Open main menu

Popular Science Monthly/Volume 40/February 1892/Is Man the Only Reasoner?


THE "whirligig of time" may he said to be bringing to the much-neglected brutes an ample revenge. The first naive view of the animal mind entertained by the savage and the child is a respectful one, and may perhaps be roughly summed up in the formula in which a little boy once set forth his estimate of equine intelligence: "All horses know some things that people don't know, and some horses know more things than a great many people." But this pristine unsophisticated view of the animal world, though its survival may be traced in mythology and religious custom, has long since been scouted by philosophers. Thinkers, from Plato downward, have, not unnaturally perhaps, regarded the faculty of rational thought, which they themselves exhibited in the highest degree, as the distinguishing prerogative of man. The Christian religion, too, with its doctrine of immortality for man and for man alone, has confirmed the tendency to put the animal mind as far below the human as possible. And so we find Descartes setting forth the hypothesis that animals are unthinking automata.

Not forever, however, was the animal world to suffer this indignity at the hands of man. Thinkers themselves prepared the way for a rapprochement between the two. More particularly the English philosophers from Locke onward, together with their French followers, pursuing their modest task of tracing back our most abstract ideas to impressions of sense, may be said by a sort of leveling-down process to have favored the idea of a mental kinship between man and brute. This work of the philosophers has been supplemented by the leveling-up work of the modern biologist. There is not the least doubt that the wide and accurate observation of animal habits by the naturalists of the last century has tended to raise very greatly our estimate of their mental powers. So that it would seem as if in the estimation of animal intelligence, scientific knowledge is coming round to the opinion of the vulgar, and as if "the conviction which forces itself upon the stupid and the ignorant, is fortified by the reasonings of the intelligent, and has its foundation, deepened by every increase of knowledge."[1]

Definiteness has been given to the question of the nature of animal intelligence by the new doctrine of evolution. If man is descended from some lower organic form, we ought to be able to make out not merely a physical but a psychical kinship between him and the lower creation; and the more favorable estimate of the animal mind taken by the modern savant is of great assistance here. Mr. Darwin has, indeed, shown in his valuable contributions to the subject, that the rude germ of all the more characteristic features of the human mind may be discovered in animals. At the same time, Mr. Darwin's investigations in this direction amounted only to a beginning. The crux of the evolutionist, the tracing of the continuity of crude, formless animal inference, up to the highest structural developments of logical or conceptual thought, still remained. And so, the most powerful attack on the theory of man's descent has come from the philosopher, the logician, and the metaphysical philologist, who have combined to urge the old argument that conceptual thought indissolubly bound up with language sets an impassable barrier between man and brute.

Mr. Darwin's unfinished work has now been taken up by one who adds to the biological knowledge of the expert a considerable acquaintance with psychology. In his previous volume, Mental Evolution in Animals, Dr. Romanes took a careful psychological survey of the animal world for the purpose of tracing out the successive grades of its mental life. In his recent volume. Mental Evolution in Man (Origin of Human Faculty), he essays to trace forward this general movement of mental evolution to the point where logical reasoning or "conceptual thought" may be distinctly seen to emerge. That is to say, he adroitly seeks to leap the "impassable" barrier by merely denying its existence. Human reasoning and animal inference are not two widely dissimilar modes of intellection. The one is merely a more complex expansion of the other. If you start either at the human or the animal bank you can pass to the opposite one by a series of steppingstones. In other words, the higher human product can be seen to have been evolved out of the lower by a continuous process of growth.

Dr. Romanes's present contribution to the theory of evolution is thus emphatically the construction of hypothetical steppingstones for the purpose of passing smoothly from the territory of animal to that of human reasoning. In order to this, he has on the one hand to follow up animal intellection to its most noteworthy achievements, and on the other hand to trace the process of human intellection down to its crudest forms in the individual and in the race.

As it is obviously language which marks off human thought from its analogue in the animal world, our author is naturally concerned to limit the function of language. While allowing as a matter of course that the "conceptual thought" of the logician involves language as its proper instrument or vehicle, he urges that there is a good deal of rudimentary generalizing prior to, and therefore independent of, language. To establish this a careful examination of the higher processes of animal "ideation" has to be carried out. In doing this Dr. Romanes introduces a number of psychological distinctions of a somewhat technical kind. Of these the most important perhaps is that between the time-honored concept of the logician and the recept. This last corresponds to Mr. Galton's generic image or the common image (Gemeinbild) of the German psychologists. It is an image formed out of a number of slightly dissimilar percepts corresponding to different members of a narrow concrete class, such as dog or water. According to our author, animal reasoning remains on the plane of recepts. It is carried on by pictorial representations. At the same time it involves a process of classification or generalizing. A diving-bird must be supposed to have a generalized idea (recept) of water, a dog a generalized idea of man, and so forth. Nay, more, this receptual ideation enables the animal to reach "unperceived abstractions," as the idea of the quality of hollowness in the ground, and even "generic ideas of principles" as when the writer's own monkey having discovered the way to take the handle out of the hearth-brush by unscrewing it, proceeded to apply the principle of the screw to the fire-irons, bell-handle, etc.

The author's whole account of this receptual ideation or the logic of recepts is interesting and persuasive. He has, it must be owned, clearly made out the existence of a very creditable power among animals of carrying out processes analogous to our own reasonings without any aid from language. Yet a doubt may be entertained whether the author has really got at the bottom of these mental feats. The whole account of the recept is a little unsatisfactory, owing to the circumstance that the writer does not make it quite clear in what sense it involves generalization. He writes in some places as if the fact of the generic image having been formed out of a number of percepts corresponding to different members of a class, e. g., different sheets of water seen by the diving-bird, gives it a general representative character. But this, as indeed Dr. Romanes himself appears to recognize in other places, is by no means a necessary consequence. A generic image may form itself more readily than a particular one, just because the animal is unable to note differences sufficiently to distinguish one sheet of water or one man from another. A baby's application of the common epithet "dada" to all bearded persons suggests not that it is carrying out any process of conscious generalization, but rather that it is failing to discriminate where there are striking and interesting features of similarity. It would seem as if an idea only acquires a properly general function after certain higher intellectual processes have been carried out. These may be roughly described as the active manipulation of percepts and images, by analytical resolution of these into their constituent features, and a due relating or ordering of these elements. Only in this way does it appear possible to reach a rudimentary form of a properly general notion; that is to say, an idea which is consciously apprehended as representing common features among a number of distinct objects. Mere superposition of images may result in a new typical image; but the mind in which such an image forms itself can not know this to be generic or general till these processes which underlie active thought have been carried out. Now, we ourselves carry out these operations of resolving into elements and recombining these elements (analysis and synthesis) largely by the help of class-symbols or general names, which come to be general symbols just because we make use of them for the purpose of noting down and keeping distinct the results of our successive comparisons and analyses. And the really pressing question for the evolutional psychologist is. How does this manipulation of the mind's imagery get carried out where the serviceable instrument of language is absent? That it does get carried out to some extent may be readily allowed. A sagacious and well-bred collie, who combines with a judicious preference for his owner a certain mild complacency toward mankind at large (with some possible exceptions), may be rightly regarded as having attained to a rudimentary consciousness of the distinction between the general and the particular, the "class" and its constituent members. But how this has been attained Dr. Romanes's account of receptual ideation hardly helps us to understand.

The recept or generic image is the first of the psychological stepping-stones leading across the unfordable Rubicon, and it is also the principal stepping-stone. Should this prove to be unstable, the transit would certainly become exceedingly doubtful.

From the recept we pass to the concept, which, according to our author, is in its simplest form a named recept. The addition of the name or sign is thus the differentiating character of the concept. We may have generic images, but no concepts apart from names or other signs.

In order to understand how the concept is marked off from the recept, we must accordingly inquire into fhe psychological conditions and concomitants of the naming process. And this our author does at some length. He gives us a full and detailed account of names and of signs in general, distinguishing different grades of sign-making from the merely indicative pointing or other gesture up to the bestowal of a general symbol with a consciousness of its significance as connoting certain common qualities. Into much of this it is not needful for us to follow Dr. Romanes, but brief reference may be made to one or two points of special importance as bearing on the evolution of the higher conceptual thought. One of the most curious features of Dr. Romanes's theory of concepts and naming is the proposition that the name is bestowed on the idea, and has for its psychological condition an act of introspection. He tells us that before we can bestow a name on a recept we must be able to set this recept before our mind as an object of our own thought. Or, to express the truth in the author's own words, self-consciousness is the necessary presupposition of naming and so of conceptual thought. Before I can name an idea I must reflect on the idea as mine, and before I can judge in the logical sense, I must realize the truth of the proposition as such, that is presumably as truth for me, so that self-consciousness would seem to come in necessarily at all stages of conceptual thought.

This doctrine seems by no means as clear and convincing as the author supposes. He is, as he clearly tells us, confining himself to the psychological treatment of his subject. This being so, it may fairly be urged that in making an act of subjective introspection an essential factor in the process of naming he is psychologically wrong. Is a child when inventing a name for his toy horse or doll reflecting on his idea as his and naming this idea? Is he not rather thinking wholly about the object, and is not the name given to this external object and not to the idea in the namer's mind at all?[2] No doubt the completed process of logical reflection on names and propositions brings in the subjective element—that is to say the mind's consciousness of its ideas and judgments as representations of the realities thought about. But this reference to self, this act of introspection, so far from being involved in every act of conceptual thought, is directly excluded from it.

This brings one to the next point. In naming things the mind is busily occupied, not with itself and its ideas, but with the "not-self," the qualities and relations of the things perceived or represented. And this suggests first of all that naming, properly so called, only begins when things come to be apprehended as such, that is to say, as wholes or unities. And here the question occurs whether an animal, say a dog, that is just coming on to understand a name or two, as that of the baby of the house, can be said to have an organized percept precisely analogous to our own percepts? Dr. Romanes does not raise the question, but, in view of the light thrown by modern psychology on the complexity of the process of perception, it might not have been redundant. But waiving this point as possibly smacking of the frivolous, we have to ask whether an animal at the stage of mental development at which it appears to begin to understand names, and even to make use of them, is capable of carrying out the processes that go along with, and in fact constitute, naming in its true and complete sense. These processes have already been referred to in connection with the subject of general ideas. To name an object appears to mean to apprehend that object as a complex of qualities, to make mental separation of these, and so to relate it to other objects both by way of similarity (classification) and dissimilarity (individuation). To use a name intelligently at all would seem to imply that these processes have been carried out in a rough fashion at least. This being so, we must be prepared when we endow an animal with the power of naming, whether under the form of understanding or that of using names, to say that it is carrying out in a rudimentary way at least these thought-processes. How, it may be asked, does Dr. Romanes deal with this point?

The answer to this question will be found by turning to new distinctions or "stepping-stones" in the movement of thought-evolution. Our author attaches importance to the distinction between higher and lower forms of the concept. Not only is there the generic image to carry us on smoothly from image to concept, but within the limits of the concept itself there are higher and lower forms. Since, according to our author, a concept is any named idea, a proper understanding of these conceptual grades can only be obtained by a glance at his scheme of names.

There are, according to Dr. Romanes, four stadia in the evolution of the complete logical sign or general name. Of these the first is (a) the indicative sign—that is, a significant tone or gesture intentionally expressive of a mental state, as the characteristic tones by which animals express their emotions. These are not names at all. Next to these in the order of evolution come (b) denotative signs. These, whether used by children or animals, e. g., talking birds, simply mark "particular objects, qualities, and actions." They are learned by association, and are not consciously employed as names. By the use of such a sign the talking bird merely fixes 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. Next in order (c) follow connotative signs which involve the "classificatory attribution of qualities to objects." This attribution of qualities may be effected either by a receptual or a properly conceptual mode of ideation. For example, a parrot had come to use a barking sound when a particular dog appeared on the scene. This sign was afterward extended to other dogs, showing that there was a certain recognition of the common qualities or attributes of the dog. Similarly when the writer's own child, among its first words, used the term star for all brightly shining objects. Here again there was perception of likeness, but no setting the term before its mind as an object of thought. Lastly (d), we have the denominative sign which means a connotative sign consciously bestowed as such with a full conceptual appreciation of its office and purpose as a name.

In this scheme Dr. Romanes evidently recognizes the point we are now dealing with, viz., the implication of a true thought-process in the proper use of a name. He seems to be trying to dispense with this as long as possible, with the view of securing a number of intermediate stepping-stones. Can he be said to have succeeded? Does this hierarchy of signs with its parallel scale of ideation carry us up to logical thought? Is it even intelligible? Let us briefly examine it.

To begin with, it staggers one not a little to find that long before the "classificatory attribution of qualities" is possible, the animal somehow manages to mark "particular qualities," whatever these may mean. How, one asks, can a sign be appended to a quality without becoming a "connotative sign"—that is, attributing a quality to a thing? But let us pass to the really important point, viz., the alleged power of the animal, e. g., the talking bird, to extend a sign to different members of a class, and so to attribute common qualities or resemblances to these, while it is unable to form a concept in the full sense. This extension, we are told, takes place in the case of the sign-using bird by receptual ideation. And here the critic may as well confess himself fairly beaten. On the one hand, Dr. Romanes tells us that such a named recept is a concept (lower concept), and, moreover, that the sign employed is a connotative sign; on the other hand, he hastens to assure us that it is not a name, and therefore presumably not a concept, in the rigorous or perfect sense, since the sign is not consciously employed as a sign. Here we seem to have a steppingstone which it is impossible to define, a sort of tertium quid between the image and the concept which is at once neither and both; Surely if a sound is used for the purpose of marking resemblances and attributing qualities, it is a genuine name, and the mental process underlying it is a germ of true conceptual thought. To say that the parrot attributes qualities, and attributes them in a "classificatory" way too, seems indeed to mean that the bird has got a considerable way along the conceptual path, and is fairly within sight of our distinctions of thing and quality, individual and class. Why logical reflection on this name as such should be needed to raise such a performance to the dignity of a true conceptual act, one is at a loss to understand. And, indeed, the author himself appears to recognize all this in a dim way at least, when he adds that the connotative sign may be the accompaniment not only of receptual but of truly conceptual ideation. At the same time this addition may very well complete the reader's perplexity, for it appears to render the next stage of evolution, the denominative sign, unnecessary.

Altogether the author's account of sign-accompanied ideation is not quite satisfactory. To begin with, one misses an adequate psychological treatment of signs in general, their nature and function in our mental processes, such as M. Taine has given us in the beginning of his work On Intelligence. Then our author has left us very much in the dark as to what it is that the sign does for the intellective process, when it begins to be used. On the one hand, since we are told that the mere addition of a name transforms the generic image into a "concept," we naturally expect the function of the sign to be a large and important one. On the other hand, we gather that signs can be used at the level of receptual ideation, where, consequently, true conceptual thought is wholly excluded.

This confusion seems to have its main source in the curious theory that while an idea may be general, it can not become a true concept till it is introspectively regarded as our idea; and its counterpart, that while a sign may be a true sign and even subserve the attribution of qualities to objects, it can not grow into the full stature of a name till it is reflected on as a name. By this doctrine Dr. Romanes seems unwittingly to have substituted the logical for the psychological definition of the concept, and so to have put the latter higher up in the evolutional scale than it ought to be. To this it must be added that the author appears to have been overanxious, with the view of making the transit smooth, to multiply distinctions. Such intermediate forms as Dr. Romanes here attempts to interpolate in the process of intellectual development can not in truth do away with the broad distinctions which psychologists are in the habit of drawing. Thus the recept only appears to connect the image and the concept just because it tries to be both at the same time. So the lower stadium of the sign only gives an appearance of bridging over the interval between signless ideation and sign-aided thought, just because it aims at once at being something less than a true sign, and this true sign itself.

If our criticisms are just. Dr. Romanes can not be said to have succeeded in his main object, viz., the obliteration of all qualitative difference between human and animal intellection by the interposition of psychological links which can be seen to have the essential characters of both. And here one is naturally led to ask whether the author is after all on the right track. For he is a master of his facts and shows considerable power in the marshaling of his arguments, and, as even a hasty perusal of the volume can show anybody, he has here concentrated his force in a severe and sustained effort. Where he has failed it is conjecturable that others may fail also. And so it behooves us to see whether he has approached the problem in the right way, or, at least, in the only possible way.

The introduction of all this technical mechanism of receptual ideation, lower concepts, and the rest, has for its avowed object the avoidance of all introduction of qualitative change in the process of intellectual evolution. Dr. Romanes tells us plainly at the outset that he is going to establish identity of kind between the animal and the human type of intellection. And, no doubt, if it were possible to do this in the way here attempted—that is to say, by interposing transitional forms which virtually efface all qualitative unlikeness—it would be a great advantage to the evolutionist. But it may be said that it is not the only way of satisfying the requirements of the evolution hypothesis. Dr. Romanes pertinently remarks, in meeting a priori objections to the derivation of human from animal intellection, that in the life of the human individual we actually have a series of transitions from animal to human psychosis. Now, a glance at the intellectual development of the individual shows us that distinct qualitative differences are introduced. Not to speak of the obvious fact that every new sensation effects a qualitative addition to the infant's mental life, there is the more important fact that the first image of the absent mother or nurse introduces a new sphere of mental activity. The child that dreams and imagines is already a different being from the infant that merely touches and sees. Similarly it may be said that the first conscious process of breaking up its sense-presentations, the first distinct apprehension of relations, is epoch-making just because it marks the oncoming of a new mode of mental activity, a qualitative extension of its conscious life.

To say this, however, is not to say that the process of development is wanting in continuity. For, first of all, these higher forms of activity introduce themselves in the most gradual way, and only slowly disentangle themselves from the lower forms which constitute their matrix. Thus the image little by little lifts itself butterfly-like out of its chrysalis, the percept. Similarly, what we call thinking, with its conscious comparing and relating of the products of sense-perception, emerges in the most gradual way out of lower forms of psychosis.

But this is not all, or the main thing. While the higher and lower forms of intellection undoubtedly exhibit qualitative differences, it may be possible to transcend these differences by going deeper, and detecting the veritable elements of the intellective process. This deeper analysis is emphatically the work of modern psychology, and, as every reader of Mr. Herbert Spencer knows, is of vast assistance to the evolutionist in following the psychical process from its rudest conceivable form in the lower grades of animal life up to the highest achievements of human thought. The luminous idea that all intelligence is at bottom a combination of two elementary processes, differentiation and integration, seems to lift one at once high above the perplexities with which our author so laboriously deals. It enables us to say that animal intelligence, just because it is intelligence, must be identical in substance with our own. The qualitative differences between perception and conception, or, to take Dr. Romanes's example, "the logic of recepts" and the logic of concepts, which obstinately persist so long as we look at the process ab extra, now appear as mere results of different degrees of complexity, of unlike modes of combination of the ultimate elements; just as to the physiologist the manifold variety of color resolves itself into different modes of combination of two or three elementary sentient processes.

When once this fundamental identity of all intellective processes is clearly apprehended, the question where exactly in the evolutionist's tree the twig of thought proper, or better, perhaps, of conscious generalization, branches off, sinks to its proper place as a question of quite secondary importance. At the same time we may agree with Dr. Romanes that the point has its real historical or genealogical interest, and that he has not done amiss to devote a volume to its discussion.

The question turns mainly on the point how much the animal can do by means of pure imagining and the aid of association. Our author clearly recognizes that this will carry animals some way, and may give to their mental operations the appearance of a true generalizing process. But he has not fixed the limits of this pictorial or suggestive inference with the precision one looks for, partly, no doubt, because his whole view of the generic image as somehow involving a generalizing process tended to obscure from him the real point. One might safely, perhaps, hazard the assertion that the diving-bird can get on very well without anything like a general idea of water, a pure (generic) image being all that seems necessary. On the other hand, one is disposed, on the evidence of the facts adduced by our author, to put the beginnings of the true generalizing process pretty low down. It certainly seems to be involved in the mental life of the ants, as elicited by Sir John Lubbock's experiments, and described by Dr. Romanes (p. 94 and following). And since these particular actions plainly imply the use of signs, and apparently signs capable of indicating such abstract ideas as those of quantity, there seems no reason why we should hesitate to call ants thinkers in the sense of being able to form general notions. The same applies to the mechanical inventions of the spider, described by Mr. Larkin (p. 62). Similarly, it is difficult to deny the rudiment of "conceptual thought" to a fox who can reason on the matter of traps in the way described by Leroy (p. 56), or to a dog that was cured of his dread of imagined thunder by being shown the true cause of the disturbing noise, viz., the shooting bags of apples on to a floor (pp. 59, 60). No doubt there is a danger in straightway endowing animals with mental qualities identical with our own, when their actions resemble ours. There may, of course, be two psychological explanations of the same action. We can not, however, escape our limitations, and, if we are to deal with animal ways at all, we are bound to interpret them in terms of our own mental processes.

The hesitation of the evolutionist to attribute rudimentary thought to animals, in which Dr. Romanes evidently shares, is no doubt due to the firmly established assumption that we generalize by help of language. To the nominalist more especially it savors of rank heresy to hint that animals apparently destitute of signs may be capable of generalizing their perceptions and reaching a dim consciousness of the distinction between the universal and the particular.

But is the nominalist's assumption that language is the indispensable instrument of thought above challenge? A considerable part of Dr. Romanes's volume deals with the relations of thought to language. He gives us a fairly good summary of the results of research into the origin of language. It can not be said that these throw much light on the question. Perhaps it is unreasonable to expect that they should. Our author contends with some skill as against Prof. Max Müller that the earliest traces of human language suggest a highly pictorial and non-conceptual mode of ideation. And in his ingenious hypothetical account of the genealogy of man as the articulate reasoner our author inclines to the idea that, so far from language making the thinker, the endowment of language has to be ingrafted on a high quality of intelligence, and even then to undergo considerable development before it becomes a mechanism for conceptual thought.

The whole subject is still a dark and perplexing one, and we must refrain from dogmatizing. It may, however, be contended that the evidence on the whole supports the view that the generalizing process is up to a certain and not very high point independent of language. That is to say, an animal unassisted by any system of general signs may make a start along the path of comparing its observations, resolving them into their constituents, and separating out some of these as common qualities. Whether in these nascent operations of thought there is some substitute for our mechanism of signs, we do not know and perhaps never shall know. However this be, they remain nascent processes never rising above a certain level. The addition of some kind of sign which can be used as a mark of common features or qualities seems to be indispensable to any high degree of generalization, and to any elaborate process of reasoning. It is the want of such signs, and not the lack of the "power of abstraction," that keeps certain animals, for example the dog, from being rational animals in as complete a sense as a large number of our own species.—Nineteenth Century.


  1. Prof. Huxley, Hume, p. 104.
  2. I believe that observers of children will indorse the remark that children regard names as objective realities mysteriously bound up with the things, and in a manner necessary to them. A nameless object is, for a child, something incomplete—almost uncanny.