38o 



NATURE 



'\_Fcb. 2 1, 1 1 



an anteceient presumption that the cjmmon sense of which it is 

 the expression is in the right, and that any ingeniously-c in- 

 structed armament of scep"icism is in the wronj. We miy there- 

 fire approach Mr. Lliyd Morgan's ar^unent with the antecedent 

 ))resumption that there mu^t be somethin.; wrong about it some- 

 where ; and I do not think that it requires much refl;ction to 

 s;e where the error lies. 



According to thearguiientas stated by my critic, there is a true 

 s:ience of human psychology, because, althiugh my linowledge 

 of another human mind is no leis ejective than is my liaowled^e 

 of a do:j's mind, yet "by means of language human beings can 

 communicate to each other the results which each has obtained, 

 and each human being is able to submit tlicse rdudts to the test of 

 subjective verificalioa." But how, let us ask, in its last analysis 

 is this verifioa'ion obtained? By language, no doubt ; but what 

 in its last analysis is language? As spoken by my neighbour, it 

 is for me nothing more than my own interpretation of a meaning 

 presented by the observable activities of an organism. Therefore, 

 if on such a basis I am entitled to affirm that such interpretations 

 as I make are of the nature of " subjective verifications " of con- 

 clusions drawn from the introspective observation of my own 

 mind, why am I not entitled to a similar view when the eject of 

 my cmtemplation is the mind of a dog? The dog cannot speak, 

 but he can display other activities which, so far as they go, are 

 quite as valid as a basis on which to construct my " subjective 

 verification " as are the activities manifested in language. Of 

 course language is able to convey immeasurably more information 

 touching the ejective mind than can be conveyed by any other 

 kind of activity ; but this fact is merely due to the further fact 

 that language is a system of activities expressly designed for this 

 very purpose. The higher v.ilue of language in this respect is 

 therefore nothing more than an expression of the higher develop- 

 ment of intelligence, which enables the mind to perceive the 

 desirability of devising a system o'f bodily activities expressly 

 designed to serve as the vehicle of communication between sub- 

 ject and eject, — as is proved by the fact that any system of 

 bodily activities which may be agreed upon (such as gesture, lip- 

 reading, writing, &c.) are alike available for this purpose. Lan- 

 guage, then, of any kind is merely a conventional system of bodily 

 activities which, because znten ied to convey information from 

 mind to mind, we call signs. But now, the element of intention 

 on the part of my neighbour is in no wise essential to my ejective 

 interpretation of his bodily activities, or to what Mr. Morgan 

 calls ray sulijective verification of them, The involuntary groan 

 of pain, the p.allor of fear, and a thousand other unintended 

 " expressions of the emotions," as well as a thousand other un- 

 intended expressions of thought (e.g. the act of pocket-picking 

 under the eye of an unseen detective), are, as it is ])roverbially 

 said, " more eloquent than words." 



I submit, therefore, that, although a dog cannot give us any 

 large measure of ejective information intentionally, or by pur- 

 posive signs (he does, however, give us some even of this), we 

 have still abundant material furnished by his other bodily 

 activities for constructing our ejective inferences. For example, 

 the dog gives very much the same indications of pain under the 

 whip that a boy gives under the cane ; therefore the gamekeeper 

 has no more doubt that he is hurting the dog than the school- 

 master has that he is hurting the boy — uor would the school- 

 master be more .satisfied on this point by asking the boy whether 

 the cane did hurt. 



If I have been followed thus far, I should be inclined to go 

 still further, and to say that in my opinion the " unpremeditated 

 art " of natural movements (whether in men or animals) is a 

 surer b.asis on which to build ejective conclusions than is the 

 more indirect information supplied by intentional gesture or 

 language, so far as the low or simple intelligence to which animals 

 attain is concerned. Poets and moralists are fond of insisting 

 upon this point as regards young children, where the level of 

 intelligence may be even considerably .above that of the most 

 intelligent animal. The immense service of langu.age in ejective 

 analysis is rendered in those higher and more complex regions of 

 intellectual life to which man- alone attains. Still, I doubt not 

 that if animals could speak, so that we could interrogate them 

 as to their mental operations, we should obtain a great deal of 

 supplementary information ; only of course this suppo-ition is 

 an impossible one, seeing that, if an animal could speak, its 

 intelligence would no longer be "animal intelligence." 



On the whole, then, as concerns the question whether there is 

 a science of comparative psychology, I should say that there cer- 

 tainly is such a science, in the same sense as there is a science of 



human psychology. For it seems to me, in view of the above 

 considerations, that the argument adduced by Mr. Lloyd Morgan 

 against the former is quite as applicable against the latter. In 

 both cases alike our ejective inferences can only be founded on 

 the observable activities of organisms, and if it is true that of 

 these observable activities language affords an exceptionally 

 meaning class, it is also true that where language is absent the 

 mental processes which stand to be ejectively analysed are of a 

 comparatively simple nature. I therefore see no reason to recede 

 from the position which I have taken up in the works to which 

 Mr. Lloyd Morgan refers, where I observe with reference to the 

 peculiar standing of psychology (both human and comparative) 

 among the sciences in the matter which we have been consider- 

 ing—" although the evidence derived from ejects is practically 

 regarded as good in the case of mental organisations inferred to 

 be closely analogous to our own, this evidence clearly ceases to 

 be trustworthy in the ratio in which the analogy fails ; so that 

 when we come to the case of very low animals — where the 

 analogy is least — we feel uncertain whether or not to ascribe to 

 them any ejective existence " (" Mental Evolution in Animals," 

 p. 22). And again, with reference to such objections as that of 

 Mr. Morgan — "Scepticism of this kind is logically bound to 

 deny evidence of mind, not only in the case of the lower ani- 

 mals, but also in that of the higher, and even in that of men 

 other than the sceptic himself. . . . This is evident because, as 

 I have already observed, the only evidence we can have of 

 ejective mind is that which is furnished by objective activities ; 

 and, as the subjective mind can never become assimilated 

 with the ejective, so as to learn by direct feeling the mental 

 processes which there accompany the objective activities, it is 

 clearly impossible to satisfy any one who chooses to doubt the 

 validity of inference, that in any case, other than his own, men- 

 tal processes ever do accompany objective activities" ("Ani- 

 mal Intelligence," p. 16). And, by parity of reasoning, the 

 same argument may be used against Mr. Morgan's sceptical 

 objection to comparative psychology as a science. In whatever 

 measure he is on principle a sceptic touching the inferences which 

 this science is able to draw as to the exi-.tence and nature of 

 animal psychology, in that measure I think he ought in con- 

 sistency also to be a sceptic with reference to the same points in 

 the science of human psychology. 



Coming now to Mr. Morgan's strictures on my psychological 

 definition of instinct, I understand that they are made, not with 

 reference to any defect in my definition as a psychological 

 definition, but with reference to the possibility of any such 

 definition whatever. In his view there can, from the nature of 

 the case, be no psychological definition of instinct ; there can 

 only be a physiological definition of the cerebral processes which 

 are concerned in actions termed instinctive. Here, then, is a 

 broad issue, although it only constitutes a part of the still broader 

 one which we have just been considering. 



I may say first of all that, if we want a physiological definition 

 of instinct, I do not think that the one which is furnished by 

 Mr. Lloyd Morgan is valid. This definition is that reflex actions 

 are due to a general type of nervous organisation, instinctive 

 actions to a specific type, and intelligent actions to an individual 

 nervous organisation. Now, this threefold definition presents 

 none of that "definiteness of application" which Mr. Morgan 

 implies, nor does it tend, as he supposes, to add any " clearness 

 to our ideas concerning the things of which we speak." For it 

 is open to the fatal objection of arbitrarily classifying as in- 

 stinctive many actions which are now universally regarded as 

 reflex ; while, conversely, a still greater number of actions now 

 universally regarded as instinctive \\ ould, under this definition, 

 become classified as reflex. That is to say, there are, on the one 

 hand, many reflex actions which we should all feel it absurd to 

 call insiinctive, and which are nevertheless manifested by only 

 one species (in our own organisations, for example, we may 

 mention the "patellar reflex," and the convulsions produced by 

 tickling the soles of the feet) ; and, on the other hand, there 

 is a very much greater number of instinctive actions which we 

 should all feel it absurd to call reflex, but which are nevertheless 

 manifested by many species of a genus, others by many genera 

 of an order, and so on, until in such cases as those of nidifica- 

 tion, incubation, &c., we arrive at instincts general to a whole 

 class. The truth, therefore, is that a zoological classification, 

 being made with reference to the whole organisations of animals, 

 has no such special application to the refined structure of their 

 nervous systems (which, indeed, we can only appreciate by its 

 effects on conr!ucl) as would be required for the groundwork of 



