judgment, and Conceptual
Judgment; Indication, Denotation, and Denomination;—these are all
manifestations, in different regions of psychological inquiry, of the
same psychological distinctions. And we have seen that the distinction
between a Recept and a Concept, which is thus carried through all
the fabric of mind, is really the only distinction about which there
can be any dispute. Moreover, we have seen that the distinction
is on all hands allowed to depend on the presence or absence of
self-consciousness. Lastly, we have seen that even in the province of
self-consciousness itself the same distinction admits of being traced:
there is a form of self-consciousness which may be termed receptual, as
well as that which may be termed conceptual. The whole question before
us thus resolves itself into an inquiry touching the relation between
these two forms of self-consciousness: is it or is it not observable
that the one is developmentally continuous with the other? Can we or
can we not perceive that in the growing child the powers of receptual
self-consciousness, which it shares with a brute, pass by slow and
natural stages into those powers of conceptual self-consciousness which
are distinctive of a man?

This question was fully considered in Chapter XI. I had previously
shown that so far as the earliest, or indicative phase of language is
concerned, no difference even of degree can be alleged between the
infant and the animal. I had also shown that neither could any such
difference be alleged with regard to the earlier stages of the next
two phases—namely, the denotative and the receptually connotative.
Moreover, I had shown that no difference of kind could be alleged
between this lower receptual utterance which a child shares with a
brute, and that higher receptual utterance which it proceeds to develop
prior to the advent of self-consciousness. Lastly, I had shown that
this higher receptual utterance gives to the child a psychological
instrument whereby to work its way from a merely receptual to an
incipiently conceptual consciousness of self. Such being the state
of the facts as established by my previous analysis, I put to my
opponents the following dilemma. Taking the case of a child about two
years old, who is able to frame such a rudimentary, communicative, or
pre-conceptual proposition as “Dit ki” (Sister is crying), I proceeded
thus.

“Dit” is the denotative name of one recept, “ki” the denotative name of
another: the object and the action which these two recepts severally
represent happen to occur together before the child’s observation:
the child, therefore, denotes them simultaneously—_i.e._ brings
them into _apposition_. The apposition in consciousness of these two
recepts, with their corresponding denotations, is thus effected _for_
the child by the logic of events: it is not effected _by_ the child in
the way of any intentional or self-conscious grouping of its ideas,
such as we have seen to be the distinguishing feature of the logic of
concepts. Here, then, comes the dilemma. For I say, either you here
have conceptual judgment, or else you have not. If you say that this
is conceptual judgment, you destroy the basis of your own distinction
between man and brute, because then you must also say that brutes
conceptually judge—the child as yet not having attained to conceptual
self-consciousness. If, on the other hand, you say that here you have
not conceptual judgment, inasmuch as you have not self-consciousness,
I ask at what stage in the subsequent development of the child’s
intelligence you would consider conceptual judgment to arise. Should
you answer that it first arises when conceptual self-consciousness
first supplies the condition to its arising, I must refer you to the
proof already given that the advent of self-consciousness is itself a
gradual process, the precedent conditions of which are supplied far
down in the animal series. But if this is so, where the faculty of
stating a truth perceived passes into the higher faculty of perceiving
the truth as true, there is a continuous series of gradations
connecting the one faculty with the other. Up to the point where this
continuous series of gradations begins, the mind of the child is, as I
have already proved, indistinguishable from the mind of an animal by
any one principle of psychology. Will you, then, maintain that up to
this time the two orders of psychical existence are identical in kind,
but that during its ascent through this final series of gradations the
human intelligence becomes distinct in kind from that of animals, and
_therefore also from its own previous self_? If so, your argument here
ends in a contradiction.

In confirmation of this my general argument, two subsidiary
considerations were then added. The first was that although the
advance to true self-consciousness from lower grades of mental
development is no doubt a very great and important matter, still
it is not so great and important in comparison with what this
development is afterwards destined to become, as to make us feel that
it constitutes any distinction _sui generis_—or even, perhaps, the
principal distinction—between the man and the brute. For even when
self-consciousness does arise, and has become fairly well developed,
the powers of the human mind are still in an almost infantile
condition. In other words, the first genesis of true self-consciousness
marks a comparatively low level in the evolution of the human mind—as
we might expect that it should, if its genesis depends upon, and
therefore lies so near to, those precedent conditions in merely animal
psychology to which I have assigned it. But, if so, does it not
follow that, great as the importance of self-consciousness afterwards
proves to be in the development of distinctively human ideation,
in itself, or in its first beginning, it does not betoken any very
perceptible advance upon those powers of pre-conceptual ideation which
it immediately follows? There is thus shown to be even less reason
for regarding the first advent of conceptual self-consciousness as
marking a psychological difference of kind, than there would be so
to regard the advent of those higher powers of conceptual ideation
which subsequently—though as gradually—supervene between early
childhood and youth. Yet no one has hitherto ventured to suggest that
the intelligence of a child and the intelligence of a youth display a
difference of kind.

The second subsidiary consideration which I adduced was, that even
in the case of a fully developed self-conscious intelligence, both
receptual and pre-conceptual ideation continue to play an important
part. The vast majority of our verbal propositions are made for the
practical purposes of communication, or without the mind pausing to
contemplate the propositions in the light of self-consciousness. No
doubt in many cases, or in those where highly abstract ideation is
concerned, this independence of the two faculties is more apparent
than real: it arises from each having undergone so much elaboration
by the assistance which it has derived from the other, that both are
now in possession of a large body of organized material on which to
operate, without requiring, whenever they are exercised, to build up
the structure of this material _ab initio_. When I say “Heat is a mode
of motion,” I am using what is now to me a mere verbal sign, which
expresses an external fact: I do not require to examine my own ideas
upon the abstract relation which the proposition sets forth, although
for the original attainment of these ideas I had to exercise many and
complex efforts of conceptual thought. But although I hold this to be
the true explanation of the apparent independence of predication and
introspection in all cases of highly abstract thought, I am convinced,
on the ground of adequate reasons given, that in all cases where
those lower orders of ideation are concerned to which I have so often
referred as receptual and pre-conceptual, the independence is not only
apparent, but real. Now, if the reasons which I have assigned for this
conclusion are adequate—and they are reasons sanctioned by Mill,—it
follows that the ideation concerned in ordinary predication becomes so
closely affiliated with that which is expressed in the lower levels
of sign-making, that even if the connecting links were not supplied
by the growing child, no one would be justified, on psychological
grounds alone, in alleging any difference of kind between one level and
another. The object of all sign-making is communication, and from our
study of the lower animals we know that communication first has to do
exclusively with recepts, while from our study of the growing child we
know that it is the signs used in the communication of recepts which
first lead to the formation of concepts. For concepts are first of all
named recepts, known as such; and we have seen in previous chapters
that this kind of knowledge (_i.e._ of names as names) is rendered
possible by introspection, which, in turn, is reached by the naming of
self as an agent. But even after the power of conceptual introspection
has been fully reached, demand is not always made upon it for the
communication of merely receptual knowledge; and therefore it is that
not every proposition requires to be introspectively contemplated as
such before it can be made. Given the power of denotative nomination on
the one hand, and the power of even the lowest degree of connotative
nomination on the other, and all the conditions are furnished to
the formation of non-conceptual statements, which differ from true
propositions only in that they do not themselves become objects of
thought. And the only difference between such a statement when made by
a young child, and the same statement when similarly made by a grown
man, is that in the former case it is not even _potentially_ capable of
itself becoming an object of thought.

       *       *       *       *       *

The investigation having been thus concluded so far as comparative
psychology was concerned, I next turned upon the subject the
independent light of comparative philology. Whereas we had hitherto
been dealing with what on grounds of psychological analysis alone
we might fairly infer were the leading phases in the development of
distinctively human ideation, we now turned to that large mass of
direct evidence which is furnished by the record of Language, and is
on all hands conceded to render a kind of unintentional record of the
pre-historic progress of this ideation.

The first great achievement of comparative philology has been that of
demonstrating, beyond all possibility of question, that language as
it now exists did not appear ready-made, or by way of any specially
created intuition. Comparative philology has furnished a completed
proof of the fact that language, as we now know it, has been the result
of a gradual evolution. In the chapter on “Comparative Philology,”
therefore, I briefly traced the principles of language growth, so far
as these are now well recognized by all philologists. It was shown,
as a matter of classification, that the thousand or more existing
languages fall into about one hundred families, all the members of each
family being more or less closely allied, while members of different
families do not present evidence of genetic affinity. Nevertheless,
these families admit of being comprised under larger groups or
“orders,” in accordance with certain characteristics of structure, or
type, which they present. Of these types all philologists are agreed
in distinguishing between the Isolating, the Agglutinating, and the
Inflectional. Some philologists make a similar distinction between
these and the Polysynthetic, while all are agreed that from the
agglutinative the Incorporating type has been derived, and from the
inflectional the Analytic.

Passing on from classification to phylogeny, we had to consider the
question of genetic relationship between the three main orders, _inter
se_, and also between the Polysynthetic type and the Agglutinating.
The conflict of authoritative opinion upon this question was shown to
have no bearing upon the subject-matter of this treatise, further than
to emphasize the doctrine of the polyphylectic origin of language—the
probability appearing to be that, regarded as types, both the isolating
and the polysynthetic are equally archaic, or, at all events, that they
have been of equally independent growth. In this connection I adduced
the hypothesis of Dr. Hale, to the effect that the many apparently
independent tongues which are spoken by different native tribes of
the New World, may have been in large part due to the inventions
of accidentally isolated children. The curious correlation between
multiplicity of independent tongues and districts favourable to the
life of unprotected children—in Africa as well as in America—seemed
to support this hypothesis; while good evidence was given to show that
children, if left much alone, do invent for themselves languages which
have little or no resemblance to that of their parents.

Without recapitulating all that was said upon the phases and causes of
linguistic evolution in its various lines of descent, it will be enough
to remind the reader that in every case the result of philological
inquiry is here the same—namely, to find that languages become simpler
in their structure the further they are traced backwards, until we
arrive at their so-called “roots.” These are sometimes represented as
the mysterious first principles of language, or even as the aboriginal
_data_ whose origin is inexplicable. As a matter of fact, however,
these roots are nothing more than the ultimate results of philological
analysis: in no other sense than this can they be supposed “primary.”
Seeing, then, that these roots represent the materials of language
up to the place where the evolution of language no longer admits of
being clearly traced, it is evident that their antecedents, whatever
they may have been, necessarily lie beyond the reach of philological
demonstration, as distinguished from philological inference. This,
of course, is what an evolutionist knows antecedently _must be the
case somewhere_ in the course of any inquiry touching the process of
evolution, wherever he may have occasion to trace it. For the further
he is able to trace it, the nearer must he be coming to the place where
the very material which he is investigating has taken its origin;
and as it is this material itself which furnishes the evidences of
evolution, when it has been traced back to its own origin, the inquiry
reaches a vanishing point. Adopting the customary illustration of a
tree, we might say that when a philologist has traced the development
of the leaves from the twigs, the twigs from the branches, the branches
from the stems, and the stems from the roots, he has given to the
evolutionist all the evidence of evolution which in this particular
line of inquiry is antecedently possible. The germ of ideation out of
which the roots developed must obviously lie beyond the reach of the
philologist as such; and if any light is to be thrown upon the nature
of this germ, or if any evidence is to be yielded of the phases whereby
the germ gave origin to the roots, this must be done by some other
lines of inquiry finding similar germs giving rise to similar products
elsewhere. In the present instance, the only place where we can look
for such parallel processes of evolution is in the case of the growing
child, which I have already considered.

Here, then, we are in the presence of exactly the same distinction with
regard to the origin of Language, as we were at the beginning of this
treatise with regard to the origin of Man. For we there saw that, while
we have the most cogent historical proof of the principles of evolution
having governed the progress of civilization, we have no such direct
proof of the descent of man from a brutal ancestry. And here likewise
we find that, so long as the light of philology is able to guide us,
there can be no doubt that the principles of evolution have determined
the gradual development of languages, in a manner strictly analogous
to that in which they have determined the ever-increasing refinement
and complexity of social organizations. Now, in the latter case we saw
that such direct evidence of evolution from lower to higher levels of
culture renders it well-nigh certain that the method must have extended
backwards beyond the historical period; and hence that such direct
evidence of evolution uniformly pervading the historical period in
itself furnishes a strong _primâ facie_ presumption that this period
was itself reached by means of a similarly gradual development of human
faculty. And thus, also, it is in the case of language. If philology
is able to prove the fact of evolution in all known languages as far
back as the primitive roots out of which they have severally grown, the
presumption becomes exceedingly strong that these earliest and simplest
elements, like their later and more complex products, were the result
of a natural growth. Or, in the words already quoted from Geiger,
we cannot forbear concluding that language must once have had no
existence at all. Nevertheless, it is important to distinguish between
demonstrated fact and speculative inference, however strong; and,
therefore, I began by stating the stages of evolution through which
languages are now known to have passed from the root-stage upwards.
Having done this, I proceeded to consider the question touching the
origin of these roots themselves.

First, as to their number, we found that the outside estimate, in the
younger days of philological research, gave one thousand as a fair
average of the roots which go to feed any living language; but that
this estimate might now be safely reduced by three-fourths. Indeed,
in his latest work, Professor Max Müller professes to have reduced
the roots of Sanskrit to as low a number as 121, and thinks that even
this is excessive. Regarding the character of roots, we saw that some
philologists look upon them as the actual words which were used by
the pre-historic speakers, who, therefore, “talked with one another
in single syllables, indicative of ideas of prime importance, but
wanting all designation of their relations.”[335] On the other hand,
it is now the generally accepted belief, that “roots are the phonetic
and significant types discovered by the analysis of the comparative
philologist as common to a group of allied words,”[336]—or, as it
were, composite phonograms of families of words long since extinct as
individuals. We saw, however, that this difference of opinion among
philologists does not affect the present inquiry, seeing that even
the phonetic-type theory does not question that the unknown words out
of the composition of which a root is now extracted must have been
genetically allied with one another, and exhibited the closeness of
their kinship by a close similarity of their sounds.

A much more important question for us is the character of these roots
with respect to their significance. In this connection we found that
they indicate what Professor Max Müller calls “general ideas,” or
“concepts;” bear testimony to an already and, comparatively speaking,
advanced stage of social culture; are all expressive either of actions
or states; and betray no signs of imitative origin. Taking each of
these characters separately, we found that although all the 121 roots
of Sanskrit are expressive of general ideas, the order of generality
is so low as for the most part to belong to that which I had previously
called “lower concepts,” or “named recepts.” Next, that they all bear
intrinsic testimony to their own comparatively recent origin, and,
therefore, are “primitive” only in the sense of representing the last
result of philological analysis: they certainly are very far from
primitive in the sense of being aboriginal. Again, that they are all
of the nature of verbs was shown to be easily explicable; and, lastly,
the fact that none of them betray any imitative source is not to be
wondered at, even on the supposition that onomatopœia entered largely
into the composition of aboriginal speech. For, on the one hand, we
saw that in the struggle for existence among aboriginal and early
words, those only could have stood any chance of survival—_i.e._ of
leaving progeny—which had attained to some degree of connotative
extension, or “generality;” and, on the other hand, that in order to
do this an onomatopoetic word must first have lost its onomatopoetic
significance. A large body of evidence was adduced in support of the
onomatopoetic theory, and certain objections which have been advanced
against it were, I think, thoroughly controverted. Later on, however,
we saw that the question as to the degree in which onomatopœia entered
in to the construction of aboriginal speech is really a question of
secondary interest to the evolutionist. Whether in the first instance
words were all purely arbitrary, all imitative, or some arbitrary and
some imitative,—in any case the course of their subsequent evolution
would have been the same. By connotative extension in divergent lines,
meanings would have been progressively multiplied in those lines
through all the progeny of ever-multiplying terms—just in the same
way as we find to be the case in “baby-talk,” and as philologists have
amply proved to be the case with the growth of languages in general.

That speech from the first should have been concerned with the naming
of generic ideas, or higher recepts, as well as with particular objects
of sense, is what the evolutionist would antecedently expect. It must
be remembered that the kind of classification with which recepts are
concerned is that which lies nearest to the automatic groupings of
sensuous perception: it depends on an absence of any power analytically
to distinguish less perceptible points of difference among more
conspicuous points of resemblance—or non-essential analogies among
essential analogies with which they happen to be frequently associated
in experience. On the other hand, the kind of classification with
which concepts are concerned is that which lies furthest from the
automatic groupings of sensuous perception: it depends on the power
of analytically distinguishing between essentials and non-essentials
among resemblances which occur associated together in experience.
Classification there doubtless is in both cases; but in the one it is
due to the obviousness of analogies, while in the other it is due to
the mental dissociation of analogies as apparent and real. Or else, in
the one case it is due to constancy of association in experience of the
objects, attributes, actions, &c., classified; while in the other case
it is due to a conscious disregard of such association.

Now, if we remember these things, we can no longer wonder that the
palæontology of speech should prove early roots to have been expressive
of “generic,” as distinguished from “general” ideas. The naming of
actions and processes so habitual, or so immediately apparent to
perception, as those to which the “121 concepts” tabulated by Professor
Max Müller refer, does not betoken an order of ideation very much
higher than the pre-conceptual, in virtue of which a young child is
able to give expression to its higher receptual life, prior to the
advent of self-consciousness. In view of these considerations, my only
wonder is that the 121 root-words do not present _better_ evidence of
conceptual thought. This, however, only shows how comparatively small a
part self-conscious reflection need play in the practical life of early
man, even when so far removed from the really “primitive” condition of
hitherto wordless man as was that of the pastoral people who have left
this record of ideation in the roots of Aryan speech.

After having thus explained the absence of words significant of
“particular ideas” among the roots of existing language, as well
as the generic character of those which the struggle for existence
has permitted to come down to us, we went on to consider sundry
other corroborations of our previous analysis which are yielded
by the science of philology. First we saw that this science has
definitely proved two general facts with regard to the growth of
predication—namely, that in all the still existing radical languages
there is no distinction between noun, adjective, verb, or particle;
and that the structure of all other languages shows this to have been
the primitive condition of language-structure in general: “every
noun and every verb was originally by itself a complete sentence,”
consisting of a subject and predicate fused into one—or rather, let
us say, not yet differentiated into the _two_, much less into the
_three_ parts which now go to constitute the fully evolved structure
of a proposition. Now, this form of predication is “condensed” only
because it is undeveloped; it is the undifferentiated protoplasm of
predication, wherein the “parts of speech” as yet have no existence.
And just as this, the earliest stage of predication, is distinctive
of the pre-conceptual stage of ideation in a child, so it is of the
pre-conceptual ideation of the race. Abundant evidence was therefore
given of the gradual evolution of predicative utterance, _pari
passu_ with conceptual thought—evidence which is woven through the
whole warp and woof of every language which is now spoken by man. In
particular, we saw that pronouns were originally words indicative
of space relations, and strongly suggestive of accompanying acts of
pointing—“I” being equivalent to “this one,” “He” to “that one,” &c.
Moreover, just as the young child begins by speaking of itself in the
third person, so “Man regarded himself as an object before he learnt
to regard himself as a subject,”[337] as is proved by the fact that
“the objective cases of the personal as well as of the other pronouns,
are always older than the subjective.”[338] Pronominal elements
afterwards became affixed to nouns and verbs, when these began to be
differentiated from one another; and thus various applications of a
primitive and highly generalized noun or verb were rendered by means
of these elements, which, as even Professor Max Müller allows, “must
be considered as remnants of the earliest and almost pantomimic phase
of language, in which language was hardly as yet what we mean by
language, namely _logos_, a gathering, but only a pointing.” Similarly,
Professor Sayce remarks of this stage in the evolution of predicative
utterance—which, be it observed, is precisely analogous to that
occupied by a young child whose highly generalized words require to
be assisted by gestures—“It is certain that there was a time in the
history of speech when articulate or semi-articulate sounds uttered
by primitive man were made the significant representations of thought
by the gestures with which they were accompanied: and this complex of
sound and gesture—a complex in which, be it remembered, the sound had
no meaning apart from the gesture—was the earliest sentence.” Thus it
was that “grammar has grown out of gesture”—different parts of speech,
with the subsequent commencements of declension, conjugation, &c.,
being all so many children of gesticulation: but when in subsequent
ages the parent was devoured by this youthful progeny, they continued
to pursue an independent growth in more or less divergent lines of
linguistic development.

For instance, we have abundant evidence to prove that, even after
articulate language had gained a firm footing, there was no distinction
between the nominative and genitive cases of substantives, nor between
these and adjectives, nor even between any words as subject-words
and predicate-words. All these three grammatical relations required
to be expressed in the same way, namely, by a mere apposition of the
generalized terms themselves. In course of time, however, these three
grammatical differentiations were effected by conventional changes
of position between the words apposed, in some cases the form of
predication being A B, and that of attribution or possession B A, while
in other branches of language-growth the reverse order has obtained.
Eventually, however, “these primitive contrivances for distinguishing
between the predicate, the attribute, and the genitive, when the three
ideas had in course of ages been evolved by the mind of the speaker,
gradually gave way to the later and more refined machinery of suffixes,
auxiliaries, and the like.”[339]

And so it is with all the other so-called “parts of speech,” in
those languages which, in having passed beyond the primitive stage,
have developed parts of speech at all. “These are the very broadest
outlines of the process by which conceptual roots were predicated, by
which they came under the sway of the categories—became substantives,
adjectives, adverbs, and verbs, or by whatever other names the results
thus obtained may be described. The minute details of this process, and
the marvellous results obtained by it, can be studied in the grammar of
every language or family of languages.”[340] Thus, philology is able to
trace back, stage by stage, the form of predication as it occurs in the
most highly developed, or inflective language, to that earliest stage
of language in general, which I have called the indicative.

Many other authorities having been quoted in support of these general
statements, and also for the purpose of tracing the evolution of
predicative utterance in more detail, I proceeded to give illustrations
of different phases of its development in the still existing languages
of savages; and thus proved that they, no less than primitive man, are
unable to “supply the blank form of a judgment,” or to furnish what
my opponents regard as the criterion of human faculty. Therefore, the
only policy which can possibly remain for these opponents to take up,
is that of abandoning their Aristotelian position: no longer to take
their stand upon the grounds of purely _formal_ predication as this
happens to have been developed in the Indo-European branch of language;
but altogether upon those of _material_ predication, or, as I may say,
upon the meaning or substance of a judgment, as distinguished from its
grammar or accidents.

In other words, it may possibly still be argued that, although the
issue is now thrown back from the “blank form” of predication on which
my opponents have hitherto relied, to the hard fact of predication
itself, this hard fact still remains. Even though I have shown that in
the absence of any parts of speech predication requires to be conducted
in a most inefficient manner; still, it may be said, predication _is_
conducted, and _must be_ conducted—for assuredly it is only in order
to conduct it that speech can ever have existed at all.

Now, I showed that if my opponents do not adopt this change of
position, their argument is at an end. For I proved that, after
all the foregoing evidence, there is no longer any possibility of
question touching the continuity of growth between the predicative
germ in a sentence-word, and the fully evolved structure of a formal
proposition. But, on the other hand, I next showed that this change
of position, even if it were made, could be of no avail. For, if the
term “predication” be thus extended to a “sentence-word,” it thereby
becomes deprived of that distinctive meaning upon which alone the
whole argument of my adversaries is reared: it is conceded that no
distinction obtains between speaking and pointing: the predicative
phase of language has been identified with the indicative: man and
brute are acknowledged to be “brothers.” That is to say, if it be
maintained that the indicative signs of the infant child or the
primitive man are predicative, no shadow of a reason can be assigned
for withholding this designation from the indicative signs of the
lower animals. On the other hand, if this term be denied to both, its
application to the case of spoken language in its fully evolved form
must be understood to signify but a difference of phase or degree,
seeing that the one order of sign-making has been now so completely
proved to be but the genetic and improved descendant of the other.
In short, the truth obviously is that we have _a proved continuity
of development between all stages of the sign-making faculty_; and,
therefore, that any attempt to draw between one and another of them a
distinction of kind has been shown to be impossible.

The conclusions thus reached at the close of Chapter XIV. with regard
to the philology of predication were greatly strengthened by additional
facts which were immediately adduced in the next Chapter with regard
to the philology of conception. Here the object was to throw the
independent light of philology upon a point which had already been
considered as a matter of psychology, namely, the passage of receptual
denotation into conceptual denomination. This is a point which had
previously been considered only with reference to the individual: it
had now to be considered with reference to the race.

First it was shown that, owing to the young child being surrounded by
an already constructed grammar of predicative forms, the earlier phases
in the evolution of speech are greatly foreshortened in the ontogeny
of mankind, as compared with what the study of language shows them to
have been in the phylogeny. Gesture-signs are rapidly starved out when
a child of to-day first begins to speak, and so to learn the use of
grammatical forms. But early man was under the necessity of elaborating
his grammar out of his gesture-signs—and this at the same time as he
was also coining his sentence-words. Therefore, while the acquisition
of names and forms of speech by infantile man must have depended in
chief part upon gestures and grimace, this acquisition by the infantile
child is actively inimical to both.

Next we saw that the philological doctrine of “sentence-words” threw
considerable additional light on my psychological distinction between
ideas as general and generic. For a sentence-word is the expression
of an idea hitherto _generalized_, that is to say _undifferentiated_.
Such an idea, as we now know, stands at the antipodes of thought from
one which is due to what is called a _generalization_—that is to say,
a conceptual synthesis of the results of a previous analysis. And the
doctrine of sentence-words recognizes an immense historical interval
(corresponding with the immense psychological interval) between the
generic and the general orders of ideation.

Again, we saw that in all essential particulars the semiotic
construction of this the most primitive mode of articulate
communication which has been preserved in the archæology of spoken
language, bears a precise resemblance to that which occurs in the
natural language of gesture. As we saw, “gesture-language has no
grammar properly so called;” and we traced in considerable detail the
analogies—so singularly numerous and exact—between the forms of
sentences as now revealed in gesture and as they first emerged in the
early days of speech. In other words, the earliest record that speech
is able to yield as to the nature of its own origin, clearly reveals to
us this origin as emerging from the yet more primitive language of tone
and gesture. For this is the only available explanation of their close
family resemblance in the matter of syntax.

Furthermore, we have seen that in gesture language, as in the forms of
primitive speech now preserved in roots, the purposes of predication
are largely furthered by the mere apposition of denotative terms. A
generalized term of this kind (which as yet is neither noun, adjective,
nor verb), when brought into apposition with another of the same kind,
serves to convey an idea of relationship between them, or to state
something of the one by means of the other. Yet apposition of this kind
need betoken no truly conceptual thought. As we have already seen, the
laws of merely sensuous association are sufficient to insure that when
the objects, qualities, or events, which the terms severally denote,
happen to occur together in Nature, they _must_ be thus brought into
corresponding apposition by the mind: it is the logic of events which
inevitably guides such pre-conceptual utterance into a statement of
the truth that is perceived: the truth is _received into_ the mind,
not _conceived by_ it. And it is obvious how repeated statements of
truth thus delivered in receptual ideation, lead onwards to conceptual
ideation, or to statements of truth as true.

Now, if all this has been the case, it is obvious that aboriginal
words can have referred only to matters of purely receptual
significance—_i.e._ “to those physical acts and qualities which are
directly apprehensible by the senses.” Accordingly, we find in all the
earliest root-words, which the science of philology has unearthed,
unquestionable and unquestioned evidence of “fundamental metaphor,” or
of a conceptual extension of terms which were previously of no more
than receptual significance. Indeed, as Professor Whitney says, “so
pervading is it, that we never regard ourselves as having read the
history of any intellectual or moral term till we have traced it back
to its physical origin.” Without repeating all that I have so recently
said upon this matter, it will be enough once more to insist on the
general conclusions to which it led—namely, psychological analysis
has already shown us the psychological priority of the recept; and now
philological research most strikingly corroborates this analysis by
actually finding the recept in the body of every concept.

Lastly, I took a brief survey of the languages now spoken by many
widely separated races of savages, in order to show the extreme
deficiency of conceptual ideation that is thus represented. In the
result, we saw that what Archdeacon Farrar calls “the hopeless
poverty of the power of abstraction” is so surprising, that the most
ardent evolutionist could not well have desired a more significant
intermediary between the pre-conceptual intelligence of _Homo alalus_,
and the conceptual thought of _Homo sapiens_.

       *       *       *       *       *

Having thus concluded the Philology of our subject, I proceeded, in the
last chapter, to consider the probable steps of the transition from
receptual to conceptual ideation in the race.

First I dealt with a view which has been put forward on this matter
by certain German philologists, to the effect that speech originated
in wholly meaningless sounds, which in the first instance were due to
merely physiological conditions. By repeated association with the
circumstances under which they were uttered, these articulate sounds
are supposed to have acquired, as it were automatically, a semiotic
value. The answer to this hypothesis, however, evidently is, that
it ignores the whole problem which stands to be solved—namely, the
genesis of those powers of ideation which first put a soul of meaning
into the previously insignificant sounds. That is to say, it begs the
whole question which stands for solution, and, therefore, furnishes
no explanation whatsoever of the difference which has arisen between
man and brute. Nevertheless, the principles set forth in this the
largest possible extension of the so-called interjectional theory, are,
I believe, sound enough in themselves: it is only the premiss from
which in this instance they start that is untrue. This premiss is that
aboriginal man presented no rudiments of the sign-making faculty, and,
therefore, that this faculty itself required to be created _de novo_ by
accidental associations of sounds with things. But we have seen, as a
matter of fact, that this must have been very far from having been the
case; and, therefore, while recognizing such elements of truth as the
“purely physiological” hypothesis in question presents, I rejected it
as in itself not even approaching a full explanation of the origin of
speech.

Next I dealt with the hypothesis that was briefly sketched by Mr.
Darwin. Premising, as Geiger points out, that the presumably superior
sense of sight, by fastening attention upon the movements of the mouth
in vocal sign-making, must have given our simian ancestry an advantage
over other species of quadrumana in the matter of associating sounds
with receptual ideas; we next endeavoured to imagine an anthropoid ape,
social in habits, sagacious in mind, and accustomed to use its voice
extensively as an organ of sign-making, after the manner of social
quadrumana in general. Such an animal might well have distanced all
others in the matter of making signs, and even proceeded far enough to
use sounds in association with gestures, as “sentence-words”—_i.e._
as indicative of such highly generalized recepts as the presence of
danger, &c.,—even if it did not go the length of making denotative
sounds, after the manner of talking-birds. Moreover, as Mr. Darwin has
pointed out, there is a strong probability that this simian ancestor
of mankind was accustomed to use its voice in musical cadences, “as do
some of the gibbon-apes at the present day;” and this habit might have
laid the basis for that semiotic interruption of vocal sounds in which
consists the essence of articulation.

My own theory of the matter, however, is slightly different to this.
For, while accepting all that goes to constitute the substance of Mr.
Darwin’s suggestion, I think it is almost certain that the faculty of
articulate sign-making was a product of much later evolution, so that
the creature who first presented this faculty must have already been
more human than “ape-like.” This _Homo alalus_ stands before the mind’s
eye as an almost brutal object, indeed; yet still, erect in attitude,
shaping flints to serve as tools and weapons, living in tribes or
societies, and able in no small degree to communicate the logic of his
recepts by means of gesture-signs, facial expressions, and vocal tones.
From such an origin, the subsequent evolution of sign-making faculty in
the direction of articulate sounds would be an even more easy matter to
imagine than it was under the previous hypothesis. Having traced the
probable course of this evolution, as inferred by the aid of sundry
analogies; and having dwelt upon the remarkable significance in this
connection of the inarticulate sounds which still survive as so-called
“clicks” in the lowly-formed languages of Africa; I went on to detail
sundry considerations which seemed to render probable the prolonged
existence of the imaginary being in question—traced the presumable
phases of his subsequent evolution, and met the objection which might
be raised on the score of _Homo alalus_ being _Homo postulatus_.

In conclusion, however, I pointed out that whatever might be the truth
as touching the time when the faculty of articulation arose, the
course of mental evolution, after it did arise, must have been the
same. Without again repeating the sketch which I gave of what this
course must have been, it will be enough to say, in the most general
terms, that I believe it began with sentence-words in association with
gesture-signs; that these acted and reacted on one another to the
higher elaboration of both; that denotative names, for the most part
of onomatopoetic origin, rapidly underwent connotative extensions;
that from being often and necessarily used in apposition, nascent
predications arose; that these gave origin, in later times, to the
grammatical distinctions between adjectives and genitive cases on
the one hand, and predicative words on the other; that likewise
gesture-signs were largely concerned in the origin of other grammatical
forms, especially of pronominal elements, many of which afterwards went
to constitute the material out of which the forms of declension and
conjugation were developed; but that although pronouns were thus among
the earliest words which were differentiated by mankind as separate
parts of speech, it was not until late in the day that any pronouns
were used especially indicative of the first person. The significance
of this latter fact was shown to be highly important. We have already
seen that the whole distinction between man and brute resides in
the presence or absence of conceptual thought, which, in turn, is
but an expression of the presence or absence of self-consciousness.
Consequently, the whole of this treatise has been concerned with the
question whether we have here to do with a distinction of kind or of
degree—of origin or of development. In the case of the individual,
there can be no doubt that it is a distinction of degree, or
development; and I had previously shown that in this case the phase
of development in question is marked by a change of phraseology—a
discarding of objective terms for the adoption of subjective when the
speaker has occasion to speak of self. And now I showed that in the
fact here before us we have a precisely analogous proof: in exactly the
same way as psychology marks for us “the transition in the individual,”
philology marks for us “the transition in the race.”

In the foregoing _résumé_ of the present instalment of my work I have
aimed only at giving an outline sketch of the main features. And
even these main features have been so much abbreviated that it is
questionable whether more harm than good will not have been done to my
argument by so imperfect a summary of it. Nevertheless, as a general
result, I think that two things must now have been rendered apparent
to every impartial mind. First, that the opponents of evolution have
conspicuously failed to discharge their _onus probandi_, or to justify
the allegation that the human mind constitutes a great and unique
exception to the otherwise uniform law of evolution. Second, that not
only is this allegation highly improbable _a priori_, and incapable
of proof _a posteriori_, but that all the evidence that can possibly
be held to bear upon the subject makes directly on the side of its
disproof. The only semblance of an argument to be adduced in its
favour rests upon the distinction between ideation as conceptual and
non-conceptual. That such a distinction exists I freely admit; but
that it is a distinction of kind I emphatically deny. For I have shown
that the comparatively few writers who still continue to regard it as
such, found their arguments on a psychological analysis which is of a
demonstrably imperfect character; that no one of them has ever paid
any attention at all to the actual process of psychogenesis as this
occurs in a growing child; and that, with the exception of Professor
Max Müller, the same has to be said with regard to their attitude
towards the “witness of philology.” Touching the psychogenesis of a
child, I have shown that there is unquestionable demonstration of a
gradual and uninterrupted passage from the one order of ideation to the
other; that so long as the child’s intelligence is moving only in the
non-conceptual sphere, it is not distinguishable in any one feature
of psychological import from the intelligence of the higher mammalia;
that when it begins to assume the attributes of conceptual ideation,
the process depends on the development of true self-consciousness out
of the materials supplied by that form of pre-existing or receptual
self-consciousness which the infant shares with the lower animals;
that the condition to this advance in mental evolution is given by a
perceptibly progressive development of those powers of denotative and
connotative utterance which are found as far down in the psychological
scale as the talking birds; that in the growing intelligence of a child
we have thus as complete a history of “ontogeny,” in its relation to
“phylogeny,” as that upon which the embryologist is accustomed to rely
when he reads the morphological history of a species in the epitome
which is furnished by the development of an individual; and, therefore,
that those are without excuse who, elsewhere adopting the principles
of evolution, have gratuitously ignored the direct evidence of
psychological transmutation which is thus furnished by the life-history
of every individual human being.

Again, as regards the independent witness of philology, if we were
to rely on authority alone, the halting and often contradictory
opinions which from time to time have been expressed by Professor Max
Müller with reference to our subject, are greatly outweighed by those
of all his brother philologists. But, without in any way appealing
to authority further than to accept matters of fact on which all
philologists are agreed, I have purposely given Professor Max Müller an
even more representative place than any of the others, fully stated the
nature of his objections, and supplied what appears to me abundantly
sufficient answers. So far as I can understand the reasons of his
dissent from conclusions which his own admirable work has materially
helped to support, they appear to arise from the following grounds.
First, a want of clearness with regard to the principles of evolution
in general:[341] second, a failure clearly or constantly to recognize
that the roots of Aryan speech are demonstrably very far from primitive
in the sense of being aboriginal: third, a want of discrimination
between ideas as general and generic, or synthetic and unanalytical:
fourth, the gratuitous and demonstrably false assumption that in order
to name a mind must first conceive. Of these several grounds from
which his dissent appears to spring, the last is perhaps the most
important, seeing that it is the one upon which he most expressly
rears his objections. But if I have proved anything, I have proved
that there is a power of affixing verbal or other signs as marks of
merely receptual associations, and that this power is _invariably_
antecedent to the origin of conceptual utterance in the only case
where this origin admits of being directly observed—_i.e._, in the
psychogenesis of a child. Again, in the case of pre-historic man, so
far as the palæontology of speech furnishes evidence upon the subject,
this makes altogether in favour of the view that in the race, as in
the individual, denotation preceded denomination, as antecedent and
consequent. Nay, I doubt whether Max Müller himself would disagree with
Geiger where the latter tersely says, in a passage hitherto unquoted,
“Why is it that the further we trace words backwards the less meaning
do they present? I know not of any other answer to be given than that
the further they go back the less conceptuality do they betoken.”[342]
Nor can he refuse to admit, with the same authority, that “conceptual
thought (_Begriff_) allows itself to be traced backwards into an ever
narrowing circle, and inevitably tends to a point where there is no
longer either thought or speech.”[343] But if these things cannot be
denied by Max Müller himself, I am at a loss to understand why he
should part company with other philologists with regard to the origin
of conceptual terms. With them he asserts that there can be no concepts
without words (spoken or otherwise), and with them he maintains that
when the meanings of words are traced back as far as philology can
trace them, they obviously tend to the vanishing point of which Geiger
speaks. Yet, merely on the ground that this vanishing point can never
be actually reached by the investigations of philology—_i.e._, that
words cannot record the history of their own birth,—he stands out for
an interruption of the principle of continuity at the place where words
originate. A position so unsatisfactory I can only explain by supposing
that he has unconsciously fallen into the fallacy of concluding that
because all A is B, therefore all B is A. Finding that there can be no
concepts without names, he concludes that there can be no names without
concepts.[344] And on the basis of such a conclusion he naturally finds
it impossible to explain how either names or concepts could have had
priority in time: both, it seems, must have been of contemporaneous
origin; and, if this were so, it is manifestly impossible to account
for the natural genesis of either. But the whole of this trouble is
imaginary. Once discard the plainly illogical inference that because
names are necessary to concepts, therefore concepts are necessary
to names, and the difficulty is at an end. Now, I have proved, _ad
nauseam_, that there are names and names: names denotative, and names
denominative; names receptual, as well as names conceptual. Even if
we had not had the case of the growing child actually to prove the
process—a case which he, in common with all my other opponents, in
this connexion ignores,—on general grounds alone, and especially from
our observations on the lower animals, we might have been practically
certain that the faculty of sign-making _must_ have preceded that of
_thinking the signs_. And whether these pre-conceptual signs were
made by gesture, grimace, intonation, articulation, or all combined,
clearly no difference would arise so far as any question of their
influence on psychogenesis is concerned. As a matter of fact, we
happen to know that the semiotic artifice of articulating vocal tones
for purposes of denotation, dates back so far as to bring us within
philologically measurable distance of the origin of denomination, or
conceptual thought—although we have seen good reason to conclude that
before that time tone, gesture, and grimace must have been much more
extensively employed in sign-making by aboriginal man than they now
are by any of the lower animals. So that, upon the whole, unless it
can be shown that my distinction between denotation and denomination
is untenable—unless, for instance, it can be shown that an infant
requires to think of names as such before it can learn to utter
them,—then I submit that no shadow of a difficulty lies against the
theory of evolution in the domain of philology. While, on the other
hand, all the special facts as well as all the general principles
hitherto revealed by this science make entirely for the conclusion,
that pre-conceptual denotation laid the psychological conditions which
were necessary for the subsequent growth of conceptual denomination;
and, therefore, yet once again to quote the high authority of Geiger,
“Speech created Reason; before its advent mankind was reasonless.”[345]

And if this is true of philology, assuredly it is no less true of
psychology. For “the development of speech is only a copy of that chain
of processes, which began with the dawn of [human] consciousness, and
eventually ends in the construction of the most abstract idea.”[346]
Unless, therefore, it can be shown that my distinction between ideation
as receptual and conceptual is invalid, I know not how my opponents
are to meet the results of the foregoing analysis. Yet, if this
distinction should be denied, not only would they require to construct
the science of psychology anew; they would place themselves in the
curious position of repudiating the very distinction on which their
whole argument is founded. For I have everywhere been careful to place
it beyond question that what I have called receptual ideation, in all
its degrees, is identical with that which is recognized by my opponents
as non-conceptual; and as carefully have I everywhere shown that with
them I fully recognize the psychological difference between this order
of ideation and that which is conceptual. The only point in dispute,
therefore, is as to the possibility of a natural transition from the
one to the other. It is for them to show the impossibility. This they
have hitherto most conspicuously failed to do. On the other hand, I
now claim to have established the possibility beyond the reach of a
reasonable question. For I claim to have shown that the _probability_
of such a transition having previously occurred in the race, as it
now occurs in every individual, is a probability that has been raised
tower-like by the accumulated knowledge of the nineteenth century.
Or, to vary the metaphor, this probability has been as a torrent,
gaining in strength and volume as it is successively fed by facts and
principles poured into it by the advance of many sciences.

Of course it is always easy to withhold assent from a probability,
however strong: “My belief,” it may be said, “is not to be wooed; it
shall only be compelled.” Indeed, a man may even pride himself on the
severity of his requirements in this respect; and in popular writings
we often find it taken for granted that any scientific doctrine is
then only entitled to be regarded as scientific when it has been
demonstrated. But in science, as in other things, belief ought to be
proportionate to evidence; and although for this very reason we should
ever strive for the attainment of better evidence, scientific caution
of such a kind must not be confused with a merely ignorant demand
for impossible evidence. Actually to demonstrate the transition from
non-conceptual to conceptual ideation in the race, as it is every day
demonstrated in the individual, would plainly require the impossible
condition that conceptual thought should have observed its own origin.
To demand any demonstrative proof of the transition in the race would
therefore be antecedently absurd. But if, as Bishop Butler says,
“probability is the very guide of life,” assuredly no less is it the
very guide of science; and here, I submit, we are in the presence of
a probability so irresistible that to withhold from it the embrace of
conviction would be no longer indicative of scientific caution, but of
scientific incapacity. For if, as I am assuming, we already accept the
theory of evolution as applicable throughout the length and breadth of
the realm organic, it appears to me that we have positively _better_
reasons for accepting it as applicable to the length and breadth of
the realm mental. In other words, looking to all that has now been
said, I cannot help feeling that there is actually better evidence of
a psychological transition from the brute to the man, than there is of
a morphological transition from one organic form to another, in any
of the still numerous instances where the intermediate links do not
happen to have been preserved. Thus, for example, in my opinion an
evolutionist of to-day who seeks to constitute the human mind a great
exception to the otherwise uniform principle of genetic continuity, has
an even more hopeless case than he would have were he to argue that a
similar exception ought to be made with regard to the structure of the
worm-like creature Balanoglossus.

If this comparison should appear to betray any extravagant estimate
on my part of the cogency of the evidence which has thus far been
presented, I will now in conclusion ask it to be remembered that
my case is not yet concluded. For hitherto I have almost entirely
abstained from considering the mental condition of _savages_. The
reason why this important branch of my subject has not been touched
is because I reserve it for the next instalment of my work. But when
we leave the groundwork of psychological principles on which up to
this point we have been engaged, and advance to the wider field of
anthropological research in general, we shall find much additional
evidence of a more concrete kind, which almost uniformly tends to
substantiate the conclusions already gained. The corroboration thus
afforded is indeed, to my thinking, superfluous; and, therefore,
will not be adduced in this connection. Nevertheless, while tracing
the principles of mental evolution from the lowest levels which are
actually occupied by existing man, we shall find that no small light
is incidentally thrown upon the demonstrably still more primitive
intelligence of pre-historic man. Thus shall we find that we are
led back by continuous stages to a state of still human ideation,
which brings us into contact almost painfully close with that of the
higher apes. This, indeed, is a side of the general question which my
opponents are prone to ignore—just as they ignore the parallel side
which has to do with the psychogenesis of a child. And, of course,
when they thus ignore both the child and the savage, so as directly to
contrast the adult psychology of civilized man with that of the lower
animals, it is easy to show an enormous difference. But where the
question is as to whether this is a difference of degree or of kind,
the absurdity of disregarding the intermediate phases which present
themselves to actual observation is surely too obvious for comment.
At all events I think it may be safely promised, that when we come to
consider the case of savages, and through them the case of pre-historic
man, we shall find that, in the great interval which lies between such
grades of mental evolution and our own, we are brought far on the way
towards bridging the psychological distance which separates the gorilla
from the gentleman.




                                INDEX.


  A

  Abstraction. _See_ Ideas

  Addison, Mrs. K., on sign-making by a jackdaw, 97

  Adjectives, appropriately used by parrots, 129, 130, 152;
    early use of, by children, 219;
    not differentiated in early forms of speech, 295 _et seq._;
    origin of Aryan, 306;
    and in language generally, 385-86.

  Adverbs not differentiated in early forms of speech, 306

  African Bushmen. _See_ Hottentots

  African languages. _See_ Languages

  Agglomerative. _See_ Languages

  Agglutinating. _See_ Languages

  American languages. _See_ Languages

  Analytic. _See_ Languages

  Anatomy, evidence of man’s descent supplied by, 19

  Animals. _See_ Brutes

  Animism of primitive man, 275

  Ants, intelligence of, 52, 53;
    sign-making by, 91-95

  Apes, brain-weight of, 16;
    bodily structure of, 19;
    counting by, 58, 215;
    understanding of words by, 125, 126;
    unable to imitate articulate sounds, 153-157;
    psychological characters of anthropoid, in relation to the descent
        of man, 364-370;
    singing, 370, 373-378;
    other vocal sounds made by, 374;
    erect attitude assumed by, 381, 382

  Appleyard on language of savages, 349

  Apposition. _See_ Predication

  Aristotle, on intelligence of brutes, 12,
    and of man, 20;
    his classification of the animal kingdom, 79;
    his logic based on grammar of the Greek language, 314, 320

  Articulation, chap. vii.;
    classification of different kinds of, 121;
    meaningless, 121, 122;
    understanding of, 122-129;
    by dogs, 128;
    use of, with intelligent signification by talking birds, 129-139;
    arbitrary use of, by young children, 138-144;
    relation of, to tone and gesture, 145-162;
    importance of sense of sight to development of, 366, 367;
    probable period and mode of genesis of in the race, 370-373

  Aryan languages. _See_ Languages

  Aryan race, civilization of, 272;
    antiquity of, 273

  Audouin on a monkey recognizing pictorial representations, 188

  Axe, discovery of, by neolithic man, 214


  B

  Barter only used by man, 19

  Basque language. _See_ Language

  Bateman, Dr. F., on speech-centre of brain, 134, 135

  Bates, on intelligence of ants, 92, 93;
    on a monkey recognizing pictorial representations, 188.

  Bats the only mammals capable of flight, 156

  Bear, intelligence of, 51;
    understanding tones of human voice, 124

  Beattie, Dr., on intelligence of a dog, 100

  Bees, sign-making by, 90

  Bell, Professor A. Graham, on teaching a dog to articulate, 128;
    on the ideation of deaf-mutes, 150

  Belt on intelligence of ants, 52, 92

  Benfry on roots of Sanskrit, 267

  Berkeley on ideas, 21, 22

  Binet on analogies between perception and reason, 32
    and sensation, 37, 46

  Bingley on bees understanding tones of human voice, 124

  Bleek, on origin of pronouns, 302;
    on the sentence-words of African Bushmen, 316, 337, 338;
    on onomatopœia, 339;
    on the clicks of Hottentots and African Bushmen, 373

  Bonaparte, Prince Lucien, on possible number of articulate sounds, 373

  Bopp on the origin of speech, 240

  Bowen, Professor F., on psychology of judgment, 167

  Boyd Dawkins, Professor, on discovery of axe by neolithic man, 214

  Bramston, Miss, on intelligence of a dog, 56

  Brazil, climate and native languages of, 262, 263

  Brown, Thomas, on generalization, 44

  Browning, A. H., on intelligence of a dog, 99, 100

  Brutes, mind of, compared with human, 6-39;
    emotions of, 7;
    instincts of, 8;
    volition of, 8;
    intellect of, 9;
    Mr. Mivart on psychology of, 10, 177;
    as machines, 11;
    rationality of, 11, 12;
    soul of, 12;
    Bishop Butler on immortality of, 12;
    instances of intelligence of, 51-63;
    ideas of causality in, 58-60;
    appreciation of principles by, 60, 61;
    sign-making by, 88-102;
    understanding of words by, 123-127;
    articulation by, 128-138, 152;
    reasons why none have become intellectual rivals of man, 154-157;
    self-consciousness in relation to, 175-178;
    recognizing pictorial representations, 188, 189;
    conditions to genesis of self-consciousness manifested by, 195-199;
    counting by, 56-58, 214, 215;
    psychology of, in relation to the descent of man, 364-384

  Buffon, on intelligence of brutes, 12, 117;
    his parrot, 201

  Bunsen, on onomatopœia, 282;
    on Egyptian language, 297, 298;
    on the substantive verb, 309

  Burton on sign-making by Indians, 105

  Bushmen, clicks in the language of, 291

  Butler, Bishop, on immortality of brutes, 12


  C

  California, climate and native languages of, 261, 262

  Caldwell on language of savages, 349

  Carlyle on fundamental metaphor, 344

  Carpenter, Commander Alfred, on monkeys using stones to open oysters,
      382

  Casalis on poverty of savage languages in abstract terms, 351

  Cat, intelligence of, 59, 98, 99;
    use of signs by, 158

  Caterpillars, sign-making by, 95, 96

  Causation, ideas of, in brutes, 58-60;
    origin of idea of, in man, 210

  Cebus, intelligence of, 60, 61;
    different tones uttered by, 96

  Champollion on Egyptian hieroglyphics, 311

  Charlevoix on language of savages, 349

  Cheyenne language. _See_ Languages

  Child, psychogenesis of, 4, 5;
    emotions and instincts of, 7, 8;
    intelligence of, as regards classification, 26, 27, 41, 66, 67;
    instinctive and imitative articulation by, 121, 122;
    understanding of words by infantile, 123;
    spontaneous invention of words by, 138-143;
    indicative stage of language in, 158, 218-222, 324;
    denotation and connotation of, 179, 191, 218-231, 283-285;
    recognizing portraits, &c., 188, 189;
    rise of self-consciousness in, 200-212;
    use of personal pronoun by, 201, 232, 408, 409;
    hypothesis of languages having been originated by, 259-263;
    undifferentiated language of, 296, 297, 317;
    stages of language in, 157-193, 328;
    differences between infantile and primitive man, as regards
        development of speech, 329-334;
    order of development of articulate sounds in, 372, 373

  Cicero on the origin of speech, 240

  Chimpanzee. _See_ Apes

  Chinese language. _See_ Language

  Classification, in relation to abstraction, 31, 32;
    powers of, exhibited by a young child, 26, 66, 67;
    by lower animals generally, 27-30 (_see_ also under Precepts);
    of ideas, 34-39, 193;
    conceptual, 78-80, 174;
    of the animal kingdom by the early Jews and by Aristotle, 78, 79;
    of language, 85-89;
    of mental faculties artificial, 234;
    of languages, 245-251

  Clicks of Hottentots, 291

  Clothes only worn by man, 19

  Communication. _See_ Language

  Complex ideas. _See_ Ideas

  Compound ideas. _See_ Ideas

  Comte, Auguste, on the logic of feelings and of signs, 42, 46, 47

  Conception. _See_ Concepts

  Concepts, defined, 34;
    logic of, 47, and chap. iv.;
    as named recepts, 74, 75;
    as higher and lower, 76, 185;
    in relation to particular and generic ideas, 76-78;
    in relation to judgment and self-consciousness, 168-191;
    Max Müller’s alleged, 221;
    in relation to non-conceptual faculties, 234-237;
    attainment of, by the individual, 230-232;
    original, 269-281;
    philological proof of derivation of, from recepts, 343-349

  Concrete ideas. _See_ Ideas

  Connotation, 88, 89, 136, 137, 157, 159-162, 169, 170, 179-184, 218,
      219, 283, 284, 294 _et seq._, 368, 383, 384

  Conscience. _See_ Morality

  Coptic language. _See_ Language

  Copula, the, 172, 173, 230, 309, 314, 387

  Counting, by rooks, 56, 57, 214, 215;
    by an ape, 58, 215;
    by sensuous computation and by separate notation, 57, 215;
    by savages, 215

  Crawford on Malay language, 351

  Cronise on the climate of California, 261

  Crows, intelligence of, 56, 57

  Cuvier on speech as the most distinctive characteristic of man, 371


  D

  Dammaras, counting by, 215

  Darwin, Charles, on intelligence of savage man in relation to his
        cerebral development, 16, 17;
    on intelligence of animals, 51, 52, 54;
    on pointing of sporting dogs, 97;
    on expression of emotions, 103;
    on psychogenesis of child, 123, 158;
    on self-consciousness, 199;
    on descent of man, 369, 370, 374-376, 380

  Dayak language. _See_ Language

  Deaf-mutes, sign-making by, 105-120;
    ideation of, 149, 150, 339-341;
    invention of articulate signs by, 122, 263, 367

  De Fravière on sign-making by bees, 90

  Demonstrative elements. _See_ Pronouns

  Denomination, 88, 89, 161, 162, 168-170, 294, _et seq._

  Denotation, 88, 89, 157, 158, 159, 162, 168, 179-184, 218, 219, 294
      _et seq._, 368-369, 383, 384, 386

  De Quatrefages, on distinctions between animal and human intelligence,
       17-19;
    on intelligence of a dog, 198;
    on poverty of savage languages in abstract terms, 351

  Dog, seeking water in hollows, 51;
    making allowance for driftway, 52;
    generic ideas shown by, 54, 352;
    chasing imaginary pigs, 56;
    idea of causation shown by, 59, 60;
    pointing and backing of, 97, 98;
    other gesture signs made by, 99, 100, 221;
    understanding of written signs by, 101, 102;
    understanding of words by, 124, 125;
    alleged articulation by, 128;
    Indian sign for barking, 146;
    recognizing pictorial representations, 188;
    practising concealment and hypocrisy, 198;
    ejective ideation of, 198;
    receptual self-consciousness of, 199;
    counting by, 215;
    begging before a bitch, 221;
    deaf-mute’s articulate name of, 367

  Donaldson on demonstrative elements, 244

  _Dublin Review_ on psychology of judgment, 166, 167

  Dumas, Alex., on sign-making, 111

  Du Ponceau on language of savages, 349, 351


  E

  Ecitons. _See_ Ants

  Egyptian language. _See_ Language

  Elephant, intelligence of, 98

  Ellis on early English pronunciation, 373

  Emerson on fundamental metaphor, 344

  Emotions of man and brutes compared, 7

  Empty words, 246

  _Encyclopædia Britannica_ (1857), on the origin of speech, 240

  English language. _See_ Language

  Etruscan language. _See_ Language


  F

  Farrar, Archdeacon, on demonstrative elements, 244;
    on invention of languages by children, 263;
    on roots of language, 268, 358;
    on origin of the verb, 275;
    on paucity of words in vocabulary of English labourers, 280;
    on onomatopœia, 284-288, 290;
    on objective phraseology of young children and early man, 301;
    on the substantive verb, 309;
    on fundamental metaphor, 344;
    on language of savages in respect of abstraction, 350;
    on absence of subjective personal pronouns in early forms of speech,
        421

  Feejee language. _See_ Language

  Fire only made by man, 19

  Fitzgerald, P. F., on self-consciousness, 212

  Flight, capability of, in insects, reptiles, birds, and mammals, 156,
      157

  Forbes, James, on intelligence of monkeys, 100

  Fox, intelligence of, 55, 56

  Frogs, understanding by, of tones of human voice, 124


  G

  Galton, Francis, on ideas as generic images, 23;
    on relation of thought to speech, 83;
    on intelligence of Dammaras, 215

  Garnett, on nature and analysis of the verb, 275, 307, 309-312;
    on sentence-words, 300;
    on primitive forms of predication, 318;
    on fundamental metaphor, 344, 358;
    on absence of subjective cases of pronouns in early forms of speech,
        421

  Geiger, on ideas, 45;
    on dependence of thought upon language, 83;
    on understanding of words by brutes, 127;
    on roots of language, 268, 273, 336;
    on distinction between ideas as general and generic, 279;
    on increasing conceptuality of terms with increase of culture, 280;
    on the impossibility of language having ever consisted exclusively
        of general terms, 282;
    on Heyse’s theory of the origin of speech, 289;
    on onomatopœia, 292;
    on the vanishing point of language, 314, 354;
    on fundamental metaphor as illustrated by names of tools, 345, 346,
      and words of moral significance, 346, 347;
    on the sense of sight in relation to the origin of speech, 366, 367;
    on _Homo alalus_, 380

  General ideas. _See_ Ideas

  Generalization. _See_ Ideas

  Generic ideas. _See_ Recepts

  Genitive case, philology of, 305, 385

  Geoffroy Saint-Hilaire, Isid., on a monkey recognizing pictorial
      representations, 188

  Geology, imperfect record of, 19

  Gesture. _See_ Language

  Gibbon. _See_ Apes

  Goethe on obliteration of original meanings of words, 284

  Goodbehere, S., on sign-making by a pony, 97

  Gorilla. _See_ Apes

  Greek. _See_ Language

  Green, Professor, on self-consciousness, 212

  Grimace. _See_ Language

  Grimm, on the origin of speech, 240;
    on names for thunder, 286;
    on fundamental metaphor, 344


  H

  Haeckel, Professor, on _Homo alalus_, 370, 380;
    on sounds made by apes, 374

  Hague on sign-making by ants, 93, 94

  Hale, Dr. H., on spontaneous invention of words by children, 138-144;
    on the origin of languages, 259-263

  Hamilton, Sir William, on ideas as abstract and general, 24, 25, 79,
      80

  Harper, F., on Greek tenses, 301

  Haughton, Sir Graves, on roots of languages, 275

  Hebrew. _See_ Language

  Hegel, on absence in brutes of the idea of causality, 58;
    on self-consciousness, 212

  Heinieke on words spontaneously invented by deaf-mutes, 367

  Hen, different tones used by, as signs to chickens, &c., 96

  Herder, on the origin of speech, 240;
    on the original concretism of language, 359

  Herzen on self-consciousness, 212

  Heyse, on onomatopœia, 285, 287;
    on the origin of speech, 289;
    on fundamental metaphor, 344;
    on poverty of savage languages in abstract terms, 351

  Hobbes on the copula, 172, 173

  Hogg on a dog understanding words, 125

  Holden on the vocabularies of children, 372, 373

  _Homo._ _See_ Man

  Horace on the origin of speech, 240

  Horse, sign-making by, 97

  Hoste, Sir W., on intelligence of monkeys, 101

  Hottentots, language of, 291, 373, 374

  Houzeau, on dogs seeking water in hollows, 51;
    on tones used by the common hen as signs, 96;
    on talking birds, 129, 130;
    on danger signals of birds, 369

  Hovelacque, on demonstrative elements, 244;
    on auxiliary words, 247;
    on formulæ of language-structure, 248;
    on affinities of languages, 250, 255;
    on limitations of consonantal sounds in various languages, 373

  Huber on sign-making by insects, 88-90

  Human. _See_ Man

  Humboldt on the origin of speech, 240

  Hun, Dr. E. R., on spontaneous invention of words by young children,
      140-143

  Hungarian language. _See_ Language

  Huxley, Professor, on importance of the evolution theory in relation
        to anthropology, 2, 3;
    on animal automatism, 11;
    on the brain-weight of man as compared with that of anthropoid apes,
        16;
    on ideas, 23, 43;
    on importance of language to development of human thought, 134;
    on smallness of anatomical difference which determines or prevents
        power of articulation, 153, 370, 371;
    on psychology of judgment, 164;
    on erect attitude assumed by gibbon and gorilla, 381, 382


  I

  Icelandic language. _See_ Language

  Ideas, definition and classification of, 20-39;
    as recepts, chap. iii.;
    as concepts, chap. iv.;
    as general and generic, 38, 39, 68, 69, 276-281, 336, 337;
    as abstract, 20-39, 70-80;
    of causation in brutes, 58-60,
      and in man, 210;
    of uneducated deaf-mutes, 149-151;
    psychological classification of artificial, 234-237;
    of savages, 337, 338, 349-353

  Idiots, psychology of, 104, 105;
    meaningless and imitative articulation by, 121;
    ideation of, 152

  Incorporating. _See_ Languages

  Indians, sign-making by, 105-113;
    languages of, 249, 255, 259, 260

  Indicative phase of language. _See_ Language

  Indicative signs, or stage of language. _See_ Language

  Indo-European languages. _See_ Languages

  Infant. _See_ Child

  Inflectional. _See_ Languages

  Instinct, defined, 7;
    of man and brutes compared, 7, 8

  Intellect of man and brutes compared, 9

  Introspection. _See_ Self-consciousness

  Isolating. _See_ Languages


  J

  Jackdaw, sign-making by, 97

  James on language of savages, 349

  Javanese language. _See_ Language

  Johnson, Capt., on intelligence of monkeys, 100, 101

  Jones, Sir W., on the origin of speech, 240

  Judgment, unconscious or intuitive, 48, 49, 189;
    J. S. Mill upon, 48;
    psychology of, 163-237;
    G. H. Lewes upon, 164;
    Professor Huxley upon, 164;
    St. G. Mivart upon, 165, 166;
    Professor Max Müller upon, 165;
    in relation to recepts, concepts, and thought, 163-193;
    Professor Sayce upon, 170;
    pre-conceptual, 227-230, 278, 384, 386;
    blank form of, 166, 167, 319, 320


  K

  Khetshua language. _See_ Language

  Kleinpaul on gesture language, 120


  L

  Landois on sign-making by bees, 90

  Langley, S. P., on intelligence of a spider, 62, 63

  Language, in relation to brain-weight, 16;
    abstraction dependent on, 25, 30-39;
    not always necessary to thought, 81-83;
    etymology and different signification of the word, 85;
    categories of, 85-89;
    as sign-making exhibited by brutes, 88-102;
    of tone and gesture, 104-120;
    articulate, spontaneously imitated by children, 138-143;
    of tone and gesture in relation to words, 145-162;
    stages of, as indicative, denotative, connotative, denominative, and
        predicative, 157-193;
    in relation to self-consciousness, 212;
    growth of, in child, 218-237;
    theories concerning origin of, in race, 238-242, 361-384;
    evolution of, 240-245, 264, 265;
    roots of, 241-245, 248, 249;
    differentiation of, into parts of speech, 294-320, 339-342;
    demonstrative elements of, 243-245;
    of savages deficient in abstract terms, 349-353;
    nursery, 365, 366;
    Chinese, 246, 253, 256, 257, 265, 266, 298, 300, 317, 338, 373;
    Magyar, 253;
    Turkish, 253;
    Basque, 258, 260, 311;
    Etruscan, 258;
    Hungarian, 259;
    Malay, 259, 301, 305, 311, 351;
    Latin, 267;
    Egyptian, 297, 298, 310, 311;
    English, 247, 259, 266, 338, 348, 373;
    Khetshua, 263;
    Hebrew, 266, 309;
    Greek, 301, 310, 320;
    Taic, 305;
    Sanskrit, 266-277, 301, 309, 354;
    Zend, 309;
    Lithuanic, 309;
    Icelandic, 309;
    Coptic, 310;
    Javanese, 311;
    Malagassy, 311;
    Philippine, 311;
    Syriac, 311;
    Dayak, 317;
    Feejee, 318;
    Cheyenne, 348;
    Australian, 351;
    Eskimo, 351;
    Zulu, 351;
    Tasmanian, 352;
    Kurd, 352;
    Japanese, 373;
    Hottentot, 373, 374

  Languages, number of, 245;
    classification of, 245-251;
    isolating, radical, or monosyllabic, 245, 246, 267, 268;
    agglutinative or agglomerative, 247;
    inflective or transpositive, 247, 248;
    polysynthetic or incapsulating, 249;
    incorporating, 245-250;
    analytic, 250;
    affinities of, 250-259;
    native American, 249, 255, 259-263, 265, 311, 342, 348, 349, 351;
    African, 260, 263, 291, 337, 338, 351, 373, 374;
    Aryan and Indo-European, 266-278, 298, 304, 309, 314, 423;
    Semitic, 266, 311;
    Romance, 308;
    Polynesian, 318

  Latham, Dr., on the growth of language, 241;
    on language of savages in respect of abstraction, 351, 352

  Latin, roots of, 267.
    _See_ also Language

  Laura Bridgman, her syntax, 116;
    her instinctive articulate sounds, 122

  Lazarus, on ideas, 44, 45;
    on origin of speech, 361

  Lee, Mrs., on talking birds, 130

  Lefroy, Sir John, on intelligence of a dog, 99

  Leibnitz on teaching a dog to articulate, 128

  Leroy on intelligence of wolf, 53;
    of stag, 54, 55;
    of fox, 55, 56;
    of rooks, 56, 57

  Lewes, G. H., on the logic of feelings and of signs, 47;
    on judgment, 164;
    on pre-perception, 185

  Links between ape and man missing, 19

  Lithuanic language. _See_ Language

  Locke on ideas, 20-23, 28-30, 65, 342

  Logic, of recepts, chap. iii.;
    of concepts, 47, and chap. iv.

  Long on gesture-language, 120

  Lubbock, Sir John, on communication by ants, 94, 95;
    on teaching a dog written signs, 101, 102

  Lucretius on the origin of speech, 240

  Ludwig on demonstrative elements, 244


  M

  Magyar language. _See_ Language

  Malagassy language. _See_ Language

  Malay language. _See_ Language

  Malle, Dureau de la, on intelligence of brutes, 12

  Mallery, Lieut.-Col., on sign-making by Indians and deaf-mutes, &c.,
        105-112, 117-120;
    on teaching a dog to articulate, 128;
    on sign for a barking dog, 146;
    on genetic relation between gestures and words, 342, 348, 349

  Man, antecedent remarks on psychology of, 4-6;
    points of resemblance between his psychology and that of brutes,
        6-10;
    points of difference, 10-39;
    intelligence of savage, 13, 16, 17, 215, 337, 338, 349-353,
      and of palæolithic and neolithic, 14, 213, 214;
    corporeal structure of, 19;
    animism of savage and primitive, 275;
    speechless, 277;
    differences between infantile, and infantile child as regards
        development of speech, 329-334;
    use of personal pronoun by early, 300, 301, 387-389;
    hypotheses as to mode of origin of, from brute, 361-389;
    superior use by, of the sense of sight, 366, 367;
    possibly speechless condition of early, 370-379

  Mansel, Dean, on ideas as general and abstract, 42

  Maudsley, Dr., on self-consciousness, 212

  Maury on poverty of savage languages in abstract terms, 351

  M’Cook, Rev. Dr., on sign-making by ants, 95

  Metaphor, importance of, in evolution of speech, 343-349

  Meunier, on the understanding of words by brutes, 125;
    on talking birds, 130

  Midas, a, recognizing pictorial representations, 188

  Mill, James, on the copula, 173

  Mill, John Stuart, on ideas as abstract and concrete, 25;
    on the logic of feelings and of signs, 41, 42;
    on judgment, 48;
    on connotation and denomination, 169;
    on conception, 172;
    on the copula, 173;
    on predication, 236

  Milligan on poverty of savage languages in abstract terms, 352

  Mind, undergoes evolution, 4-6;
    of man and brute compared, 7-39;
    classification of faculties of artificial, 234

  Missing links, 19

  Mivart, St. George, on psychology of brutes, 10, 177;
    on animal automatism, 11;
    on superiority of savage mind to simian, 16;
    on absence in brutes of the idea of causality, 58;
    on relation of thought to speech, 83;
    on categories of language, 85, 86;
    on rationality of brutes, 87;
    on psychology of judgment, 165-167;
    on thought and reflection, 177, 178

  Mixed ideas. _See_ Ideas

  Moffat, R., on invention of languages by children, 263

  Monboddo on the origin of speech, 240

  Monkeys, general intelligence of, 60, 61, 100, 101;
    discovering mechanical principles, 60, 61, 213, 214;
    more intelligent and imitative than parrots, 153;
    recognizing pictorial representations, 188;
    understanding words, 369;
    using stones to open oysters, 382

  Monosyllabic. _See_ Languages

  Morality, alleged to distinguish man from brute, 17-19, 346;
    terms relating to, derived from ideas morally indifferent, 346, 347

  Morshead, E. J., on comparative psychology, 37

  Moschkan, Dr. A., on talking birds, 130

  Müller, F., on sign-making by bees, 90

  Müller, J., on absence in brutes of the idea of causality, 58

  Müller, Professor Friedrich, on ideas, 45;
    on language, as not identical with thought, 83;
    on classification of languages, 245;
    on sentence-words, 296;
    on undifferentiated language of child, 297;
    on origin of pronouns, 302;
    on the genitive case, 305;
    on the origin of speech, 362

  Müller, Professor F. Max, on ideas, 42, 43;
    on language as necessary to thought, 81, 83;
    on psychology of judgment, 165;
    on the copula, 173;
    on origin of the personal pronoun, 210;
    on evolution of language, 241;
    on demonstrative elements, 244, 423;
    on roots of Sanskrit, 267-289;
    on undifferentiated language of young children, 296, 317;
    on sentence-words, 298-300, 317;
    on gesture origin of pronouns, 302,
      and of language in general, 354;
    on origin of adjectives, 306;
    on the origin of verbs, 307;
    on Chinese sentence-words, 317;
    on Aristotle’s logic as based on Greek grammar, 320, 321;
    on philology proving that human thought has proceeded from the
        abstract to the concrete, 334-336;
    on names necessarily implying concepts, 336, 337;
    on fundamental metaphor, 344, 345;
    on imperfection of early names, 356;
    on the evolution of parts of speech, 423;
    on the general theory of evolution, 432, 433


  N

  Names, in relation to abstract and generic ideas, 31, 32, 57, 58,
        70-78, 174, 273-281, 336-339;
    not always necessary for thoughts, 81-83;
    or thoughts for them, 226, 336-339

  Natterer, J., on the languages of Brazil, 263

  Negro, intelligence of, 13;
    Mr. Mivart’s use of the term to illustrate the psychology of
        predication, 166, 235

  Neuter insects, instincts of, 297-299

  Nodier, on onomatopœia, 288;
    on metaphor, 344

  Noiré, on ideas, 43;
    on the origin of speech, 288, 289, 379-381;
    on the origin of pronouns, 302;
    on fundamental metaphor, 344, 345

  Nominalism, 145

  Noun-substantives, appropriately used by parrots, 129, 152;
    early use of, by children, 218;
    of earlier linguistic growth than verbs or pronouns, 275;
    not differentiated in early forms of speech, 295 _et seq._;
    oblique cases of, as attribute-words, 306, 385


  O

  Onomatopœia, in nursery-language, 136, 244;
    in relation to the origin of speech, 282-293, 339

  Orang-outang. _See_ Apes

  Oregon, climate and native languages of, 262


  P

  Palæontology. _See_ Geology

  Parrots, talking of, 128-138;
    use of indicative signs by, 158;
    denotative and connotative powers of, 179-191, 222-226;
    statements made by, 189, 190

  Particular ideas. _See_ Ideas

  Parts of speech, differentiation of language into, 294-320, 339-342,
      423

  Peckham, Mr. and Mrs., on memory in a spider, 207

  Perception, analogies between reason and, 32;
    constituted by fusions of sensations, 37;
    in relation to other mental faculties, 48;
    illusions of, 49

  Perez on psychogenesis of the child, 26, 41, 158, 210

  Philippine language. _See_ Language

  Philology. _See_ Language

  Pickering on poverty of savage languages in abstract terms, 352

  Pictures recognized as portraits, &c., by infants, dogs, and monkeys,
      188, 189

  Pig taught to point game, 97

  Poescher on the Aryan race, 273

  Pointing, game by a pig, 97;
    of setter-dogs, 97, 98;
    as the first stage of language, 157, 158

  Polynesian languages. _See_ Languages

  Polysynthetic. _See_ Languages

  Pony, sign-making by, 97

  Pott, on the origin of speech, 240;
    on language-roots, 267;
    on names for thunder, 286;
    on fundamental metaphor, 344

  Powers on the climate of California, 261

  Pre-concepts, 185-193, 218, 219, 227-230, 278, 384, 386

  Predicate, the, 305, 306, 423

  Predication, 88, 89, 157, 162-164, 169, 171, 175, 227, 235-237, 294
      _et seq._, 384, 386, 387, 422

  Prepositions not differentiated in early forms of speech, 295
      _et seq._

  Preyer, on psychogenesis of the child, 26, 219, 221, 222;
    on sensuous computation of number, 57, 58

  Primates. _See_ Apes _and_ Monkeys

  Pritchard on Celtic languages, 275

  Progress in successive generations, 12-15

  Pronoun, first personal, 201, 232, 301, 387-389, 408, 409

  Pronouns and pronominal elements, 210, 275;
    not differentiated in early forms of speech, 295 _et seq._;
    origin of, in gestures, 301-304, 387, 421, 422

  Proposition. _See_ Predication

  Psychogenesis. _See_ Child

  Psychology. _See_ Mind


  Q

  Quadrumana. _See_ Apes _and_ Monkeys


  R

  Radical. _See_ Languages

  Ray on different tones used by the common hen, 96

  Reason in relation to perception, 32;
    to sensation, 37;
    and to other mental faculties in general, 48

  Recepts, defined, 36-39;
    logic of, 40-69;
    recognized by previous writers, 40-45;
    in relation to the intellectual faculties, 48-50, 234;
    examples of, in the animal kingdom, 51-63;
    as primitive as percepts, 64-69;
    of water-fowl, 74;
    in relation to judgment and self-consciousness, 176-193;
    as higher and lower, 184-193;
    counting by, 214, 215;
    naming by, 218, 219;
    of the framers of Sanskrit, 277-279;
    philologically prior to concepts, 343-349

  Reflection in relation to reflex action, 48.
    _See also_ Thought

  Reflex action, 48

  Religion alleged to distinguish man from brute, 17, 19, 346

  Renan on roots of Hebrew, 266

  Rengger on different tones uttered by the cebus, 96

  Reptiles, understanding by, of tones of human voice, 124

  Ribot, Professor, on self-consciousness, 212

  Richter on obliteration of the original meanings of words, 284

  Romance languages. _See_ Languages

  Romanes, on teaching an ape to count, 58;
    on intelligence of cebus, 60, 61;
    on sign-making by caterpillars, 95, 96;
    on pointing of setter-dogs, 97, 98;
    on sign-making by other dogs, 100, 221;
    on infant intelligence, 122, 159, 160, 188, 189, 218-220, 232, 283,
        324;
    on dogs and apes understanding words, 124-126;
    on talking birds, 129, 130;
    on ideation of deaf-mutes, 149, 150

  Rooks, intelligence of, 56, 57

  Roots of language. _See_ Language


  S

  Sandwith on poverty of savage languages in abstract terms, 352

  Sanskrit. _See_ Language

  Sayce, Professor, on differences of degree and kind, 3;
    on terms as abbreviated judgments, 170;
    on the number of languages, 245;
    on the affinities between languages, 250-259;
    on monosyllabic origin of language, 268;
    on civilization of the Aryan race, 272;
    on antiquity of the Aryan race, 273;
    on rarity of general terms in savage languages, 280;
    on onomatopœia, 286;
    on the clicks in the language of Hottentots, etc., 291, 373, 374;
    on sentence-words, 299, 300, 303;
    on the origin of pronouns, 302;
    on the genitive case, the predicate, and the attribute, 305, 306,
        313, 423;
    on the evolution of nouns, adjectives, and verbs, 308;
    on Aristotle’s logic as based on Greek grammar, 321;
    on deficiency of savage languages in abstract terms, 352;
    on Noiré’s theory of the origin of speech, 380

  Schelling on parts of speech, 295, 296

  Schlegel on the origin of speech, 240

  Schleicher, on evolution of language, 241;
    on formulæ of language-structure, 248

  Scott, Dr., on psychology of idiots and deaf-mutes, 104, 105, 115,
      116, 121

  Scott, Sir Walter, on a dog understanding words, 125

  Self-consciousness, condition to introspective reflection or thought,
        175;
    absent in brutes, 175, 176;
    genesis of, 194-212;
    philosophy and psychology of, 194, 195;
    character of, in man and in brutes, 195-212;
    as inward and outward, or receptual and conceptual, 199, 200;
    growth of, in child, 200-212, 228, 229-234

  Semitic. _See_ Languages

  Sensation in relation to perception and reason, 37;
    and to other mental faculties in general, 48

  Sentence and sentence-words, 296 _et seq._

  Sicard, Abbé, on syntax of gesture-language, 116

  Sight, superior use of sense of, by man, 366, 367

  Signs and sign-making. _See_ Language

  Simple ideas. _See_ Ideas

  Skeat, Professor, on Aryan roots of English, 266

  Skinner, Major, on intelligence of elephants, 98

  Smith, Rev. S., on ideation of deaf-mutes, 150

  Snakes, understanding by, of tones of human voice, 124

  Solomon, quoted, 195

  Somnambulism in animals, 149

  Speech. _See_ Language

  Spider, intelligence of, 62, 63, 153, 207

  Steinthal, on ideas, 45;
    first issue of his _Zeitschrift_, 240;
    on roots of language, 277;
    on onomatopœia, 286;
    on primitive forms of predication, 318

  Stephen, Leslie, on intelligence of the dog, 54

  Stephen, Sir James, on dependence of thought upon language, 85

  Street, A. E., on vocabulary of a young child, 143, 144

  Substantive. _See_ Noun _and_ Verb

  Sullivan, Sir J., on talking birds, 130

  Sully, J., on ideas, 40, 41;
    on illusions of perception, 49;
    on rise of self-consciousness in the growing child, 201-203, 207,
        210, 212

  Sweet, on animistic thought of primitive man, 275;
    on the evolution of grammatical forms, 306, 315, 316

  Syntax, of gesture-language, 107-120;
    of different spoken languages, 246, 247;
    of gesture-language in relation to that of early speech, 339-342,
        385

  Syriac language. _See_ Language


  T

  Taine, on psychogenesis of the child, 26, 66, 67, 180, 181;
    on abstract ideas, 31, 32;
    on self-consciousness, 212

  Thought, distinguished from reason, 12;
    absent in brutes, 29, 30;
    dependent on language, 30, 31;
    simplest element of, 165, 174, 215, 216;
    animistic, of primitive and savage man, 275;
    not necessary to naming, 226, 336-339

  Toads, understanding by, of tones of human voice, 124

  Tone. _See_ Language

  Tools, said to be only used by man, 19;
    names of, derived from activities requiring only natural organs,
        345-347;
    used by monkeys, 382

  Threlkeld on language of savages, 349

  Transposition. _See_ Languages

  Tschudi, Baron von, on the Khetshua language, 262, 263

  Turkish language. _See_ Language

  Tylor, on sign-making by Indians and deaf-mutes, 105-108, 113-117;
    on articulate sounds instinctively made by deaf-mutes, 122;
    on ideation of deaf-mutes, 150


  V

  Varro on roots of Latin, 267

  Verbs, appropriately used by parrots, 130, 152;
    substantive, 167, 308-312;
    early use of, by children, 219;
    early origin of, 274;
    not differentiated in early forms of speech, 295 _et seq._;
    development of, 275, 307, 308, 385, 386

  Voice. _See_ Language

  Volition of man and brutes compared, 8


  W

  Waitz, Professor, on self-consciousness, 212;
    on the sentence as the unit of language, 296

  Wallace, A. R., on intelligence of savage man in relation to his
      cerebral development, 15, 16

  Ward on the descent of man, 365

  Wasps, sign-making by, 88-90

  Watson on understanding of words by brutes, 125

  Wedgwood, on roots of language, 268;
    on onomatopœia, 288

  Westropp, H. M., on intelligence of a bear, 51

  Whitney, Professor, on dependence of thought upon words, 83;
    on superiority of voice to gesture in sign-making, 147, 148;
    on our ignorance of polysynthetic languages, 255, 256;
    on monosyllabic origin of language, 267;
    on civilization of the Aryan race, 272;
    on the growth of language, 290;
    on priority of words to sentences, 333, 334;
    on fundamental metaphor, 343;
    on the possibly speechless condition of primitive man, 369

  Wildman on bees understanding tones of human voice, 124

  Wilkes, Dr. S., on talking birds, 131, 132, 136

  Will. _See_ Volition

  Wolf, intelligence of, 53

  Wright, Chauncey, on language in relation to brain-weight, 16;
    on self-consciousness, 199, 206, 207, 212

  Wundt, Professor, on latent period in seeing and hearing, 146;
    on self-consciousness, 197, 200, 201, 208, 211, 212;
    on evolution of language, 265;
    on the distinction between ideas as general and generic, 279, 280;
    on onomatopœia, 287, 291;
    on objective phraseology of primitive speech, 301;
    on sentence-words, 304


  Y

  Youatt on a pig being taught to point game, 97


  Z

  Zend language. _See_ Language

  Zoological affinity between man and brute, 19




                              FOOTNOTES:

[1] _Man’s Place in Nature_, p. 59.

[2] It is perhaps desirable to explain from the first that by the words
“difference of kind,” as used in the above paragraph and elsewhere
throughout this treatise, I mean difference of _origin_. This is the
only real distinction that can be drawn between the terms “difference
of kind” and “difference of degree;” and I should scarcely have deemed
it worth while to give the definition, had it not been for the confused
manner in which the terms are used by some writers—_e.g._ Professor
Sayce, who says, while speaking of the development of languages from
a common source, “differences of degree become in time differences of
kind” (_Introduction to the Science of Language_, ii. 309).

[3] See _Mental Evolution in Animals_, chapter on the Emotions.

[4] _Mental Evolution in Animals_, p. 159. “The term is a generic
one, comprising all the faculties of mind which are concerned in
conscious and adaptive action, antecedent to individual experience,
without necessary knowledge of the relation between means employed and
ends attained, but similarly performed under similar and frequently
recurring circumstances by all individuals of the same species.”

[5] Of course my opponents will not allow that this word can be
properly applied to the psychology of any brute. But I am not now
using it in a question-begging sense: I am using it only to avoid the
otherwise necessary expedient of coining a new term. Whatever view we
may take as to the relations between human and animal psychology, we
must in some way distinguish between the different ingredients of each,
and so between the instinct, the emotion, and the intelligence of an
animal. See _Mental Evolution in Animals_, p. 335, et seq.

[6] If any one should be disposed to do so, I can only reply to him
in the words of Professor Huxley, who puts the case tersely and
well:—“What is the value of the evidence which leads one to believe
that one’s fellow-man feels? The only evidence in this argument from
analogy is the similarity of his structure and of his actions to
one’s own, and if that is good enough to prove that one’s fellow-man
feels, surely it is good enough to prove that an ape feels,” etc.
(_Critiques and Addresses_, p. 282). To this statement of the case
Mr. Mivart offers, indeed, a criticism, but it is one of a singularly
feeble character. He says, “Surely it is not by similarity of structure
or actions, but by _language_ that men are placed in communication
with one another.” To this it seems sufficient to ask, in the first
place, whether language is not action; and, in the next, whether, as
expressive of _suffering_, articulate speech is regarded by us as more
“eloquent” than inarticulate cries and gestures?

[7] Of course where the term Reason is intended to signify
Introspective Thought, the above remarks do not apply, further than to
indicate the misuse of the term.

[8] I here neglect to consider the view of Bishop Butler, and others
who have followed him, that animals may have an immortal principle as
well as man; for, if this view is maintained, it serves to identify,
not to separate, human and brute psychology. The dictum of Aristotle
and Buffon, that animals differ from man in having no power of
mental apprehension, may also be disregarded; for it appears to be
sufficiently disposed of by the following remark of Dureau de la Malle,
which I here quote as presenting some historical interest in relation
to the theory of natural selection. He says: “Si les animaux n’étaient
pas suscéptibles d’apprendre les moyens de se conserver, les espèces se
seraient anéanties.”

[9] John Fiske, _Excursions of an Evolutionist_, pp. 42, 43 (1884).

[10] _Natural Selection_, p. 343. It will subsequently appear, as a
general consequence of our investigation of savage psychology, that of
these two opposite opinions the one advocated by Mr. Mivart is best
supported by facts. But I may here adduce one or two considerations of
a more special nature bearing upon this point. First, as to cerebral
_structure_, the case is thus summed up by Professor Huxley:—“The
difference in weight of brain between the highest and the lowest man
is far greater, both relatively and absolutely, than that between
the lowest man and the highest ape. The latter, as has been seen, is
represented by, say 12 ounces of cerebral substance absolutely, or by
32:20 relatively; but, as the largest recorded human brain weighed
between 65 and 66 ounces, the former difference is represented by
more than 33 ounces absolutely, or by 65:32 relatively. Regarded
systematically, the cerebral differences of man and apes are not of
more than generic value—his family distinction resting chiefly on his
dentition, his pelves, and his lower limbs” (_Man’s Place in Nature_,
p. 103). Next, concerning cerebral _function_, Mr. Chauncey Wright well
remarks:—“A psychological analysis of the faculty of language shows
that even the smallest proficiency in it might require more brain power
than the greatest proficiency in any other direction” (_North American
Review_, Oct. 1870, p. 295). After quoting this, Mr. Darwin observes of
savage man, “He has invented and is able to use various weapons, tools,
traps, &c., with which he defends himself, kills or catches prey,
and otherwise obtains food. He has made rafts or canoes for fishing,
or crossing over to neighbouring fertile islands. He has discovered
the art of making fire.... These several inventions, by which man in
the rudest state has become so preeminent, are the direct results
of the development of his powers of observation, memory, curiosity,
imagination, and reason. I cannot, therefore, understand how it is that
Mr. Wallace maintains that ‘natural selection could only have endowed
the savage with a brain a little superior to that of an ape’” (_Descent
of Man_, pp. 48, 49).

[11] _The Human Species_, English trans., p. 22.

[12] Sundry other and still more special distinctions of a
psychological kind have been alleged by various writers as obtaining
between man and the lower animals—such as making fire, employing
barter, wearing clothes, using tools, and so forth. But as all
these distinctions are merely particular instances, or detailed
illustrations, of the more intelligent order of ideation which belongs
to mankind, it is needless to occupy space with their discussion.
Here, also, I may remark that in this work I am not concerned with
the popular objection to Darwinism on account of “missing-links,” or
the absence of fossil remains structurally intermediate between those
of man and the anthropoid apes. This is a subject that belongs to
palæontology, and, therefore, its treatment would be out of place in
these pages. Nevertheless, I may here briefly remark that the supposed
difficulty is not one of any magnitude. Although to the popular
mind it seems almost self-evident that if there ever existed a long
series of generations connecting the bodily structure of man with
that of the higher apes, at least some few of their bones ought now
to be forthcoming; the geologist too well knows how little reliance
can be placed on such merely negative testimony where the record of
geology is in question. Countless other instances may now be quoted of
connecting links having been but recently found between animal groups
which are zoologically much more widely separated than are apes and
men. Indeed, so destitute of force is this popular objection held to
be by geologists, that it is not regarded by them as amounting to any
objection at all. On the other hand, the close anatomical resemblance
that subsists between man and the higher apes—every bone, muscle,
nerve, vessel, etc., in the enormously complex structure of the one
coinciding, each to each, with the no less enormously complex structure
of the other—speaks so voluminously in favour of an uninterrupted
continuity of descent, that, as before remarked, no one who is at
all entitled to speak upon the subject has ventured to dispute this
continuity so far as the corporeal structure is concerned. All the
few naturalists who still withhold their assent from the theory of
evolution in its reference to man, expressly base their opinion on
those grounds of psychology which it is the object of the present
treatise to investigate.

[13] In my previous work I devoted a chapter to “Imagination,” in
which I treated of the psychology of ideation so far as animals are
concerned. It is now needful to consider ideation with reference to
man; and, in order to do this, it is further needful to revert in
some measure to the ideation of animals. I will, however, try as far
as possible to avoid repeating myself, and therefore in the three
following chapters I will assume that the reader is already acquainted
with my previous work. Indeed, the argument running through the three
following chapters cannot be fully appreciated unless their perusal
is preceded by that of chapters ix. and x. of _Mental Evolution in
Animals_.

[14] _Human Understanding_, bk. ii., chap. ii., 10, 11. To this passage
Berkeley objected that it is impossible to form an abstract idea of
quality as apart from any concrete idea of object; _e.g._ an idea of
motion distinct from that of any body moving. (See _Principles of
Human Knowledge_, Introd. vii.-xix.). This is a point which I cannot
fully treat without going into the philosophy of the great discussion
on Nominalism, Realism, and Conceptualism—a matter which would take
me beyond the strictly psychological limits within which I desire
to confine my work. It will, therefore, be enough to point out that
Berkeley’s criticism here merely amounts to showing that Locke did not
pursue sufficiently far his philosophy of Nominalism. What Locke did
was to see, and to state, that a general or abstract idea embodies a
perception of likeness between individuals of a kind while disregarding
the differences; what he failed to do was to take the further step of
showing that such an idea is not an idea in the sense of being a mental
image; it is merely an intellectual symbol of an actually impossible
existence, namely, of quality apart from object. Intellectual symbolism
of this kind is performed mainly through the agency of verbal or other
conventional signs (as we shall see later on), and it is owing to
a clearer understanding of this process that Realism was gradually
vanquished by Nominalism. The only difference, then, between Locke
and Berkeley here is, that the nominalism of the former was not so
complete or thorough as that of the latter. I may remark that if in the
following discussion I appear to fail in distinctly setting forth the
doctrine of nominalism, I do so only in order that my investigation
may avoid needless collision with conceptualism. For myself I am a
nominalist, and agree with Mill that to say we think in concepts is
only another way of saying that we think in class names.

[15] This simile has been previously used by Mr. Galton himself, and
also by Mr. Huxley in his work on Hume.

[16] Hence, the only valid distinction that can be drawn between
abstraction and generalization is that which has been drawn by
Hamilton, as follows: “Abstraction consists in concentration of
attention upon a particular object, or particular quality of an object,
and diversion of it from everything else. The notion of the _figure_
of the desk before me is an abstract idea—an idea that makes part of
the total notion of that body, and on which I have concentrated my
attention, in order to consider it exclusively. This idea is abstract,
but it is at the same time individual: it represents the figure of this
particular desk, and not the figure of any other body.” Generalization,
on the other hand, consists in an ideal compounding of abstractions,
“when, comparing a number of objects, we seize on their resemblances;
when we concentrate our attention on these points of similarity.... The
general notion is thus one which makes us know a quality, property,
power, notion, relation, in short, any point of view under which we
recognize a plurality of objects as a unity.” Thus, there may be
abstraction without generalization; but inasmuch as abstraction has
then to do only with particulars, this phase of it is disregarded
by most writers on psychology, who therefore employ abstraction
and generalization as convertible terms. Mill says, “By _abstract_
I shall always, in Logic proper, mean the opposite of _concrete_;
by an abstract name the name of an attribute; by a concrete name,
the name of an object” (_Logic_, i. § 4). Such limitation, however,
is arbitrary—it being the same kind of mental act to “concentrate
attention upon a particular _object_,” as it is to do so upon any
“particular _quality_ of an object.” Of course in this usage Mill is
following the schoolmen, and he expressly objects to the change first
introduced (apparently) by Locke, and since generally adopted. But it
is of little consequence in which of the two senses now explained a
writer chooses to employ the word “abstract,” provided he is consistent
in his own usage.

[17] The age here mentioned closely corresponds with that which is
given by M. Perez, who says:—“At seven months he compares better
than at three; and he appears at this age to have visual perceptions
associated with ideas of _kind_: for instance, he connects the
different flavours of a piece of bread, of a cake, of fruit, with their
different forms and colours” (_First Three Years of Childhood_, English
trans., p. 31).

[18] _Die Seele des Kindes_, s. 87.

[19] Taine, _Intelligence_, p. 18.

[20] _Human Understanding_, bk. ii., ch. ii., §§ 5-7.

[21] If required, proof of this fact is to be found in abundance in the
chapter on “Imagination,” _Mental Evolution in Animals_, pp. 142-158.
It is there shown that imagination in animals is not dependent only on
associations aroused by sensuous impressions from without, but reaches
the level of carrying on a train of mental imagery _per se_.

[22] _Loc. cit._, pp. 397-399. Allusion may also be here conveniently
made to an interesting and suggestive work by another French writer,
M. Binet (_La Psychologie du Raisonnement_, 1886). His object is to
show that all processes of reasoning are fundamentally identical with
those of perception. In order to do this he gives a detailed exposition
of the general fact that processes of both kinds depend on “fusions”
of states of consciousness. In the case of perception the elements
thus fused are sensations, while in the case of reasoning they are
perceptions—in both cases the principle of association being alike
concerned.

[23] _Mental Evolution in Animals_, p. 118.

[24] In this connection I may quote the following very lucid statements
from a paper by the Secretary of the Victoria Institute, which is
directed against the general doctrine that I am endeavouring to
advance, _i.e._ that there is no distinction of kind between brute and
human psychology.

“Abstraction and generalization only become intellectual when they are
utilized by the intellect. A bull is irritated by a red colour, and not
by the object of which redness is a property; but it would be absurd to
say that the bull voluntarily abstracts the phenomenon of redness from
these objects. The process is essentially one of abstraction, and yet
at the same time it is essentially automatic.” And with reference to
the ideation of brutes in general, he continues:—“Certain qualities
of an object engage his attention to the exclusion of other qualities,
which are disregarded; and thus he abstracts automatically. The image
of an object having been imprinted on his memory, the feelings which
it excited are also imprinted on his memory, and on the reproduction
of the image these feelings and the actions resulting therefrom are
reproduced, likewise automatically: thus he acts from experience,
automatically still. The image may be the image of the same object, or
the image of another object of the same species, but the effect is the
same, and thus he generalizes, automatically also.” Lastly, speaking of
inference, he says:—“This method is common to man and brute, and, like
the faculties of abstraction, &c., it only becomes intellectual when
we choose to make it so.” (E. J. Morshead, in an essay on _Comparative
Psychology_, _Journ. Vic. Inst._, vol. v., pp. 303, 304, 1870.) In
the work of M. Binet already alluded to, the distinction in question
is also recognized. For he says that the “fusion” of sensations which
takes place in an act of perception is performed automatically (_i.e._
is receptual); while the “fusion” of perceptions which are concerned in
an act of reason is performed intentionally (_i.e._ is conceptual).

[25] The more elaborate analysis of German psychologists has yielded
five orders instead of three; namely, _Wahrnehmung_, _Anschauung_,
_Vorstellungen_, _Erfahrungsbegriff_, and _Verstandesbegriff_. But for
the purposes of this treatise it is needless to go into these finer
distinctions.

[26] _Outlines of Psychology_, p. 342. The italics are mine. It will
be observed that Mr. Sully here uses the term “generic” in exactly the
sense which I propose.

[27] _First Three Years of Childhood_, English trans., pp. 180-182.

[28] _Examination of Hamilton’s Philosophy_, p. 403.

[29] To this, Max Müller objects on account of its veiled
conceptualism—seeing that it represents the “notion” as
chronologically prior to the “name” (_Science of Thought_, p. 268).
With this criticism, however, I am not concerned. Whether “the many
pictures” which the mind thus forms, and blends together into what
Locke terms a “compound idea,” deserve, when so blended, to be called
“a general notion” or a “concept”—this is a question of terminology
of which I steer clear, by assigning to such compound ideas the term
recepts, and reserving the term notions, or concepts, for compound
ideas _after they have been named_.

[30] _Logos_, p. 175, quoted by Max Müller, who adds:—“The followers
of Hume might possibly look upon the faded images of our memory as
abstract ideas. Our memory, or, what is often equally important, our
oblivescence, seems to them able to do what abstraction, as Berkeley
shows, never can do; and under its silent sway many an idea, or cluster
of ideas, might seem to melt away till nothing is left but a mere
shadow. These shadows, however, though they may become very vague,
remain percepts; they are not concepts” (_Science of Thought_, p.
453). Now, I say it is equally evident that these shadows are _not_
percepts: they are the result of the _fusion_ of percepts, no one of
which corresponds to their generic sum. Seeing, then, that they are
neither percepts nor concepts, and yet such highly important elements
in ideation, I coin for them the distinctive name of recepts.

[31] _Life of Hume_, p. 96.

[32] Steinthal and Lazarus, however, in dealing with the problem
touching the origin of speech, present in an adumbrated fashion this
doctrine of receptual ideation with special reference to animals. For
instance, Lazarus says, “Es gibt in der gewöhnlichen Erfahrung kein
so einfaches Ding von einfacher Beschaffenheit, dass wir es durch
_eine_ Sinnesempfindung wahrnehmen könnten; erst aus der Sammlung
seiner Eigenschaften, d. h. erst aus der _Verbindung_ der mehreren
Empfindungen ergibt sich _die Wahrnehmung eines Dinges_: erst indem
wir die weisse Farbe sehen, die Härte fühlen und den süssen Geschmack
empfinden, erkennen wir ein Stück Zucker” (_Das Leben der Seele_
(1857), 8, ii. 66). This and other passages in the same work follow
the teaching of Steinthal; _e.g._ “Die Anschauung von einem Dinge
ist der Complex der sämmtlichen Empfindungserkenntnisse, die wir
von einem Dinge haben ... die Anschauung ist eine Synthesis, aber
eine unmittelbare, die durch die Einheit der Seele gegeben ist.”
And, following both these writers, Friedrich Müller says, “Diese
Sammlung und Einigung der verschiedenen Empfindungen gemäss der in
den Dingen verbundenen Eigenschaften heisst Anschauung” (_Grundriss
der Sprachwissenschaft_, i. 26). On the other hand, their brother
philologist, Geiger, strongly objects to this use of the term
_Anschauung_, under which, he says, “wird theils etwas von der
Sinneswahrnehmung gar nicht Unterschiedenes verstanden, theils auch
ein dunkles Etwas, welches, ohne dass die Bedingungen und Ursachen zu
erkennen sind, die Einheit der Wahrnehmungen zu kleineren und grössern
Complexen bewirken soll.... So dass ich eine solche ‘Synthesis’ nicht
auch bei dem Thiere ganz ebenso wie bei dem Menschen voraussetze: ich
glaube im Gegentheile, dass es sich mit der Sprache erst entwickelt”
(_Ursprung der Sprache_, 177, 178). Now, I have quoted these various
passages because they serve to render, in a brief and instructive
form, the different views which may be taken on a comparatively simple
matter owing to the want of well-defined terms. No doubt the use of
the term _Anschauung_ by the above writers is unfortunate; but by it
they appear to me clearly to indicate a nascent idea of what I mean by
a recept. They all three fail to bring out this idea in its fulness,
inasmuch as they restrict the powers of non-conceptual “synthesis” to
a grouping of simple perceptions furnished by different sense-organs,
instead of extending it to a synthesis of syntheses of perceptions,
whether furnished by the same or also by different senses. But these
three philologists are all on the right psychological track, and their
critic Geiger is quite wrong in saying that there can be no synthesis
of (non-conceptual) ideas without the aid of speech. As a matter of
fact the _dunkles Etwas_ which he complains of his predecessors as
importing into the ideation of animals, is an _Etwas_ which, when
brought out into clearer light, is fraught with the highest importance.
For, as we shall subsequently see, it is nothing less than the
needful psychological condition to the subsequent development both of
speech and thought. The term _Apperception_ as used by some German
psychologists is also inclusive of what I mean by receptual ideation.
But as it is also inclusive of conceptual, nothing would here be gained
by its adoption. Indeed F. Müller expressly restricts its meaning
to conceptual ideation, for he says, “Alle psychischen Processe bis
einschliesslich zur Perception lassen sich ohne Sprache ausführen und
vollkommen begreifen, die Apperception dagegen lässt sich nur an der
Hand der Sprache denken” (_loc. cit._ i., 29).

[33] As stated in a previous foot-note, this truth is well exhibited by
M. Binet, _loc. cit._

[34] The word Logic is derived from λόγος, which in turn is derived
from λέδω, to arrange, to lay in order, to pick up, to bind together.

[35] The terms Logic of Feelings and Logic of Signs were first
introduced and extensively employed by Comte. Afterwards they were
adopted, and still more extensively employed by Lewes, who, however,
seems to have thought that he so employed them in some different sense.
To me it appears that in this Lewes was mistaken. Save that Comte is
here, as elsewhere, intoxicated with theology, I think that the ideas
he intended to set forth under these terms are the same as those which
are advocated by Lewes—although his incoherency justifies the remark
of his follower:—“Being unable to understand this, I do not criticize
it” (_Probs. of Life and Mind_, iii., p. 239). The terms in question
are also sanctioned by Mill, as shown by the above quotation (p. 42).

[36] _Mental Evolution in Animals_, p. 62.

[37] Special attention, however, may be drawn to the fact that the
term “unconscious judgment” is not metaphorical, but serves to convey
in a technical sense what appears to be the precise psychology of
the process. For the distinguishing element of a judgment, in its
technical sense, is that it involves an element of _belief_. Now,
as Mill remarks, “when a stone lies before me, I am conscious of
certain sensations which I receive from it; but if I say that these
sensations come to me from an external object which I perceive, the
meaning of these words is, that receiving the sensations, I intuitively
believe that an external cause of those sensations exists” (_Logic_,
i., p. 58). In cases, such as that mentioned in the text, where the
“unconscious judgment” is wrong—_i.e._ the perception illusory—it
may, of course, be over-ridden by judgment of a higher order, and thus
we do not end by believing that the bowl is a sphere. Nevertheless,
so far as it is dependent on the testimony of our senses, the mind
judges erroneously in perceiving the bowl as a sphere. In his work on
_Illusions_, Mr. Sully has shown that illusions of perception arise
through the mental “application of a rule, valid for the majority of
cases, to an exceptional case.” In other words, an erroneous judgment
is made by the non-conceptual faculties of perception—this judgment
being formed upon the analogies supplied by past experience. Of course,
such an act of merely perceptual inference is not a judgment, strictly
so called; but it is clearly _allied_ to judgment, and convenience
is consulted by following established custom in designating it
“unconscious,” “intuitive,” or “perceptual judgment.”

[38] _Descent of Man_, p. 76.

[39] See _Animal Intelligence_, pp. 465, 466.

[40] Of course the words “general idea” and “concept” here are open to
that psychological objection for the avoidance of which I have coined
the terms generic idea and recept.

[41] In my previous works I have already quoted facts of animal
intelligence narrated by this author, but not any of those which I am
now about to use.

[42] _Intelligence of Animals_, English trans., p. 20.

[43] _Ibid._, p. 107. This identical illustration appears to have
occurred independently both to Mr. Darwin and Mr. Leslie Stephen. All
these writers use the terms “abstract” and “general” as above; but,
of course, as shown in my last chapter, this is merely a matter of
terminology—in my opinion, however, objectionable, because appearing
to assume, without analysis, that the ideation of brutes and of men is
identical in kind.

[44] _Ibid._, pp. 43, 44.

[45] _Ibid._, p. 39.

[46] _Ibid._, p. 30. In the present connection, also, I may refer
to the chapter on Imagination in my previous work, where sundry
illustrations are given of this faculty as it occurs in animals; for
wherever imagination leads to appropriate action, there is evidence
of a Logic of Recepts, which in the higher levels of imagination,
characteristic of man, passes into a Logic of Concepts.

Since publishing the chapter just alluded to, I have received an
additional and curious illustration of the imaginative faculty in
animals, which I think deserves to be published for its own sake. Of
course we may see in a general way that dogs and cats resemble children
in their play of “pretending” that inanimate objects are alive, and
this betokens a comparatively high level of the imaginative faculty.
The case which I am about to quote, however, appears to show that this
kind of imaginative play may extend in animals, as in children, to the
still higher level of not only pretending that inanimate objects are
alive, but of “peopling space with fancy’s airy forms.” I shall quote
the facts in the words of my correspondent, who is Miss Bramston, the
authoress.

“_Watch_ is a collie dog belonging to the Archbishop of Canterbury; but
lives with me a good deal, as Lambeth does not suit him. He is a very
remarkable dog in many ways, which I will not inflict on you. He is
very intelligent, understands many words, and can perform tricks. What
I mention him for, however, is that he is the only dog I ever met with
a dramatic faculty. His favourite drama is chasing imaginary pigs. He
used now and then to be sent to chase real pigs out of the field, and
after a time it became a custom for Miss Benson to open the door for
him after dinner in the evening, and say, ‘Pigs!’ when he always ran
about, wildly chasing imaginary pigs. If no one opened the door, he
went to it himself wagging his tail, asking for his customary drama. He
now reaches a further stage, for as soon as we get up after our last
meal he begins to bark violently, and if the door is open he rushes
out to chase imaginary pigs with no one saying the word ‘pigs’ at all.
He usually used to be sent out to chase pigs after prayers in the
evening, and when he came to my small house it was amusing to see that
he recognized the function of prayers, performed with totally different
accompaniments, to be the same as prayers performed in an episcopal
chapel, so far as he expected ‘Pigs’ to be the end of both. The word
‘Pigs,’ uttered in any tone, will always set him off playing the same
drama.”

[47] _Ibid._, pp. 125, 126.

[48] Professor Preyer has ascertained experimentally the number of
objects (such as shot-corns, pins, or dots on a piece of paper), which
admit of being simultaneously estimated with accuracy. (_Sitzungs
berichten der Gesellschaft für Medicin und Naturwissenshaft_, 29 Juli,
1881.) The number admits of being largely increased by practice, until,
with an exposure to view of one second’s duration, the estimate admits
of being correctly made up to between twenty and thirty objects. (See
also _Mental Evolution in Animals_, p. 138.)

[49] _Lessons from Nature_, pp. 219, 220.

[50] See _Animal Intelligence_, pp. 422-424.

[51] I may here observe that the earliest age in the infant at which
I have observed such appreciation of causality to occur is during the
sixth month. With my own children at that age I noticed that if I
made a knocking sound with my concealed foot, they would look round
and round the room with an obvious desire to ascertain the cause
that was producing the sound. Compare, also, _Mental Evolution in
Animals_, pp. 156-158, on emotions aroused in brutes by sense of the
_mysterious_—_i.e._ the _unexplained_.

[52] The reader is referred to the whole biography of this monkey
(_Animal Intelligence_, pp. 484-498) for a number of other facts
serving to show to how high a level of intelligent grouping—or of
“logic”—recepts may attain without the aid of concepts. In the same
connection I may refer to the chapter on “Imagination” in _Mental
Evolution in Animals_, and also to the following pages in _Animal
Intelligence_:—128-40; 181-97, 219-222, 233, 311-335, 337, 338, 340,
348-352, 377-385, 397-410, 413-425, 426-436, 445-470, 478-498.

[53] Taine, _On Intelligence_, pp. 16, 17.

[54] _Lectures_, vol. ii., p. 290.

[55] _Science of Thought_, p. 35. For his whole argument, see pp. 30-64.

[56] _Ursprung der Sprache_, s. 91.

[57] _Grundriss der Sprachwissenshaft_, i., s. 16. It will be observed
that there is an obvious analogy between the process above described,
whereby conceptual ideation becomes degraded into receptual, and that
whereby, on a lower plane of mental evolution, intelligence becomes
degraded into instinct. In my former work I devoted many pages to
a consideration of this subject, and showed that the condition to
intelligent adjustments thus becoming instinctive is invariably to be
found in frequency of repetition. Instincts of this kind (“secondary
instincts”) may be termed degraded recepts, just as the recepts spoken
of in the text are degraded concepts; neither could be what it now is,
but for its higher parentage. Any one who is specially interested in
the question whether there can be thought without words, may consult
the correspondence between Prof. Max Müller, Mr. Francis Galton,
myself, and others, in _Nature_, May and June, 1887 (since published
in a separate form); between the former and Mr. Mivart, in _Nature_,
March, 1888. Also an article by Mr. Justice Stephen in the _Nineteenth
Century_, April, 1888. Prof. Whitney has some excellent remarks on this
subject in his _Language and the Study of Language_, pp. 405-411.

[58] From this it will be seen that by using such terms as “inference,”
“reason,” “rational,” &c., in alluding to mental processes of the lower
animals, I am in no way prejudicing the question as to the distinction
between man and brute. In the higher region of recepts both the man
and the brute attain in no small degree to a perception of analogies
or relations: this is inference or ratiocination in its most direct
form, and differs from the process as it takes place in the sphere
of conceptual thought only in that it is not itself an object of
knowledge. But, considered as a process of inference or ratiocination,
I do not see that it should make any difference in our terminology
whether or not it happens to be itself an object of knowledge.
Therefore I do not follow those numerous writers who restrict such
terms to the higher exhibitions of the process, or to the ratiocination
which is concerned only with introspective thought. It may be a matter
of straw-splitting, but I think it is best to draw our distinctions
where the distinctions occur; and I cannot see that it modifies the
process of inference, as inference, whether or not the mind, in virtue
of a superadded faculty, is able to think about the process as a
process—not any more, for instance, than the process of association
is altered by its becoming itself an object of knowledge. Therefore, I
hope I have made it clear that in maintaining the rationality of brutes
I am not arguing for anything more than that they have the power, as
Mr. Mivart himself allows, of drawing “practical inferences.” Hitherto,
then, my difference with Mr. Mivart—and, so far as I know, with all
other modern writers who maintain the irrationality of brutes—is only
one of terminology.

[59] See _Animal Intelligence_, p. 158.

[60] _Animal Intelligence_, pp. 114-116.

[61] Kreplin, quoted by Büchner.

[62] The best instances of sign-making among Invertebrata other than
the Hymenoptera which I have met with is one that I have myself
observed and already recorded in _Mental Evolution in Animals_ (p. 343,
note). The animal is the processional caterpillar. These larvæ migrate
in the form of a long line, crawling Indian file, with the head of the
one touching the tail of the next in the series. If one member of the
series be removed, the next member in advance immediately stops and
begins to wag its head in a peculiar manner from side to side. This
serves as a signal for the next member also to stop and wag his head,
and so on till all the members in front of the interruption are at a
standstill, all wagging their heads. But as soon as the interval is
closed up by the advance of the rear of the column, the front again
begins to move forward, when the head-wagging ceases.

[63] _Fac. Ment. des Animaux_, tom. ii., p. 348.

[64] Darwin, _Descent of Man_, pp. 84, 85.

[65] _Nature_, April 10, 1884, pp. 547, 548.

[66] For information on all these points, see Darwin, _Expression of
the Emotions_.

[67] Quoted by Tylor, _Early History of Mankind_, p. 80.

[68] Burton, _City of the Saints_, p. 151.

[69] _Loc. cit._, p. 78.

[70] _Sign-language among the North American Indians, &c._, by
Lieut.-Col. Garrick Mallery (_First Annual Report of the Bureau of
Ethnology, Washington_, 1881).

[71] Mallery, _loc. cit._, p. 320. The author gives several very
interesting records of such conversations, and adds that the mutes show
more aptitude in understanding the Indians than _vice versâ_, because
to them “the ‘action, action, action,’ of Demosthenes is their only
oratory, and not a heightening of it, however valuable.”

[72] _Loc. cit._, p. 39.

[73] See especially Tylor, _loc. cit._, pp. 28-30, where an interesting
account is given of the elaborate and yet self-speaking signs whereby
an adult deaf-mute gave directions for the drawing up of his will.

[74] _Early History of Mankind_, pp. 24-32.

[75] _Loc. cit._, p. 54.

[76] Further information of a kind corroborating what has been given
in the foregoing chapter concerning gesture-language may be found in
Long’s _Expedition to the Rocky Mountains_, and Kleinpaul’s paper in
_Völkerpsychologie_, _&c._, vi. 352-375. The subject was first dealt
with in a philosophical manner by Leibnitz, in 1717, _Collectanea
Etymologia_, ch. ix.

[77] For meaningless articulation by idiots, see Scott’s _Remarks on
Education of Idiots_. The fact is alluded to by most writers on idiot
psychology, and I have frequently observed it myself. But the case of
uneducated deaf-mutes is here more to the purpose. I will, therefore,
furnish one quotation in evidence of the above statement. “It is a very
notable fact bearing upon the problem of the Origin of Language, that
even born-mutes, who never heard a word spoken, do of their own accord
and without any teaching make vocal sounds more or less articulate,
to which they attach a definite meaning, and which, when once made,
they go on using afterwards in the same unvarying sense. Though these
sounds are often capable of being written down more or less accurately
with our ordinary alphabets, this effect on those who make them can,
of course, have nothing to do with the sense of hearing, but must
consist only in particular ways of breathing, combined with particular
positions of the vocal organs” (Tylor, _Early History of Mankind_, p.
72, where see for evidence). The instinctive articulations of Laura
Bridgman (who was blind as well as deaf) are in this connection even
still more conclusive (see _ibid._, pp. 74, 75).

[78] Writers on infant psychology differ as to the time when words are
first understood by infants. Doubtless it varies in individual cases,
and is always more or less difficult to determine with accuracy. But
all observers agree—and every mother or nurse could corroborate—that
the understanding of many words and sentences is unmistakable long
before the child itself begins to speak. Mr. Darwin’s observations
showed that in the case of his children the understanding of words and
sentences was unmistakable between the tenth and twelfth months.

[79] See _Animal Intelligence_: for Fish, p. 250; for Frogs and Toads,
p. 225; for Snakes, p. 261; for Birds and Mammals in various parts of
the chapters devoted to these animals. The case quoted on the authority
of Bingley regarding the tame bees of Mr. Wildman, which he had taught
to obey words of command (p. 189), would, if corroborated, carry the
faculty in question into the invertebrated series.

[80] Although the ages at which talking proper begins varies much in
different children, it may be taken as a universal rule—as stated in
the last foot-note—that words, and even sentences, are understood long
before they are intelligently articulated; although, as previously
remarked, even before any words are _understood_ meaningless syllables
may be spontaneously or instinctively articulated.

[81] See, for instance, Watson’s _Reasoning Power in Animals_, pp.
137-149, and Meunier’s _Les Animaux Perfectibles_, ch. xii.

[82] _Ursprung der Sprache_, p. 122.

[83] Some cases are on record of dogs having been taught to
articulate. Thus the thoughtful Leibnitz vouches for the fact (which
he communicated to the _Académie Royale_ at Paris, and which that body
said they would have doubted had it not been observed by so eminent
a man), that he had heard a peasant’s dog distinctly articulate
thirty words, which it had been taught to say by the peasant’s son.
The _Dumfries Journal_, January, 1829, mentions a dog as then living
in that town, who uttered distinctly the word “William,” which was
the name of a person to whom he was attached. Again, Colonel Mallery
writes:—“Some recent experiments of Prof. A. Graham Bell, no less
eminent from his work in artificial speech than in telephones, shows
that animals are more physically capable of pronouncing articulate
sounds than has been supposed. He informed the writer that he recently
succeeded by manipulation in causing an English terrier to form a
number of the sounds of our letters, and particularly brought out
from it the words ‘How are you, grandmama,’ with distinctness.” As I
believe that the barrier to articulation in dogs is anatomical and not
psychological, I regard it as merely a question of observation whether
this barrier may not in some cases be partly overcome; but, as far as
the evidence goes, I think it is safer to conclude that the instances
mentioned consisted in the animals so modulating the tones of their
voices as to resemble the sounds of certain words.

[84] Mr. Darwin writes:—“It is certain that some parrots, which have
been taught to speak, connect unerringly words with things, and persons
with events. I have received several detailed accounts to this effect.
Admiral Sir J. Sullivan, whom I know to be a careful observer, assures
me that an African parrot, long kept in his father’s house, invariably
called certain persons of the household, as well as visitors, by their
names. He said ‘Good morning’ to every one at breakfast, and ‘Good
night’ to each as they left the room at night, and never reversed
these salutations. To Sir J. Sullivan’s father he used to add to the
‘good morning’ a short sentence, which was never repeated after his
father’s death. He scolded violently a strange dog which came into the
room through an open window, and he scolded another parrot (saying,
‘You naughty polly!’), which had got out of its cage, and was eating
apples on the kitchen table. Dr. A. Moschkan informs me that he knew a
starling which never made a mistake in saying in German ‘good morning’
to persons arriving, and ‘good-bye, old fellow’ to those departing. I
could add several other cases” (_Descent of Man_, p. 85). Similarly
Houzeau gives some instances of nearly the same kind (_Fac. Ment. des
Anim._, tom, ii., p. 309, _et seq._); and Mrs. Lee, in her _Anecdotes_
records several still more remarkable cases (which are quoted by
Houzeau), as does also M. Meunier in his recently published work on
_Les Animaux Perfectibles_. In my own correspondence I have received
numerous letters detailing similar facts, and from these I gather
that parrots often use comical phrases when they desire to excite
laughter, pitiable phrases when they desire to excite compassion, and
so on; although it does not follow from this that the birds understand
the meanings of these phrases, further than that they are as a whole
appropriate to excite the feelings which it is desired to excite. I
have myself kept selected parrots, and can fully corroborate all the
above statements from my own observations.

[85] _Journal of Mental Science_, July, 1879.

[86] This term has been previously used by some philologists to signify
ejaculation by man. It will be observed that I use it in a more
extended sense.

[87] _Man’s Place in Nature_, p. 52. I may here appropriately allude
to a paper which elicited a good deal of discussion some years ago. It
was read before the Victoria Institute in March, 1872, by Dr. Frederick
Bateman, under the title “Darwinism tested by Recent Researches in
Language;” and its object was to argue that the faculty of articulate
speech constitutes a difference of kind between the psychology of man
and that of the lower animals. This argument Dr. Bateman sought to
establish, first on the usual grounds that no animals are capable of
using words with any degree of understanding, and, second, on grounds
of a purely anatomical kind. In the text I fully deal with the first
allegation: as a matter of fact, many of the lower animals understand
the meanings of many words, while those of them which are alone capable
of imitating our articulate sounds not unfrequently display a correct
appreciation of their use as signs. But what I have here especially
to consider is the anatomical branch of Dr. Bateman’s argument. He
says:—“As the remarkable similarity between the brain of man and that
of the ape cannot be disputed, if the seat of human speech could be
positively traced to any particular part of the brain, the Darwinian
could say that, although the ape could not speak, he possessed the germ
of that faculty, and that in subsequent generations, by the process of
evolution, the ‘speech centre’ would become more developed, and the
ape would then speak.... If the scalpel of the anatomist has failed to
discover a _material locus habitandi_ for man’s proud prerogative—the
faculty of Articulate Language; if science has failed to trace speech
to a ‘material centre,’ has failed thus to connect matter with mind, I
submit that speech is the barrier between men and animals, establishing
between them a difference not only of degree but of kind; the Darwinian
analogy between the brain of man and that of his reputed ancestor,
the ape, loses all its force, whilst the common belief in the Mosaic
account of the origin of man is strengthened.” Now, I will not wait to
present the evidence which has fully satisfied all living physiologists
that “the faculty of Articulate Language” has “a _material locus
habitandi_;” for the point on which I desire to insist is that it
cannot make one iota of difference to “the Darwinian analogy” whether
this faculty is restricted to a particular “speech-centre,” or has
its anatomical “seat” distributed over any wider area of the cerebral
cortex. Such a “seat” there must be in either case, if it be allowed
(as Dr. Bateman allows) that the cerebral cortex “is undoubtedly the
instrument by which this attribute becomes externally manifested.”
The question whether “the material organ of speech” is large or small
cannot possibly affect the question on which we are engaged. Since Dr.
Bateman wrote, a new era has arisen in the localization of cerebral
functions; so that, if there were any soundness in his argument, one
would now be in a position immensely to strengthen “the Darwinian
analogy;” seeing that physiologists now habitually utilize the brains
of monkeys for the purpose of analogically localizing the “motor
centres” in the brain of man. In other words, “the Darwinian analogy”
has been found to extend in physiological, as well as in anatomical
detail, throughout the entire area of the cortex. But, as I have shown,
there is no soundness in his argument; and therefore I do not avail
myself of these recent and most wonderfully suggestive results of
physiological research.

[88] I may, however, add the following corroborative observations, as
they have not been previously published. I owe them to the kindness
of my friend Mr. A. E. Street, who kept a diary of his children’s
psychogenesis. When about two years of age one of these children
possessed the following vocabulary:—

 Af-ta (in imitation of the sound which the nurse used to make when
 pretending to drink) = _drinking_ or a _drink_, _drinking-vessel_, and
 hence a _glass_ of any kind.

 Vy = a _fly_.

 Vy-’ta = _window_, _i.e._ the ‘ta or af-ta (_glass_) on which a fly
 walks.

 Blow = _candle_.

 Blow-hattie = a _lamp_, _i.e._ candle with a hat or shade.

 ’Nell = a _flower_, _i.e._ smell.

These words are clearly all of imitative origin. The following,
however, seem to have been purely arbitrary:—

 Numby = _food_ of any kind (onomatopoetic).

 Nunny = _dress_ of any kind.

 Milly = _dressing_, and any article used in dressing, _e.g._ a pin.

 Lee = _the name for her nurse_, though no one else called the woman by
 any other name than nurse.

 Diddle-iddle = _a hole_; hence _a thimble_; hence _a finger_.

 Wasky = _the sea_.

 Bilu-bilu = _the printed character_ “&,” invented on learning the
 first letters of her alphabet, and always afterwards used.

[89] Touching the comparative rapidity with which signs admit of being
made to the eye and ear respectively, it may be pointed out that there
is a physiological reason why the latter should have the advantage; for
while the ear can distinguish successive sensations separated only by
an interval of .016 sec., the eye cannot do so unless the interval is
more than .047 sec. (Wundt).

[90] _Encyclop. Brit._, 9th ed., art. _Philology_.

[91] It will be remembered that in a previous chapter I argued the
impossibility of estimating the reflex influence of speech upon
gesture, in the case of the high development attained by the latter in
man. In the text I am now considering the converse influence of gesture
upon speech, and find that it is no more easy precisely to estimate.
There can be no doubt, however, that the reciprocal influence must
have been great in both directions, and that it must have proceeded
from gesture to speech in the first instance, and afterwards, when the
latter had become well developed as a system of auditory signs, from
speech to gesture. More will require to be said upon this point in a
future chapter.

[92] “The remark made by Tiedemann on the imperative intention of
tears, is confirmed by similar observations of Charles Darwin’s. At
the age of eleven weeks, in the case of one of his children, a little
sooner in another, the nature of their crying changed according to
whether it was produced by hunger or suffering. And this means of
communication appeared to be very early placed at the service of the
will. The child seemed to have learnt to cry when he wished, and to
contract his features according to the occasion, so as to make known
that he wanted something. This development of the will takes place
towards the end of the third month.” (Perez, _First Three Years of
Childhood_, English trans., p. 101.)

[93] Several writers of repute have habitually used the word “Judgment”
in a most unwarrantable manner—Lewes, for instance, making it stand
indifferently for an act of sensuous determination and an act of
conceptual thought. I may, therefore, here remark that in the following
analysis I shall not be concerned with any such gratuitous abuses of
the term, but will understand it in the technical sense which it bears
in logic and psychology. The extraordinary views which Mr. Huxley
has published upon this subject I can only take to be ironical. For
instance, he says:—“Ratiocination is resolvable into predication,
and predication consists in marking in some way the existence, the
co-existence, the succession, the likeness and unlikeness, of things or
their ideas. Whatever does this, reasons; and I see no more ground for
denying to it reasoning power, because it is unconscious, than I see
for refusing Mr. Babbage’s engine the title of a calculating machine
on the same grounds” (_Critiques and Addresses_, p. 281). If this
statement were taken seriously, of course the answer would be that Mr.
Babbage’s engine is called a calculating machine only in a metaphorical
sense, seeing that it does not evolve its results by any process at
all resembling, or in any way analogous to, those of a human mind. It
would be an absurd misstatement to say that a machine either reasons
or predicates, _only_ because it “marks in some way the existence,
the co-existence, the succession, and the likeness and unlikeness of
things.” A rising barometer or a striking clock do not predicate, any
more than a piece of wood, shrieking beneath a circular saw, feels.
To denominate purely mechanical or unconscious action—even though it
should take place in a living agent and be perfectly adjustive—reason
or predication, would be to confuse physical phenomena with psychical;
and, as I have shown in my previous work, even if it be supposed that
the latter are mere “indices” or “shadows” of the former, _still the
fact of their existence must be recognized_; and the processes in
question have reference to them, not to their physical counterparts.
It is, therefore, just as incorrect to say that a calculating machine
really calculates, or predicates the result of its calculations, as it
would be to say that a musical-box composes a tune because it plays a
tune, or that the love of Romeo and Juliet was an isosceles triangle,
because their feelings of affection, each to each, were, like the
angles at the base of that figure, equal. But, as I have said, I take
it that Professor Huxley must here have been writing in some ironical
sense, and therefore purposely threw his criticisms into a preposterous
form.

[94] The “images answering respectively to ‘a thing being,’ and ‘a
thing not being,’ and to ‘at the same time’ and ‘in the same sense,’”
must indeed be “vague.” How is it conceivable that “the imagination”
can entertain any such “images” at all, apart from the “abstract ideas”
of the “mind”? Such ideas as “a thing not being,” or “being in the same
sense,” &c., belong to the sphere of conceptual thought, and cannot
have any existence at all except as “abstract ideas of the mind.”

[95] _Nature_, August 21, 1879.

[96] The statement conveyed in this sentence I am not able to
understand, and therefore will not hereafter endeavour to criticize.
If it be taken literally—and I know not in what other sense to take
it—we must suppose the writer to mean that “greenness” only occurs in
“grass,” or, which is the same thing, that only grass is green.

[97] _Lessons from Nature_, pp. 226, 227.

[98] For instance, Professor Francis Bowen, of Harvard College, in an
essay on _The Human and Brute Mind_, _Princeton Review_, 1880.

[99] Mill, following the schoolmen, uses the terms connotation and
denomination as synonymous. For the distinction which I have drawn
between them see above, p. 162.

[100] Sayce, _Introduction to the Science of Language_, i., 115.

[101] This view of a concept as already embodying the idea of existence
is not really opposed to that of Mill, where he points out that if we
pronounce the word “Sun” alone we are not necessarily affirming so much
as existence of the sun (_Logic_, i., p. 20); for, although we are not
affirming existence of that particular body, we must at least have the
idea of its existence _as a possibility_: the use of the term carries
with it the implied idea of such a possibility, and therefore the idea
of existence—whether actual or potential—as already present to the
mind of the speaker.

[102] In order to avoid misapprehension, I may observe that the
criticism which Mill passes upon this analysis of the proposition by
Hobbes (_Logic_, i., p. 100) has no reference to the only matter with
which I am at present concerned—namely, the function of the copula.
Indeed, with regard to this matter I am in full agreement with both
the Mills. For James Mill, see _Analysis of the Human Mind_, i. 126,
_et seq._; Mr. John Stuart Mill writes as follows:—“It is important
that there should be no indistinctness in our conception of the nature
and office of the copula; for confused notions respecting it are among
the causes which have spread mysticism over the field of logic, and
perverted its speculations into logomachies. It is apt to be supposed
that the copula is something more than a mere sign of predication;
that it also signifies existence. In the proposition, Socrates is
just, it may seem to be implied not only that the quality _just_ can
be affirmed of Socrates, but moreover that Socrates _is_, that is to
say exists. This, however, only shows that there is an ambiguity in
the word _is_; a word which not only performs the function of a copula
in affirmations, but has also a meaning of its own, in virtue of which
it may itself be made the predicate of a proposition” (_Logic_, i., p.
86). In my chapters on Philology I shall have to recur to the analysis
of predication, and then it will be seen how completely the above view
has been corroborated by the progress of linguistic research.

[103] Of course concepts may be something more than mere recepts known
as such: they may be the knowledge of other concepts. But with this
higher stage of conceptual ideation I am not here concerned.

[104] _Nature_, August 21, 1879.

[105] Taine, _Intelligence_, pp. 399, 400.

[106] Or, as we may now more closely define it, a denominated recept.
A merely denotated recept (such as a parrot’s name for its recept of
dog) is not conceptual, even in the lowest degree. In other words,
named recepts, merely as such, are not necessarily concepts. Whether
or not they are concepts depends on whether the naming has been an
act of denotation or of denomination—conscious only, or likewise
_self_-conscious.

[107] I coin this word on the pattern already furnished by
“pre-perception,” which was first introduced by Lewes, and is now in
general use among psychologists.

[108] Touching the power of recognizing pictorial representations among
animals, this unquestionably occurs in dogs (see _Animal Intelligence_,
pp. 455, 456), and there is some evidence to show that it is likewise
displayed by monkeys. For Isidore Geoffroy St. Hilaire relates of a
species of Midas (_Corinus_) that it distinguished between different
objects depicted on an engraving; and Audouin “showed it the portraits
of a cat and a wasp, at which it became much terrified: whereas, at the
sight of a figure of a grasshopper or a beetle, it precipitated itself
on the picture, as if to seize the objects there represented” (Bates,
_Nat. on Amaz._, p. 60). The age at which a young child first learns to
recognize pictorial resemblances no doubt varies in individual cases. I
have not met with any evidence on this subject in the writings of other
observers of infant psychology. The earliest age at which I observed
any display of this faculty in my own children was at eight months,
when my son stared long and fixedly at my own portrait in a manner
which left no doubt on my mind that he recognized it as resembling the
face of a man. Moreover, always after that day when asked in that room,
“Where’s papa?” he used at once to look up and point at the portrait.
Another child of my own, which had not seen this portrait till she was
sixteen months old, immediately recognized it at first sight, as was
proved by her pointing to it and calling it “Papa.” Two months later I
observed that she also recognized pictorial resemblances of animals,
and for many months afterwards her chief amusement consisted in looking
through picture-books for the purpose of pointing out the animals or
persons depicted—calling “Ba-a-a” to the sheep, “Moo” to the cows,
grunting for the pigs, &c., these sundry sounds having been taught
her as names by the nurse. She never made a mistake in this kind of
nomenclature, and spontaneously called all pictorial representations
of men “Papa,” of women “Mama,” and of children “Ilda”—the latter
being the name which she had given to her younger brother. Moreover,
if a picture-book were given into her hands upside-down, she would
immediately perceive and rectify the mistake; and whenever she happened
to see a pictorial representation of an animal—as, for instance, on
a screen or wall-paper—she would touch it and utter the sound that
was her name for that animal. With a third child, who was still wholly
speechless at eighteen months, I tried the experiment of spreading out
a number of photographic portraits, and asking him “Which is mamma?
Which is papa?” &c. Without any hesitation he indicated them all
correctly.

[109] By using the word “judgment” in all these cases I am in no
way prejudicing the argument of my opponents. The explanation which
immediately follows in the text is sufficient to show that the
qualifying terms “receptual” and “pre-conceptual” effectually guard
against any abuse of the term—quite as much, for instance, as when
psychologists speak of “perceptual judgments,” or “unconscious
judgments,” or “intuitive judgments,” in connection with still lower
levels of mental operation. And it seems to me better thus to qualify
an existing term than to add to the already large number of words I
have found it necessary to coin.

[110] I may here remark that this possibility of receptual predication
on the part of talking birds is not entirely hypothetical: I have
some evidence that it may be actually realized. For instance, a
correspondent writes of a cockatoo which had been ill:—“A friend
came the same afternoon, and asked him how he was. With his head on
one side and one of his cunning looks, he told her that he was ‘a
little better;’ and when she asked him if he had not been very ill, he
said, ‘Cockie better; Cockie ever so much better.’ ... ‘When I came
back (after a prolonged absence) he said, ‘Mother come back to little
Cockie: Mother come back to little Cockie. Come and love me and give
me pretty kiss. Nobody pity poor Cockie. The boy beat poor Cockie.’ He
always told me if Jes scolded or beat him. He always told me as soon as
he saw me, and in such a pitiful tone.... The remarkable thing about
this bird is that he does not merely ‘talk’ like parrots in general,
but so habitually _talks to the purpose_.”

[111] Lest there should still be any ambiguity about the numerous terms
which I have found it necessary to coin, I will here supply a table of
definitions.

 Lower recept = an automatic grouping of percepts.

 Higher recept = pre-concept; or a degree of receptual ideation which
 does not occur in any brute.

 Lower concept = named recept, provided that the naming be due to
 reflective thought.

 Higher concept = a named compound of concepts.

The analogues of these terms are, in the matter of naming:—

 Receptual naming = denotation, which includes pre-conceptual naming.

 Conceptual naming = denomination.

And, in the matter of judging, the analogues are:—

 Receptual judgment = automatic, “practical,” or unthinking inference.

 Pre-conceptual judgment = the higher, though still unthinking,
 inferences of a child prior to the rise of self-consciousness.

 Conceptual judgment = true judgment, whether exhibited in
 denomination, predication, or any act of inference for which
 self-conscious thought may be required.

[112] See above, Chapters II. and IV.

[113] See _Mental Evolution in Animals_, chapter on “Imagination.”

[114] In the opinion of Wundt, the most important of all conditions
to the genesis of self-consciousness is given by the muscular sense
in acts of voluntary movement (_Vorlesungen über die Menschen und
Thierseele_, 18 vol.). While agreeing with him that this is a highly
important condition, I think the others above mentioned are quite as
much, or even more so.

[115] See for cases of this, _Animal Intelligence_, pp. 410, 443, 444,
450-452, 458, 494.

[116] The following is a good example of ejective ideation in a
brute—all the better, perhaps, on account of being so familiar. I
quote it from Quatrefage’s _Human Species_, pp. 20, 21:—“I must here
beg permission to relate the remembrance of my struggles with a mastiff
of pure breed and which had attained its full size, remaining, however,
very young in character. We were very good friends and often played
together. As soon as ever I assumed an attitude of defence before him,
he would leap upon me with every appearance of fury, seizing in his
mouth the arm which I had used as a shield. He might have marked my arm
deeply at the first onset, but he never pressed it in a manner that
could inflict the slightest pain. I often seized his lower jaw with my
hand, but he never used his teeth so as to bite me. And yet the next
moment the same teeth would indent a piece of wood I tried to tear away
from them. This animal evidently knew what it was doing when it feigned
the passion precisely opposite to that which it really felt; when, even
in the excitement of play, it retained sufficient mastery over its
movements to avoid hurting me. In reality it played a part in a comedy,
and we cannot act without being conscious of it.”

[117] Not, however, wholly so. Mr. Chauncey Wright has clearly
recognized the existence of what I term receptual self-consciousness,
and assigned to it the name above adopted—_i.e._ “outward
self-consciousness.” See his _Evolution of Self-consciousness_. Mr.
Darwin, also, appears to have recognized this distinction, in the
following passage:—“It may be freely admitted that no animal is
self-conscious, if by this term is implied that he reflects on such
points as whence he comes or whither he will go, or what is life and
death, and so forth. But how can we feel sure that an old dog with
an excellent memory and some power of imagination, as shown by his
dreams, never reflects on his past pleasures or pains in the chase?
And this would be a form of self-consciousness” (_Descent of Man_, p.
83). Of course a psychologist may take technical exception to the word
“reflects” in this passage; but that this kind of receptual reflection
does take place in dogs appears to me to be definitely proved by the
facts of home-sickness and pining for absent friends, above alluded to.

[118] In the present connection the following very pregnant sentence
may be appropriately quoted from Wundt:—“Wenn wir überall auf die
Empfindung als Ausgangspunkt der ganzen Entwicklungsreihe hingewiesen
werden, so _müssen_ auch die Anfänge jener Unterscheidung des Ichs von
den Gegenständen schon in den Empfindungen gelegen sein” (_Vorlesungen
über die Menschen und Thierseele_, i. 287). And to the objection that
there can be no thought without knowledge of thought, he replies
that before there is any knowledge of thought there must be the same
order of thinking as there is of perceiving prior to the advent of
self-consciousness—_e.g._ receptual ideas about space before there is
any conceptual knowledge of these ideas as such.

[119] Sully, _loc. cit._, p. 376. See also Wundt, _loc. cit._, i. 289.
He shows that this speaking of self in the third person is not due to
“imitation,” but, on the contrary, opposed to it. For “a thousand times
the child hears that its elders do not thus speak of themselves.” The
child hears that its elders call it in the third person, and in this it
follows them. But such imitation as we here find is expressive only of
the fact that hitherto the child has not distinguished between self as
an object and self as a subject. Only later on, when this distinction
has begun to dawn, does imitation proceed to apply to the self the
first person, after the manner in which other selves (now recognized by
the child as such) are heard to do.

[120] _Loc. cit._, p. 377.

[121] _Loc. cit._, pp. 435, 436.

[122] _Philosophical Discussions_, p. 256. See also _Animal
Intelligence_, pp. 269, 270, for the case of a parrot apparently
endeavouring to recover the memory of a particular word in a phrase. In
the course of an interesting research on the intelligence of spiders
(_Journ. Morphol._, i., p. 383-419), Mr. and Mrs. Peckham have recently
found that the memory of eggs which have been withdrawn from the mother
is retained by her for a period varying in different species from less
than one to more than two days.

[123] Sully, _loc. cit._, p. 377.

[124] Wundt, _loc. cit._, ii. 289, 290. He gives cases where such a
definite memory of the moment has persisted, and elsewhere states that
such is the case in his own experience. The circumstance which here
was connected with the sudden birth of self-consciousness consisted in
rolling down stairs into a cellar—an event which no doubt was well
calculated forcibly to impress upon infant consciousness that it was
itself, and nobody else.

[125] See _Mental Evolution in Animals_, pp. 161-165. Perez records
analogous facts with regard to the infant as unmistakably displayed in
the fourteenth week (_First Three Years of Childhood_, English trans.,
p. 29).

[126] _Outlines of Psychology_, p. 378.

[127] _Vorlesungen_, _&c._, i. 289.

[128] In the above sketch of the principles which are concerned in the
development of self-consciousness, I have only been concerned with the
matter on the side of its psychology, and even on this side only so far
as my own purposes are in view. Those who wish for further information
on the psychology of the subject may consult Wundt, _loc. cit._; Sully,
_loc. cit._, and _Illusions_, ch. x.; Taine, _On Intelligence_, pt.
ii., bk. iii.; Chauncey Wright, _Evolution of Self-consciousness_; and
Waitz, _Lehrbuch der Psychologie_, 58. On the side of its physiology
and pathology Taine, Maudsley, and Ribot may be referred to (_On
Intelligence_, _Pathology of Mind_, _Diseases of Memory_), as also a
paper by Herzen, entitled, _Les Modifications de la Conscience du moi_
(_Bull. Soc. Hand. Sc. Nat._, xx. 90). _An Essay on the Philosophy of
Self-consciousness_, by P. F. Fitzgerald, is written from the side of
metaphysics. On this side, also, we are met by the school of Hegel and
the Neo-Kantians with a virtual denial of the origin and development
of self-consciousness in time. Thus, for instance, Green expressly
says:—“Should the question be asked, If this self-consciousness is
not derived from nature, what then is its origin? the answer is, that
it has no origin. It never began because it never was not. It is the
condition of there being such a thing as beginning or end. Whatever
begins or ends does so for it, or in relation to it” (_Prolegomena to
Ethics_, p. 119). To this I can only answer that for my own part I feel
as convinced as I am of the fact of my self-consciousness itself that
it had a beginning in time, and was afterwards the subject of a gradual
development. “Das Ich ist ein Entwicklungsprodukt, wie der ganze Mensch
ein Entwicklungsprodukt ist” (Wundt).

[129] “Of all the neolithic implements the axe was by far the most
important. It was by the axe that man achieved his greatest victory
over nature” (Boyd Dawkins, _Early Man in Britain_, p. 274).

[130] Galton, _Tropical South Africa_, p. 213. The author adds, “Once,
while I watched a Dammara floundering hopelessly in a calculation on
one side of me, I observed Dinah, my spaniel, equally embarrassed on
the other. She was overlooking half a dozen of her new-born puppies,
which had been removed two or three times from her, and her anxiety was
excessive, as she tried to find out if they were all present, or if any
were still missing. She kept puzzling and running her eyes over them,
backwards and forwards, but could not satisfy herself. She evidently
had a vague notion of counting, but the figure was too large for her
brain. Taking the two as they stood, dog and Dammara, the comparison
reflected no great honour on the man.” As previously stated, I taught
the chimpanzee “Sally” to give one, two, three, four, or five straws at
word of command.

[131] The boy’s name was Ernest, and was thus called by all other
members of the household. As I could not find any imitative source of
the dissimilar name used by his sister, this is probably an instance
of the spontaneous invention of names by young children, which has
already been considered at the close of my chapter on “Articulation.”
Touching the use of adjectives by young children, I may quote the
following remark from Professor Preyer:—“A very general error must be
removed, which consists in the supposition that all children on first
beginning to speak use substantives only, and later pass on to the use
of adjectives. This is certainly not the case.” And he proceeds to give
instances drawn from the daily observations of his own child, such as
the use of the word “heiss” in the twenty-third month.

[132] We shall subsequently see that at this stage of mental evolution
there is no well-defined distinction between the different parts of
speech. Therefore here, and elsewhere throughout this chapter, I use
the terms “noun,” “adjective,” “verb,” &c., in a loose and general
sense.

[133] I have seen a terrier of my own (who habitually employed this
gesture-sign in the same way as Preyer’s child, namely, as expressive
of desire), assiduously though fruitlessly “beg” before a refractory
bitch.

[134] Many dogs will significantly bark, and cats significantly mew,
for things which they desire to possess or to be done. For significant
crying by children, see above, p. 158.

[135] For the case of the ape in this connection see above, p. 126.
I took my daughter when she was seven years of age to witness the
understanding of the ape “Sally.” On coming away, I remarked to her
that the animal seemed to be “quite as sensible as Jack”—_i.e._ her
infant brother of eighteen months. She considered for a while, and then
replied, “Well, I think she is sensibler.” And I believe the child was
right.

[136] Or, if any opponent were to suggest this, he would be committing
argumentative surrender. For the citadel of his argument is, as we
know, the faculty of conception, or the distinctively human power of
objectifying ideas. Now, it is on all hands admitted that this power
is impossible in the absence of self-consciousness. Will it, then, be
suggested that my daughter had attained to self-consciousness and the
introspective contemplation of her own ideas before she had attained
to the faculty of speech, and therefore to the very _condition_ to the
naming of her ideas? If so, it would follow that there may be concepts
without names, and thus the whole fortress of my opponents would
crumble away.

[137] See pp. 81-83, where it is shown that even in cases where
conceptual thought is necessary for the original formation of a
name, the name may afterwards be used without the agency of such
thought—just in the same way as actions originally due to intelligence
may, by frequent repetition, become automatic. At the close of the
present chapter it will be shown that the same is true even of full or
formal predication.

[138] In this connection it is interesting to observe the absence of
the copula. Notwithstanding the strongly imitative tendencies of a
child’s mind, and notwithstanding that our English children hear the
copula expressed in almost every statement that is made to them, their
own propositions, while still in the preconceptual phase, dispense with
it (see above, p. 204). In thus trusting to apposition alone, without
expressing any sign of relation, the young child is conveying in
spoken language an immediate translation of the mental acts concerned
in predication. As previously noticed, we meet with precisely the
same fact in the natural language of gesture, even after this has
been wrought up into the elaborate conceptual systems of the Indians
and deaf-mutes. Lastly, in a subsequent chapter we shall see that the
same has to be said of all the more primitive forms of spoken language
which are still extant among savages. So that here again we meet with
additional proof, were any required, of the folly of regarding the
copula as an essential ingredient of a proposition.

[139] See p. 166.

[140] Thus far, it will be observed, the case of predication is
precisely analogous to that of denomination, alluded to in the
foot-note on page 226. Just as instincts may arise by way of “lapsed
intelligence,” so may originally conceptual names, and even originally
conceptual propositions, become worn down by frequent use, until they
are, as it were, degraded into the pre-conceptual order of ideation. Be
it observed, however, that the paragraphs which _follow_ in the text
have reference to a totally different principle—namely, that there may
be propositions strictly conceptual as to form, which, nevertheless,
need never at any time have been conceptual as to thought.

[141] _Logic_, vol. i., p. 108.

[142] _Encyclopædia Britannica_, eighth edition, 1857, Art. “Language.”

[143] Of course in classical times, when there was no theological
presumption against the theory of development, this alternative met
with a fuller recognition; as, for example, by the Latin authors,
Horace, Lucretius, and Cicero. Before that time Greek philosophers had
been much exercised by the question whether speech was an intuitive
endowment (analogists), or a product of human invention (anomalists);
and, earlier still, astonishing progress had been made by the
grammarians of India in a truly scientific analysis of language-growth.
But in the text I am speaking of modern times; and here I think there
can be no doubt that till the middle of the present century the
possibility of language having been the result of a natural growth was
not sufficiently recognized. Among those who did recognize it, Herder,
Monboddo, Sir W. Jones, Schlegel, Bopp, Humboldt, Grimm, and Pott, are
most deserving of mention. The same year that witnessed the publication
of the _Origin of Species_ (1859), gave to science the first issue of
Steinthal’s _Zeitschrift für Völkerpsychologie und Sprachwissenschaft_.
From that date onwards the theory of evolution in its application to
philology has held undivided sway.

[144] _Encycl. Brit._, _loc. cit._ Remembering that the above was
published two years before the _Origin of Species by means of Natural
Selection_, this clear enunciation of the struggle for existence in the
field of philology appears to me deserving of notice.

[145] _Science of Thought_, preface, p. xi.

[146] _Darwinism tested by the Science of Language_, p. 41.

[147] There is a difference of opinion among philologists as to the
extent in which modifying constants were themselves originally roots.
The school of Ludwig regards demonstrative elements as never having
enjoyed existence as independent words; but, even so, they must have
had an independent existence of some kind, else it is impossible to
explain how they ever came to be employed as constantly modifying
different roots in the same way. Moreover, as Max Müller well observes,
“to suppose that Khana, Khain, Khanana, Khaintra, Khatra, &c., all
tumbled out ready-made, without any synthetical purpose, and that
their differences were due to nothing but an uncontrolled play of the
organs of speech, seems to me an unmeaning assertion.... What must be
admitted, however, is that many suffixes and terminations had been
wrongly analyzed by Bopp and his school, and that we must be satisfied
with looking upon most of them as in the beginning simply demonstrative
and modificatory” (_loc. cit._, pp. 224 and 225). See also Farrar,
_Origin of Language_, pp. 100, _et seq._; Donaldson, _Greek Grammar_,
pp. 67-79; and Hovelacque, _Science of Language_, p. 37. It will be
remarked that this question does not affect the exposition in the text.

[148] _Grundriss der Sprachwissenschaft_, I. i. 77. This estimate is
accepted by Professor Sayce, _Introduction to the Science of Language_,
vol. ii., p. 32.

[149] Hovelacque, _Science of Language_, English trans., p. 37.

[150] This method of representation was devised by Schleicher, who
carries it further than I have occasion to do in the text. See _Memoirs
of Academy of St. Petersburg_, vol. i., No. 7, 1859.

[151] Hovelacque, _loc. cit._, p. 130.

[152] Sayce, _Introduction, &c._, i. 126.

[153] _Introduction, &c._, vol. i., p. 374.

[154] _Ibid._, vol. i., pp. 375, 376.

[155] _Ibid._, p. 120. See also his _Principles of Comparative
Philology_, 2nd ed., p. ix.

[156] Sayce, _Introduction, &c._, i., 125, 126.

[157] Hovelacque, _Science of Language_, p. 130.

[158] “What we most need to note is the very narrow limitation of
our present knowledge. Even among the neighbouring families like the
Algonquin, Troquois, and Dakota, whose agreement in style of structure
(polysynthetic), taken in connection with the accordant race-type of
their speakers, forbids us to regard them as ultimately different, no
material correspondence, agreements in words and meanings, is to be
traced; and there are in America all degrees of polysynthetism, down
to the lowest, and even to its entire absence. Such being the case, it
ought to be evident that all attempts to connect American languages as
a body with languages of the Old World are, and must be, fruitless:
in fact, all discussions of the matter are at present unscientific”
(Professor Whitney in _Encycl. Brit._, art. “Philology,” 1885).

[159] _Introduction, &c._, i. 120.

[160] _Ibid._, i. 116.

[161] “The number of separate families of speech now existing in
the world, which cannot be connected with one another, is at least
seventy-five; and the number will doubtless be increased when we have
grammars and dictionaries of the numerous languages and dialects which
are still unknown, and better information as regards those with which
we are partially acquainted. If we add to these the innumerable groups
of speech which have passed away without leaving behind even such
waifs as the Basque of the Pyrenees, or the Etruscan of ancient Italy,
some idea will be formed of the infinite number of primæval centres or
communities in which language took its rise” (Sayce, _Introduction,
&c._, ii. 323).

[162] _Life and Growth of Language_, p. 259.

[163] _Ibid._, p. 262.

[164] I may add that the hypothesis admits of corroboration from
sources not mentioned by its author. For Archdeacon Farrar wrote in
1865:—“The neglected children in some of the Canadian and Indian
villages, who are left alone for days, can and do invent for themselves
a sort of _lingua franca_, partially or wholly unintelligible to
all except themselves;” and he quotes Mr. R. Moffat as “testifying
to a similar phenomenon in the villages of South Africa (_Mission
Travels_).” He also alludes to the fact that “deaf-mutes have an
instinctive power to develop for themselves a language of signs,”
which, as we have seen in an earlier chapter, embraces the use of
arbitrary articulations, even though in this case the speakers cannot
themselves hear the sounds which they make.

While this work is passing through the press an additional paper has
been published by Dr. Hale, entitled, _The Development of Language_. It
supplies further evidence in support of this hypothesis.

[165] Wundt, _Vorlesungen, &c._, ii., 380, 381.

[166] Sayce, _Introduction to Science of Language_, ii, 13.

[167] The difference of opinion in question seems to arise from
individual prepossessions with regard to the ulterior question
whether or not the aboriginal roots of all languages must have been
polysyllabic. For my own part, and for the reasons already given, I can
see no presumption in favour of the view that primitive languages must
all have presented the “polysinthetic genius.”

[168] _Histoire des Langues Semitique_, p. 138.

[169] _Etymological Dictionary_, p. 746.

[170] See Max Müller, _Science of Thought_, p. 332.

[171] _Ibid._, p. 404.

[172] _Ethnologische Forschungen_, ii., s. 73, _et seq._ He here quotes
Varro to the effect that the roots of Latin amount to about a thousand.

[173] _Language and the Study of Language_, p. 256.

[174] Sayce, _Introduction to the Science of Language_, ii., p. 4.

[175] Geiger, _Ursprung der Sprache_, s. 16.

[176] Sayce, _loc. cit._, ii. p. 6.

[177] Wedgwood, _Etymol. Dict._, p. iii.

[178] Farrar, _Origin of Language_, p. 53.

[179] _Science of Thought_, p. 439.

[180] _Science of Thought_, p. 549.

[181] _Science of Thought_, pp. 551, 552.

[182] _Ibid._, pp. 551, 552.

[183] “The Aryan languages are the languages of a civilized race; the
parent speech to which we may inductively trace them was spoken by men
who stood on a relatively high level of culture” (Sayce, _Introduction,
&c._, i. 56). “The primitive tribe which spoke the mother-tongue
of the Indo-European family was not nomadic alone, but had settled
habitations, even towns and fortified places, and addicted itself in
part to the rearing of cattle, in part to the cultivation of the earth.
It possessed our chief domesticated animals—the horse, the ox, the
goat, and the swine, besides the dog: the bear and the wolf were foes
that ravaged its flocks; the mouse and the fly were already domestic
pests.... Barley, and perhaps also wheat, was raised for food, and
converted into meal. Mead was prepared from honey, as a cheering and
inebriating drink. The use of certain metals was known; whether iron
was one of them admits of question. The art of weaving was practised;
wool and hemp, and possibly flax, being the materials employed....
The weapons of offence and defence were those which are usual among
primitive peoples, the sword, spear, bow, and shield. Boats were
manufactured and moved by oars.... The art of numeration was learned,
at least up to a hundred; there is no general Indo-European word for
‘thousand.’ Some of the stars were noticed and named; the moon was the
chief measurer of time. The religion was polytheistic, a worship of
the personified powers of nature” (Whitney, _Language and the Study
of Language_, pp. 207, 208). For a more detailed account of this
interesting people, see Poescher, _Die Arier_.

[184] “Unsere Wurzeln sind die Urwurzeln nicht; wir haben vielleicht,
von keiner einzigen die erste, ursprüngliche Laut-form mehr vor uns,
ebensowenig wohl die Urbedeutung” (Geiger, _Ursprung der Sprache_, s.
65). And this opinion, so far as I know, is adopted as an axiom by all
other philologists.

[185] “It is impossible to bring down the epoch at which the Aryan
tribes still lived in the same locality, and spoke practically the same
language, to a date much later than the third millennium before the
Christian era” (Sayce, _Introduction_, _&c._, ii., p. 320).

[186] This fact alone would be sufficient to dispose of what I cannot
but consider, from any and every point of view, the transparent
absurdity of the doctrine that “the formation of thought is the first
and natural purpose of language, while its communication is accidental
only” (_Science of Thought_, p. 40). Such a “purpose” would imply
“thought” as already formed; and, therefore, the doctrine must suppose
a purpose to precede the conditions of its own possibility.

[187] I use the term “verbs” merely for the sake of brevity and
clearness. Of course there cannot have been verbs, strictly so-called,
before there were parts of speech of any kind. The more accurate
statement is given in the next sentence, and is the one which I desire
to be understood hereafter in the short-hand expression “verbs.”

[188] “It must be borne in mind that primitive man did not distinguish
between phenomena and volitions, but included everything under the head
of actions, not only the involuntary actions of human beings, such
as breathing, but also the movements of inanimate things, the rising
and setting of the sun, the wind, the flowing of water, and even such
purely inanimate phenomena as fire, electricity, &c.; in short, all
the changing attributes of things were conceived as voluntary actions”
(Sweet, _Words, Logic and Grammar_, p. 486).

[189] As a matter of fact, and as we shall subsequently see, there is
an immense body of purely philological evidence to show that verbs are
really a much later product of linguistic growth than either nouns or
pronouns. This is proved by their comparative paucity in many existing
languages of low development (their place being taken by pronominal
appositions, &c.); and also by tracing the origin of many of them to
other parts of speech. (See especially Garnett’s _Essays, Pritchard on
the Celtic Languages_, _Quart. Rev._, Sept. 1876; _The Derivation of
Words from Pronominal and Prepositional Roots_, _Proc. Philol. Soc._
vol. ii.; and _On the Nature and Analysis of the Verb_, ibid., vol.
iii.) Later on it will be shown that in the really primitive stages of
language-growth there is no assignable distinction between any of the
parts of speech. Archdeacon Farrar well remarks, “The invention of a
verb requires a greater effort of abstraction than that of a noun....
We cannot accept it as even _possible_ that from roots meaning _to
shine_, _to be bright_, names were formed for _sun_, _moon_, _stars_,
&c.... In some places, indeed, Professor Müller appears to hold the
correct view, that at first ‘roots’ stood for any and every part of
speech, just as the monosyllabic expressions of children do” (_Chapters
on Language_, pp. 196, 197; see, also, some good remarks on the subject
by Sir Graves Haughton, _Bengali Grammar_, p. 108).

[190] “Standst du dabei, als sich der Brust des noch stummen Urmenschen
der erste Sprachlaut entrang? und verstandst du ihn? Oder hat man
dir die Urwurzeln jener ersten Menschen vor hundert tausend Jahren
überliefert? Sind das, was du als Wurzeln hinstellst, und was wirklich
Wurzeln sein mögen, auch Wurzeln der Urzeit, unveränderte Reflexlaute?
Sind jene deine Wurzeln älter als sechstausend, als zehntausend Jahre?
und wie viel mögen sie sich in den früheren Jahrzehntausenden verändert
haben? wie mag sich ihre Bedeutung verändert haben?” (Steinthal,
_Zeits. b. Volkerpysch. u. Sprachwiss._, 1867, s. 76).

[191] _Supra_, p. 68, _et seq._

[192] _Ursprung der Sprache_, s. 74. To the same effect, and from
the side of psychology, I may quote Wundt:—“Oft hat man desshalb in
der Sprache einen Ubergang vom Abstrakten zum Konkreten zu finden
geglaubt, weil dieselbe thatsächlich zunächst umfassendere, dann
individuellere Vorstellungen bezeichnet und erst zuletzt wieder die
Namen individueller Objekte zu Gemeinnamen stempelt. Aber was am Anfang
dieser Reihe liegt ist etwas ganz anderes als was den Schluss derselben
bildet: Gemeinnamen sind wirkliche Zeichen für Allgemeinvorstellungen
und Begriffe. Jene ersten Vorstellungen, welche das Bewusstsein bildet
und die Sprache ausdrückt, sind nicht _Allgemein_vorstellungen sondern
_umfassende_ Vorstellungen. Beides ist wesentlich aus einander zu
halten” (_Vorlesungen, &c._, ii. 382). The passage then proceeds to
discuss the psychology of the subject.

[193] _Introduction, &c._, ii. 5, 6.

[194] And even as regards this minority (such as “to be,” “to think,”
“to do,” &c.), we must remember an important consideration on which
Geiger bestows a number of excellent pages. Briefly put, this
consideration is that the offspring of words are everywhere proved
to have progressively changed their meanings by successive steps and
in divergent lines: applying this general law to the case of roots,
it follows that the oldest meaning which philology is able to trace
as expressed by a root, need not be anywhere near the meaning which
attached to its remoter parents: the latter may have been much less
conceptual.

[195] Professor Max Müller says in one place, “The Science of Language,
by inquiring into the origin of general terms, has established two
facts of the highest importance, namely, first, that all terms were
originally general; and, secondly, that they could not be anything
but general” (_Science of Thought_, p. 456). Elsewhere, however, he
says, “Although during the time when the growth of language becomes
historical and most accessible, therefore, to our observation, the
tendency certainly is from the general to the special, I cannot
resist the conviction that before that time there was a pre-historic
period during which language followed an opposite direction. During
that period roots, beginning with special meanings, became more and
more generalized, and it was only after reaching that stage that they
branched off again into special channels” (_ibid._, pp. 383, 384).
Again, in his earlier work on the _Science of Language_ (vol. i.,
pp. 425-432), he argues in favour of terms having been aboriginally
general. It will thus be seen that with reference to this question he
is not consistent. Touching the first of his doctrines above quoted,
Geiger pertinently observes that against such a conclusion there lies
the obvious absurdity, that if a language were to consist exclusively
of general terms, it would be _ipso facto_ unintelligible to its own
speakers; “for what hope could there be of any mutual understanding
with a language comprising only such words as “to bind,” “to sound,”
&c.? (_Ursprung der Sprache_, s. 16). Clearly, Professor Max Müller’s
difficulties regarding this subject are quite imaginary, and would
disappear if he were to entertain the natural alternative that there is
no reason to suppose aboriginal words were exclusively restricted to
being either special or general—_i.e._ generic.

[196] Bunsen, _Philosophy of Universal History_, ii. 131.

[197] Professor Max Müller in all his works; but it is observable that
his opposition to what he calls the “bow-wow and pooh-pooh theory” was
more strenuous in his earlier publications than it is in his later.

[198] It is needless to say that innumerable instances might be quoted
of this metaphorical change in the meanings of words, even in existing
languages,—so much so, indeed, that, as Richter says, all languages
are but dictionaries of forgotten metaphors. For example, there is
a single Hebrew word of three letters which may bear any one of the
following significations:—to mix, to exchange, to stand in place of,
to pledge, to interfere, to be familiar, to disappear, to set, to do
a thing in the evening, to be sweet, a fly or beetle, an Arabian, a
stranger, the weft of cloth, the evening, a willow, and a raven. (See
Farrar, _Chapters on Language_, p. 229. He adds, “Assuming that all
these significations are ultimately deducible from one and the same
root, we see at once the extent to which metaphor must have been at
work.” For further examples of the same principle, see _ibid._, pp.
234, 251, 252.)

[199] _Science of Thought_, pp. 317, 318.

[200] Or, as Heyse puts it, many onomatopœias are not “old fruitful
roots of language, but modern inventions which remain isolated in
language, and are incapable of originating any families of words,
because their meaning is too limited and special to admit of a
manifold application” (_System_, s. 92, quoted by Farrar, _Chapters on
Language_, p. 152, who also shows that words of onomatopoetic origin
are not invariably sterile. When such origin is not so remote as to
have become wholly obscured by a widely connotative extension, it
does remain possible to trace its progeny through areas of smaller
extension).

[201] “Nichtsdestoweniger bleibt es eine wichtige psychologische
Thatsache, dass die Laute einen onomatopoetischen Werth haben, dass
wir diesen Werth heute noch fühlen. Nur ist dieses Gefühl nicht sicher
genug, um als wissenschaftlicher Beweis zu gelten, wie es denn auch bei
den verschiedenen Racen verschieden ist. Die Sprachen der mongolischen
Race haben zur Bezeichnung von Naturereignissen viele Onomatopöien,
welche wir nicht mitfühlen. Und das ist weder zu verwundern, noch ist
es ein Beweis gegen die geistige Einheit des Menschengeschlechtes. Das
Gefühl wird ja vielfach durch Associationen der Vorstellungen bestimmt.
Andere Associationen aber walten im Kaukasier, andere im Mongolen”
(_Zeits. b. Volkerpsych. u. Sprachwissen._, 1867, s. 76).

[202] _Introduction, &c._, i., p. 108. He points out that “_bilbit_,
_glut-glut_, and _puls_, are all attempts to represent the same sound.”

[203] _Chapters on Language_, p. 154.

[204] _Ueber Namen des Donners_, 1855.

[205] Steinthal’s _Zeitschrift_, &c.

[206] Professor Max Müller has argued that in the Indo-European
languages the apparently onomatopoetic words signifying “thunder” are
derived from the root _tan_, to “stretch,” and therefore were not of
imitative origin. But Farrar has satisfactorily met this objection,
even as regards this one particular case, by showing that even if not
originally onomatopoetic, these words afterwards “became so from a
feeling of the need that they should be” (_Origin of Language_, p. 82).
See also, _Chapters on Language_, pp. 178-182; Heyse, _System_, s. 93;
and Wundt, _Vorlesungen, &c._, ii. 396.

[207] See also Nodier, _Dictionnaire des Onomatopées_; and Wedgwood,
_Dictionary of English Etymology_.

[208] Probably the explanation of this apparent inconsistency is to be
found in the fact that Noiré’s special version of the onomatopoetic
theory comes within easy distance of a hypothesis which Max Müller had
himself previously sanctioned. This hypothesis, originally propounded
by Heyse in his _System der Sprachwissenschaft_, is that, just as
every inorganic substance in nature gives out a particular sound when
struck—metal one sound, wood another, stone another, &c.—so different
animals have inherent tendencies (or “instincts”) to emit distinctive
sounds. In the case of primitive man this inherent tendency was in the
direction of articulate speech. For my own part, I do not see that this
theory explains anything; and therefore agree with Geiger, who says of
it:—“Die Annahme eines jetzt erloschenen Vermögens der Sprachschöpfung
und die damit zusammenhängende von einem vollkommenen Urzustande des
Menschen ist eine Zuflucht zum Unbegreiflichen, und nicht weit von dem
Eingeständnisse entfernt, dass es uns der Natur der Dinge nach für
immer unmöglich sei, den wahren Sinn der Urwurzeln zu erkennen und den
Vorgang des Sprachursprunges zu erklären. Wir würden mit einer solchen
Annahme auf einen mystischen Standpunkt zurückgeführt sein, da doch
schon Herder das ‘Gespenst vom Wort Fähigkeit’ bekämpft und gesagt
hat: ‘Jch gebe den Menschen nicht gleich plötzlich neue Kräfte, keine
sprachschaffende Fähigkeit, wie eine willkürliche qualitas occulta’”
(_Ursprung der Sprache_, s. 24). Sayce, also, well remarks of this
hypothesis, “It really rests upon an _a priori_ conception of the
origin of speech, which is neither borne out by linguistic facts nor
easily intelligible.... Such a theory of language is plainly mystical”
(_Introduction to Science of Language_, vol. i., pp. 66, 67).

[209] _Encyclo. Brit._, art. “Philology,” vol. xviii., p. 769.

[210] See, for instance, Farrar, _Chapters on Language_, p. 184.

[211] See above, pp. 138-144.

[212] See above, pp. 121, 122.

[213] See _Vorlesungen, &c._, ii. 394, 395.

[214] See above, pp. 132-136.

[215] _Introduction to the Science of Language_, ii. 302.

[216] See above, pp. 138-143.

[217] _Der Ursprung der Sprache_, s. 31. His own answer to the
question is as follows:—“Sind die Wörter Produkte der Natur order der
Willkür? Beides und beides nicht. Kein Wort hat naturnothwendig seine
bestimmte Bedeutung; insofern sind sie alle willkürlich: aber keines
ist zu seiner Bedeutung durch menschliche Willensthätigkeit gekommen”
(_ibid._, s. 113).

[218] Schelling, _Einl. in die Philos. d. Mythologie_, s. 51.

[219] _Anthropologie der Naturvölker_, i., 272. See also, F. Müller,
_Grundriss der Sprachwissenshaft_, I. i. 49.

[220] _Science of Language_, ii. 91, 92.

[221] _Grund. d. Sprachwiss._, i., 43.

[222] _Ægypten_, i. 324.

[223] Sayce, _Introduction, &c._, i. 119, 120.

[224] _Science of Thought_, 423-440.

[225] Sayce, _Introduction, &c._, i. 111.

[226] _Ibid._, i. 113, 114.

[227] Sayce, _Introduction, &c._, i. 121.

[228] _Science of Thought_, p. 242.

[229] Garnett, _Philolo. Essays_, p. 87.

[230] _Ibid._, 77, 78.

[231] Farrar, _Origin of Language_, p. 99. The passage continues,
“We might have conjectured this from the fact already noticed, that
children learn to speak of themselves in the third person—_i.e._
regard themselves as objects—long before they acquire the power of
representing their material selves as the instrument of an abstract
entity.” He also alludes to “some admirable remarks to this effect in
Mr. F. Whalley Harper’s excellent book on the _Power of Greek Tenses_;”
and recurs to the subject in his more recently published _Chapters on
Language_, p. 62. I could quote other authorities who have commented
upon this philological peculiarity of early pronouns; but will only
add the following in order to show how the peculiarity in question may
continue to survive even in languages still spoken. “The Malay _ulun_,
‘I,’ is still ‘a man’ in Lampong, and the Kawi _ugwang_, ‘I,’ cannot
be separated from _nwang_, ‘a man’” (Sayce, _Introduction_, ii. 26).
Lastly, Wundt has pointed out that this impersonal form of speech is
distinctive, not only of early pronominal elements, but also of early
forms of predication. For instance, “Die ersten Urtheile, die in das
Bewusstsein hereinbrechen, _subjektlose_ Urtheile sind, und dass die
Prädikate derselben stets eine sinnliche Vorstellung ausdrücken. ‘Es
leuchtet es glänzt, es tönt,’—solcher Art sind die Urtheile, die
der Mensch zuerst denkt und zuerst ausspricht. Jenes Prädikat, dass
sogleich bei der Wahrnehmung eines Gegenstandes sich aufdrängt, wird
zur Bezeichnung des Gegenstandes selber. ‘Das Leuchtende, Glänzende,
Tönende,’—solcher Art find die Wörter, die ursprünglich in der Sprache
gebildet werden” (_loc. cit._, ii. 377).

[232] _Science of Thought_, p. 221.

[233] _Ibid._, p. 554.

[234] _Ibid._, 241.

[235] Sayce, _Introduction, &c._, ii. 25; see also to the same effect,
Bleek, _Ursprung der Sprache_, 70-72; F. Müller, _Grundriss der
Sprachwissenshaft_, I., i., s. 40; and Noiré, _Logos_, p. 186. The
chief ground of this scepticism is that it is difficult to conceive how
a word could ever have gained a footing if it did not from the first
present some independent predicative meaning. But it seems to me that
the force of this objection is removed if we remember the sounds which
are arbitrarily invented by young children and uneducated deaf-mutes,
not to mention the inarticulate clicks of the Bushmen. Moreover, there
is nothing inimical to the pronominal theory in the supposition that
pronominal elements, even of the most aboriginal kind, were survivals
of still more primitive sentence-words—a supposition which would of
course remove the difficulty in question. But, as explained in the
text, this difficulty, even if it could not be thus met, would really
not be one of any importance to my exposition.

[236] _Introduction, &c._, i. 117.

[237] _Introduction, &c._, ii. 301. Or, as Wundt puts it, “Die
demonstrative Wurzel ist daher eine demonstrirende Pantomime in einen
Laut übersetzt” (_Vorlesungen, &c._, ii. 392).

[238] Sayce, _Introduction, &c._, i. 415. See also F. Müller, _loc.
cit._, I. i. 2, p. 2, for another statement of the same facts referred
to by Sayce.

[239] Sayce, _Introduction, &c._, i. 416.

[240] Sweet, _Words, Logic, and Grammar_, in _Trans. Philo. Soc._,
1867, p. 493.

[241] _Science of Thought_, p. 442.

[242] See especially Garnett, _On the Nature and Analysis of the Verb_.

[243] _Science of Thought_, p. 223.

[244] _Ibid._, p. 442.

[245] Sayce, _Introduction, &c._

[246] I refer the reader to what is said on both these aspects of the
verb in question by my opponents (see pp. 165-167.)

[247] Farrar, _Origin of Language_, pp. 105, 106.

[248] Garnett, _On the Nature and Analysis of the Verb_, _Proc. Philo.
Soc._, vol. iii.

[249] Sayce, _Introduction, &c._, i. 415.

[250] Geiger, _Development of the Human Race_, English trans., p. 22.

[251] Sweet, _Words, Logic, and Grammar_, in _Trans. Philol. Soc._,
1876, pp. 486, 487.

[252] Sweet, _loc. cit._, pp. 489, 490.

[253] Bleek, _Ursprung der Sprache_, s. 69, 70.

[254] _Science of Thought_, p. 241.

[255] Steinthal, _Charakteristik, &c._, 165, 173.

[256] Garnett, _Philological Essays_, p. 310.

[257] _Ibid._, p. 311.

[258] _Ibid._, p. 312.

[259] _Ibid._, p. 314.

[260] See Chapter on Speech, p. 166.

[261] I may remark that it was Aristotle who first fell into the error
of identifying the copula with the verb _to be_, by which it happens
to be expressed in Greek. For many centuries afterwards this error
was a fruitful source of endless confusions; but it is curious to
find a wholly new fallacy springing from it in the latter half of the
nineteenth century. Touching the subject and predicate, Aristotle, of
course, never contemplated any more primitive relation between them
than that which obtained in the only forms of speech with which he
was acquainted. As regards his “categories” the following remarks by
Professor Max Müller are worth quoting:—

“These categories, which proved of so much utility to the early
grammarians, have a still higher interest to the students of the
science of language and thought. Whereas Aristotle accepted them simply
as the given forms of predication in Greek, after that language had
become possessed of the whole wealth of its words, we shall have to
look upon them as representing the various processes by which those
Greek words, and all our own words and thoughts, too, first assumed
a settled form. While Aristotle took all his words and sentences as
given, and simply analyzed them in order to discover how many kinds of
predication they contained, we ask how we ever came into possession of
such words as _horse_, _white_, _many_, _greater_, _here_, _now_, _I
stand_, _I fear_, _I cut_, _I am cut_. Anybody who is in possession
of such words can easily predicate, but we shall now have to show
that every word by itself was from the first a predication, and that
it formed a complete sentence by itself. To us, therefore, the real
question is, how these primitive sentences, which afterwards dwindled
away into mere words, came into existence. The true categories, in
fact, are not those which are taught by grammar, but those which
produced grammar, and it is these categories which we now proceed to
examine” (_Science of Thought_, p. 439).

[262] Sayce, _Introduction, &c._, ii. 229. He adds, “Had Aristotle been
a Mexican, his system of logic would have assumed a wholly different
form.”

[263] _Introduction, &c._, i, 15.

[264] In these considerations I find myself able largely to reconcile
what has always been regarded as a contradiction between the views
of Professor Whitney and those of other philologists on the subject
of sentence-words. Partly following Schleicher—who maintains the
doctrine still more unequivocally—he regards the word as having
been historically prior to the sentence. This, of course, is in
contradiction to the doctrine of the sentence having been historically
prior to the word, which, as we have seen, is the doctrine now held
by philologists in general. But, now, what the latter doctrine
really amounts to is, that words were sentences before they were
names—predicative before they were nominative; and, as I understand
it, Whitney’s objection to this doctrine is really raised on grounds of
psychology. If so, the above considerations show that he is perfectly
right. Intellectually, primitive man was fully capable of acquiring the
use of words as names; and, therefore, psychologically considered, it
was only an accident of social environment which prevented him from so
doing.

[265] _Science of Thought_, pp. 432, 433.

[266] Pp. 281, 282, note.

[267] _Ursprung der Sprache_, s. 65. For the original German, see the
passage as previously quoted on page 273, note.

[268] As pointed out in a previous chapter, curious ambiguity attaches
to this term. For, as used in biology, it means the _hitherto
undifferentiated_, while in psychology and elsewhere a “generalization”
means the _synthetically integrated_. But, as psychologists never speak
of ideas as “generalized,” I here use the word in its biological sense.
See also above, pp. 277-280.

[269] _Ursprung der Sprache_, s. 69, 70.

[270] Bleek entertains no doubt on this point.

[271] Compare also close of Chapter VII. (pp. 138-144), where the
children mentioned by Dr. Hale are shown to have adopted the syntax of
gesture-language in their spontaneously devised spoken language.

[272] Chapter VI., pp. 114-120.

[273] _Sign-Language, &c._, p. 284. On page 352, this writer further
supplies a most interesting comparison between gesture and spoken
language as both are used by the North American Indians—showing that
the syntax in the two cases is identical.

[274] Whitney, _Encyclo. Brit._, _loc. cit._, p. 770. It is interesting
to note that the psychological importance of this principle was clearly
enunciated by Locke:—“It may lead us a little towards the original
of all our notions and knowledge, if we remark how great a dependence
our words have on common sensible ideas; and how those which are made
use of to stand for actions and notions quite removed from sense, have
their rise from thence, and from obvious sensible ideas are transferred
to more abstruse significations, and made to stand for ideas that come
out under the cognizance of our senses” (_Human Understanding_, iii. i.
5).

[275] Whitney, _Encyclo. Brit._, p. 770. See also Nodier, _Notions de
Linguistique_, p. 39; Garnett, _Essays_, p. 89; Grimm, _Gesch. d. d.
Sprache_, s. 56 _et seq._; Pott, _Metaphern vom Leben, &c._, _Zeitschr.
fur Vergl. Sprachf. Jahrg._, ii., heft 2; Heyse, _System, &c._, s. 97;
and Farrar, _Origin of Language_, 130; _Chapters on Language_, pp. 67,
133, 204-246. He refers to the above, and quotes the following passages
from Emerson and Carlyle:—“As the limestone of the Continent consists
of infinite masses of shells of animalcules, so language is made up
of images and tropes, which now, in their secondary use, have long
ceased to remind us of their poetic origin” (_Essays on the Poets_).
“Language is the flesh-garment of Thought. I said that Imagination wore
this flesh-garment; and does she not? Metaphors are her stuff. Examine
Language. What, if you except a few primitive elements of natural
sound, what is it all but metaphors recognized as such, or no longer
recognized; still fluid and florid, or now solid-grown and colourless?
If those same primitive elements are the osseous fixtures in the
flesh-garment of Language—then are metaphors its muscles, its tissues,
and living integuments. An unmetaphorical style you shall in vain
seek for: is not your very _attention_ a _stretching-to_?” (_Sartor
Resartus_, ch. x.).

[276] _Science of Thought_, p. 329.

[277] _Science of Language_, p. 123.

[278] _Logos_, p. 258, _et seq._

[279] Geiger, _Address delivered before the International Congress for
Archæology and History at Bonn_, 1868.

[280] Geiger, _A Lecture to the Commercial Club of
Frankfort-on-the-Main_ (1869).

[281] Perhaps the most interesting department of fundamental metaphor
is that wherein the metaphor is found by philological research to
have reference, not to any natural object, quality, &c., but to a
pre-existing action or gesture as already made by man himself for the
purpose of conveying information, expressing his emotions, &c. For
fundamental metaphor of this kind obviously brings us within seeing
distance of the time when the audible signs of articulations were born
of the visible signs of gesture and grimace. In illustration of this
branch of our subject I will only quote one passage; but the reader
will at once perceive how easy it would be to furnish many other
instances from the etymology of words now in habitual use.

“The further a language has been developed from its primordial roots,
which have been twisted into forms no longer suggesting any reason
for their original selection, and the more the primitive significance
of its words has disappeared, the fewer points of contact can it
retain with signs. The higher languages are more precise because the
consciousness of the derivation of most of their words is lost, so that
they have become counters, good for any sense agreed upon and for no
other.

“It is, however, possible to ascertain the included gesture even in
many English words. The class represented by the word _supercilious_
will occur to all readers, but one or two examples may be given not
so obvious and more immediately connected with the gestures of our
Indians. _Imbecile_, generally applied to the weakness of old age,
is derived from the Latin _in_, in the sense of on, and _bacillum_,
a staff, which at once recalls the Cheyenne sign for _old man_
[previously mentioned]. So _time_ appears more nearly connected
with [Greek: teinô], to stretch, when information is given of the
sign for _long time_, in the Speech of Kin Chē-ĕss, in this paper,
namely, placing the thumbs and forefingers in such a position as if
a small thread was held between the thumb and forefinger of each
hand, the hands first touching each other, and then moving slowly
from each other, as if _stretching_ a piece of gum-elastic” (Mallery,
_Sign-Language, &c._, p. 350). This writer also says, with reference
to the uncivilized languages which he has specially studied, “In the
languages of North America, which have not become arbitrary, to the
degree exhibited by those of civilized man, the connection between
the idea and the word is only less obvious than that still unbroken
connection between the idea and the sign, and they remain strongly
affected by the concepts of outline, form, place, position, and feature
on which gesture is founded, while they are similar in their fertile
combination of radicals. Indian language consists of a series of
words that are but slightly differentiated parts of speech following
each other in the order suggested in the mind of the speaker without
absolute laws of arrangement, as its sentences are not completely
integrated. The sentence necessitates parts of speech, and parts of
speech are possible only when a language has reached that stage where
sentences are logically constructed. The words of an Indian tongue,
being synthetic or undifferentiated parts of speech, are in this
respect strictly analogous to the gesture elements which enter into
a sign-language. The study of the latter is therefore valuable for
comparison with the words of the former. The one language throws much
light upon the other, and neither can be studied to the best advantage
without a knowledge of the other.”

[282] There are certain writers, such as Du Ponceau, Charlevoix,
James, Appleyard, Threlkeld, Caldwell, &c., who have sought to
represent that the languages of even the lowest savages are “highly
systematic and truly philosophical,” &c. But this opinion rests on
a radically false estimate of the criteria of system and philosophy
in a language. For the criteria chosen are exuberance of synonyms,
intricacies or complications of forms, &c., which are really works
of a low development. The fallacy is now acknowledged to be such
by all philologists. Even Farrar, who at first himself fell into
this error (_Origin of Language_, p. 28), in his subsequent work
writes:—“Further examination has entirely removed this belief. For
this apparent wealth of synonyms and grammatical forms is chiefly due
_to the hopeless poverty of the power of abstraction_. It would not
only be no advantage, but even an impossible encumbrance to a language
required for literary purposes. The transnormal character of these
tongues only proves that they are the work of minds incapable of all
subtle analysis, and following in one single direction an erroneous
and partial line of development.... If language proves anything, it
proves that these savages must have lived continuously in a savage
condition” (Farrar, _Chapters on Language_, pp. 53, 54, who also refers
to numerous authorities).

[283] The term “conception” here is, of course, equivalent to my term
“pre-conception.” When my daughter uttered her first denotative word
“star,” she was, indeed, bestowing a name; but it was the name of a
recept, not of a concept.

[284] Farrar, _Chapters on Language_, pp. 198, 199.

[285] _Mithridates_, iii. 325, 397. See also Pott, _Etym. Forsch._, ii.
167; and Heyse, _System_, 132.

[286] Latham, _Races of Man_, p. 376.

[287] Quatrefages, _Rev. des Deux Mondes_, Dec. 15, 1860; Maury, _La
Terre et l’Homme_, p. 433.

[288] _Mem. sur le Syst. Gram., &c._, p. 120.

[289] _Malay Grammar_, i., p. 68, _et seq._

[290] _Journl. Ameri. Orient, Soc._, i. No. 4, p. 402.

[291] Casalis, _Grammar_, p. 7.

[292] Pickering, _Indian Languages_, p. 26.

[293] _Vocabulary of the Dialects of some of the Aboriginal Tribes of
Tasmania_, p. 34.

[294] _Introduction, &c._, vol. ii., p. 6.

[295] _Ibid._, vol. i., p. 379.

[296] _A Lecture delivered at Frankfort_, 1869.

[297] _Science of Thought_, p. 245.

[298] _Essays_, p. 89.

[299] _Chapters on Language_, p. 133.

[300] Herder, _Abhandl._, s. 122.

[301] _Das Leben der Seele_, ii. 47.

[302] _Grundriss der Sprachwissenschaft_, i. 35, 36.

[303] See, for example, F. Müller, _loc. cit._, i. 36, 37.

[304] Some of the supporters of the interjectional theory in this
extreme, not to say extravagant form, appear to go on the assumption
that primitive and hitherto speechless man already differed from the
lower animals in presenting conceptual thought. This assumption would,
of course, explain why man alone began to invest his instinctive cries,
&c., with the character of names. But, from a psychological point of
view, any such assumption is obviously a putting of the cart before
the horse. I make this remark in order to add that the objection would
not apply if the ideation were supposed to be _pre-conceptual_—_i.e._
beyond the level reached by any brute, though not yet distinctively
human. Later on, I myself espouse a theory to this effect.

[305] _E.g._ by Mr. Ward, in his _Dynamical Sociology_.

[306] Differences of opinion are entertained by philologists concerning
the value of “nursery-language,” or “baby-talk,” as a guide to the
probable stages of language-growth in primitive man. Without going into
the arguments upon this question on either side, it appears to me that
the analogy as above limited cannot be objected to even by the most
extreme sceptics upon the philological value of infantile utterance.
And it is only to this extent that I anywhere use the analogy.

[307] For cases, see Heinieke, _Beobachtungen über Stumme_, s. 137, &c.

[308] _Ibid._, s. 73.

[309] _Mental Evolution in Animals_, p. 238.

[310] The carnivorous habits of this animal (which is named as a new
species) are most interesting. It is surmised that in its wild state
it must live upon birds; but in the Zoological Gardens it is found to
show a marked preference for cooked meat over raw. It dines off boiled
mutton-chops, the bones of which it picks with its fingers and teeth,
being afterwards careful to clean its hands. It mixes a little straw
with the mutton as vegetables, and finishes its dinner with a dessert
of fruits. But a more important point is that this animal answers its
keeper in vocal tones—or rather grunts—when he speaks to it, and
these tones are understood by the keeper as indicative of different
mental states. I have spent a great deal of time in observing this
animal, but the publicity and other circumstances render it difficult
to do much in the way of experiment or tuition. With regard to teaching
her to count, see above, p. 58; and with regard to her understanding of
words, p. 126.

[311] “If there once existed creatures above the apes and below man,
who were extirpated by primitive man as his especial rivals in the
struggle for existence, or became extinct in any other way, there is
no difficulty in supposing them to have possessed forms of speech,
more rudimentary and imperfect than ours” (Professor Whitney, Art.
_Philology_, _Ency. Brit._, vol. xviii., p. 769).

[312] Houzeau gives a very curious account of his observations on this
subject in his _Facultés Mentales des Animaux_, tom. ii., p. 348.

[313] _Descent of Man_, p. 87.

[314] _Descent of Man_, p. 87.

[315] This term is used by Haeckel as synonymous with
_Pithecanthropoi_, or the ape-like men, who are supposed to have
immediately preceded _Homo sapiens_ (_History of Evolution_, English
trans., vol. ii., p. 293). In the next instalment of work I will
consider what has to be said in favour of this view from the side of
my anthropology. Meanwhile, it is sufficient to bear in mind that, as
previously stated, great as is the psychological difference introduced
by the faculty of speech, for the attainment of this faculty anatomical
changes so minute as to be imperceptible were all that seem to have
been required. “The argument, that because there is an immense
difference between a man’s intelligence and an ape’s, therefore there
must be an equally immense difference between their brains, appears
to me to be about as well based as the reasoning by which one should
endeavour to prove that, because there is a ‘great gulf’ between a
watch that keeps accurate time and another that will not go at all,
there is therefore a great structural hiatus between the two watches.
A hair in the balance-wheel, a little rust on a pinion, a bend in a
tooth of the escapement, a something so slight that only the practised
eye of the watchmaker can discover it, may be the source of all the
difference. And believing, as I do, with Cuvier, that the possession of
articulate speech is the grand distinctive character of man (whether
it be absolutely peculiar to him or not), I find it very easy to
comprehend, that some equally inconspicuous structural difference
may have been the primary cause of the immeasurable and practically
infinite divergence of the human from the simian stirps” (Huxley,
_Man’s Place in Nature_, p. 103).

[316] Here I will ask the reader to bear in mind the considerations
above adduced from Geiger, as to the encouragement which must have
been given to a semiotic use of vocal sounds by habitual attention
being given to the movements of the mouth in significant grimace—such
attention being naturally bestowed in larger measure by an intelligent
ape-like creature which was accustomed to depend chiefly on its sense
of sight, than it would be by any of the existing quadrumana.

[317] For sign-making among the social insects, see above, pp. 88-95.

[318] Here, be it observed, the element of truth which belongs to the
first of the three hypotheses that we are considering comes in. Compare
foot-note on page 364: _Homo alalus_, though not yet a conceptual
thinker, is nevertheless in possession of a higher receptual life than
has ever been attained by a brute, and is correspondingly more capable
of utilizing as signs interjectional or other sounds which emanate from
the “purely physiological grounds” of his own organization.

[319] See Preyer, _loc. cit._, for a detailed account of the order in
which the consonants are developed in the growing child. Also Professor
Holden, on the _Vocabularies of Children_, in _Proc. Amer. Philolo.
Ass._, 1877. There can be no doubt that vowel sounds must have been
of early origin in the race; but in what order the consonants may
have followed is much more doubtful. For different races now exhibit
great differences with regard to the use—and even to the capability
of using—consonantal sounds; the Chinese, for instance, changing _r_
into _l_, while the Japanese change _l_ into _r_. And, of course, the
whole science of comparative philology may be said to be based upon a
study of the laws of “phonetic change.” But it is obviously a matter
of no importance in what particular order the different articulate
sounds were first evolved. According to Prince Lucien Bonaparte, who
has investigated the matter with much care, the total number of these
sounds that can be possibly made by the human organs of vocalization
is 385. See, also, Ellis, on _Early English Pronunciation_; and, for
the limitation of consonants in various languages of existing races,
Hovelaque, _Science of Language_, English trans., pp. 49, 61, 81.

[320] “When we remember the inarticulate clicks which still form part
of the Bushman’s language, it would seem as if no line of division
could be drawn between man and beast, even when language is made the
test” (Sayce, _Introduction, &c._, ii., p. 302).

[321] _Ursprung der Sprache_, s. 52.

[322] _Introduction, &c._, ii., 302: by “thought” of course he means
what I mean by recepts.

[323] Here also compare the first of the three hypotheses, the
important elements of truth in which are, as I have already more than
once observed, to be considered as adopted by Mr. Darwin’s hypothesis,
and therefore also by the present one.

[324] The song of the gibbon has already been alluded to in a
quotation from Darwin. I may here add that the chimpanzee “Sally” not
unfrequently executes an extraordinary performance of an analogous
kind. The song, however, is by no means so “musical.” It is sung
without any regard to notation, in a series of rapidly succeeding howls
and screams—very loud, and accompanied by a drumming of the legs upon
the ground. She will only thus “break forth into singing” after more or
less sustained excitement by her keeper; but more often than not she
refuses to be provoked by any amount of endeavour on his part.

[325] Compare quotations from the German philologists in support of the
first hypothesis, pp. 361, 362.

[326] See pp. 288-290.

[327] _Welt als Entwickelung der Geists_, s. 255. This book, however,
was not published until 1874—_i.e._ some years after the _Descent of
Man_.

[328] This is likewise the view that was ably supported by Geiger on
philological grounds, _Ursprung der Sprache_, 1869; and by Haeckel on
grounds of general reasoning, _History of Creation_, English trans.,
1876.

[329] “How many of the roots of language were formed in this way it is
impossible to say; but when we consider that there is no modern word
which we can derive from such cries as the sailor makes when he hauls a
rope, or the groom when he cleans a horse, it does not seem likely that
they can have been very numerous” (Sayce, _Introduction, &c._, i., p.
110).

[330] With regard to the erect attitude, we must remember that,
although the chimpanzee and orang never adopt it, the only other
kinds of anthropoid apes—namely, gorilla and gibbon—frequently do
so when progressing on level surfaces. In the case of the gorilla,
indeed, although the fore-limbs quit the ground and the locomotion
thus becomes bipedal, the body is never fully straightened up; but in
the case of the gibbon the erect attitude may be said to be complete
when the animal is walking. (Huxley, _Man’s Place in Nature_, pp.
36-49). With regard to the selection and use of stones as tools,
Commander Alfred Carpenter, R.N., thus describes the _modus operandi_
of monkeys inhabiting islands off S. Burmah:—“The rocks at low-water
are covered with oysters. The monkeys select stones of the best shape
for their purpose from shingle of the beach, and carry them to the
low-water mark, where the oysters live, which may be as far as eighty
yards from the beach. This monkey has chosen the easiest way to open
the rock-oyster, namely, to dislocate the valves by a blow on the base
of the upper one, and to break the shell over the attaching muscle”
(_Nature_, vol. xxxvi., p. 53. In connection with this subject see also
_Animal Intelligence_, p. 481).

[331] See above, p. 220.

[332] See pp. 220-222.

[333] See pp. 179-181.

[334] See above, pp. 300, 301.

[335] Whitney.

[336] Sayce.

[337] Farrar.

[338] Garnett.

[339] Sayce.

[340] Max Müller.

[341] See especially _Science of Thought_, chaps, ii. and iv. The
following quotations may suffice to justify this statement. “If
once a genus has been rightly recognized as such, it seems to me
self-contradictory to admit that it could ever give rise to another
genus.... Once a sheep always a sheep, once an ape always an ape, once
a man always a man.... What seems to me simply irrational is to look
for a fossil ape as the father of a fossil man.... Why should it be
the settled or ready-made Pithecanthropus who became the father of
the first man, though everywhere else in nature what has once become
settled remains settled, or, if it varies, it varies within definite
limits only? (pp. 212-215).... If the germ of a man never develops into
an ape, nor the germ of an ape into a man, why should the full-grown
ape have developed into a man? (p. 117).... Let us now see what Darwin
himself has to say in support of his opinion that man does not date
from the same period which marks the beginning of organic life on
earth—that he has not an ancestor of his own, like the other great
families of living beings, but that he had to wait till the mammals had
reached a high degree of development, and that he then stepped into the
world as the young or as the child of an ape” (p. 160), &c., &c. So far
as can be gathered from these, and other statements to the same effect,
it does not appear that Professor Max Müller can ever have quite
understood the theory of evolution, even in its application to plants
and animals. For these are not criticisms upon that theory: they are
failures to appreciate in what it is that the theory itself consists.

[342] _Ursprung der Sprache_, s. 84.

[343] _Ursprung der Sprache_, s. 119.

[344] It would be no answer to say that by “names” he means only
signs of ideas which present a conceptual value—or, in other words,
that he would refuse to recognize as a name what I have called a
denotative sign. For the question here is not one of terminology, but
of psychology. I care not by what terms we designate these different
sorts of signs; the question is whether or not they differ from one
another in kind. If the term “name” is expressly reserved for signs
of conceptual origin, it would be no argument, upon the basis of this
definition, to say that there cannot be names without concepts; for, in
terms of the definition, this would merely be to enunciate a truism: it
would be merely to say that without concepts there can be no concepts,
nor, _à fortiori_, the signs of them. In short, the issue is by no
means one as to a definition of terms; it is the plain question whether
or not a non-conceptual sign is the precursor of a conceptual one. And
this is the question which I cannot find that Max Müller has adequately
faced.

[345] _Ursprung der Sprache_, s. 91. The exact words are, “Die Sprache
hat die Vernunft erschaffen: vor ihr war der Mensch vernunftlos.” It is
needless to observe that the word which I have rendered by its English
equivalent “Reason” is here used in the sense of conceptual thought.

[346] Wundt, _Vorlesungen, &c._, ii. 282.








End of Project Gutenberg's Mental Evolution in Man, by George John Romanes