hands on the handle instead of employing one hand to steady the screw. With the hammer and the nails, he steadied the nail be- tween two fingers of his left hand and used the hammer with his right. Once he pounded his fingers and instantly put them in his mouth. When tested with the form board he showed total inability to place the blocks in their proper compartments. He seemed to be able to articulate the word " mama." Witmer states that he did this with great effort. His response to words was not tested scien- tifically enough to enable us to pass judgment upon his ability in this direction. When asked certain questions as: " Where is mama? " "Where is daddy?" "Where is Peter?" "I beg pardon!" "Give mama a kiss," " Give daddy a kiss," " Give mama a drink," and so on, he did the correct thing, seemingly in response to the words. No attempt was made by Witmer to substitute different words with the same intonation, etc., and no attempt was made to give the words from behind a screen, etc. (See tests on Don and on Jasper, below. ) The behavior of Peter, while described by Witmer in a grossly anthropomorphic manner, shows ns at once that we are dealing with an organism which is far more complicated than any horse at Elberfeld. Witmer states that he ought to be compared rather with the child. Unfortunately neither the behavior of the chimpanzee nor that of the child has been subjected to sufficient analysis to make comparison feasible. At least two general state- ments may be made of the chimpanzee by way of showing similarity to and difference from man. In the first place his motor development seems capable of being extended to a point where in some instances he can actually compete 314 THE LIMITS OF TRAINING IN ANIMALS with man, and in the second place his lack of language habits puts him forever below the plane of comparison with man.
Don: The analysis of the vocal efforts of Don, " the talking dog," has been undertaken by Pfungst. The analysis is of interest only by way of the light thrown upon the instinctive sounds which can be made by the dog. Don is a German setter about 8 years of age be- longing to the royal game warden at Ebers in Gardelegen. His vocabulary is supposed to contain 8 words which may be called out when food is present and the following questions given: " Was heisst du?" "Don." "Was hast du?" "Hunger." "Was willst du?" "Hahen haben." "Was ist das?" " Kuchen." "Was bittest du dir aus? " " Ruhe." Don was supposed further to answer questions by J a and Nein and to speak the word Haherland. On examination it was found that the vocabulary is given always in order beginning with Don and ending with Ruhe. If the order is varied we find him calling himself Kuchen and stating that he desires Hunger! The first vocal effort sounding like a word was supposed to have been hahen. This attracted the attention of the owner and thereupon be- gan a series of lessons in which Don was fed for correct response. Ten repetitions some weeks apart sufficed for completing this re- sponse. The word Ruhe was supposed to have been uttered first on command of the owner's daughter. It was " fixed " in the same way. It will be seen immediately that there is nothing in the behavior of Don which is different, from the standpoint of interpretation, from what we find in the parrot. Furthermore, close analysis of even the words he speaks shows that they are not true words in the me- chanical sense — not all the vowel and consonant sounds being in- cluded. Indeed, even the number of syllables is not always constant. The dog makes only one vowel sound lying between o and u. His one guttural aspirant is like the German ch and does duty both for h and h. There is also a nasal sound lying between n and ng. \\[\\en not prolonged it will pass for d as in Don. He does not really make the sounds &, d, k, I, or r. What he really sounds is \_{ch) unguo] in the word which passes for Hunger.^ Jasper: The bull terrier owned by Mr. Dixie Taylor is in many ways a remarkable animal, but not nearly so remarkable as his o\^Tle^ believes him to be. Jasper has been several times under the observa- tion of the author. There are many things in his performance which are deserving of a more careful study than we have been able to give him. Jasper's claims to consideration come from his extraordinary ability to do certain acts in response to words. Mr. Taylor thinks the dog can respond properly to several hundred words, including those indicating direction; e.g., up, down, right, left. We have not been able to agree with Mr. Taylor respecting the dog's ability to go in a given direction. In one case in a three-story house we took Jasper to the stair-landing on the second floor. Mr. Taylor remained on the first floor, completely hidden. The author was alone with the dog. The owner then said to the dog, " Go up the stairs, Jasper." Once 3 Those of us who have seen Don's performance in this country will surely agree with the above analysis of Pfungst.
THE BEHAVIOR OF JASPER 315 he went to the window and looked out (another act he is supposed to do on command), once he tried to go below, but the third time he found the stairs and started up them. Not many tests were made on this point since there were several others which we desired to bring out. The most significant act brought out on command was the following: Jasper was told to "go into the room and turn over the waste-paper basket"; or to "go into the room and bring the waste- paper basket." When the tests were made in the Hopkins Labora- tory Mr. Taylor had first gone into the room and touched the articles. We repeated the experiments upon another occasion in which Mr. Taylor had never even seen the articles. The commands were given by the owner from such a position that the dog could not see him. On command the dog would go into the other room and mount the table and look out of the window, or he would close the door. On command, " Go to the next room and bring me a paper lying on the floor," the dog would go and execute the command. Many common objects which are left lying on the floor will be brought in on com- mand. When the situation is complicated by laying several objects in a row near together, e.g., a handkerchief and a one-dollar bill, the dog does not always bring the object designated by the word. Not enough tests have been made to give the percentage of correct re- sponses when two objects are used, nor to tell the decrease in ac- curacy when three or more objects are placed side by side. Two other tests were made in the open. After reaching the street Mr. Taylor said to Jasper, " Go behind me and put your feet on the bicycle." The bicycle was about 50 feet behind Mr. 'Taylor. The dog trotted im- mediately to it and did as he was told. The command to go down the street and put his feet on an automobile (about 100 yards dis- tant) was executed with equal readiness. Many other tests of this character were made, but they were so poorly controlled that no very accurate conclusion can be drawn. For the past 18 months Miss Elizabeth Gilman has been training a cocker spaniel of pure breed (registered) along these same lines. The dog (now about 20 months of age) was purchased a few weeks after birth by the laboratory for this special purpose. He is living in the home of Miss Gilman and is a great pet. So far, although trained every day upon the execu- tion of simple commands, little success has attended Miss Oilman's efforts. This fact has increased our respect for Jasper's accomplish- ments regardless of the possibility that subsequent analysis may show that his repertoire of responses to spoken command is much more limited than we now believe it to be.* Summary. — Unsatisfactory as this chapter is from the standpoint of the strictly factual material it presents, 4 Recently the Airedale terrier Rolf, owned by the Moekel family of Mannheim, has come into prominence. This dog is supposed to possess the same mathematical ability and conversational powers as the horses of Krall. No thorougrh scientific tests have been made upon him. William Mackenzie (" Le Probleme du Chien Pensant de Mannheim," Archives de Psychologie, 1913, XIII, 312) has given a popular presentation of this dog's accomplishments. Claparfede ("X Propos du Chien de Mannheim," J. Larguier des Bancels et Ed. Claparede, Archives de Psychologie, 1913, XIII, 377) has made a few observations upon Rolf. It seems worth while to mention that Krall " discovered " the dog of Mannheim.
316 THE LIMITS OF TRAINING IN ANIMALS nevertheless it serves to show that the behavior laboratories must be prepared to admit that the sympathetic upbring- ing of animals in the home where they are thrown into con- stant contact with human beings does produce in them a certain complex type of behavior for which the laboratory concepts, as they now exist, are inadequate to supply ex- planation. It is quite obvious that the behaviorist has no right to be content until the newly born chimpanzee and the newly born child are brought up side by side and sub- jected to the same training. At definite intervals stock should be taken and the lines of divergence and of parallel growth be determined. Such studies can be adequately conducted only in a station where the chimpanzee can be bred.
BIBLIOGRAPHY Clapabede, Ed., " Les Chevaux savants d'Elberfeld," Archives d.
Psychologie, 1912, XII, 261. " Encore les Chevaux d'Elberfeld," ibid., 1913, XIII, Dexleb, H., " tjber den dermaligen Stand des Krallismus." Prag, Ettingeb, Max, " Der Streit um rechnenden Pferde." Miinchen, Natur und Kultur, 1913. Johnson, H. M., "The Talking Dog," Science, N.S., 1912, XXXV, 749. (Abstract of Pfungst's Study.) Kball, Kabl, Denkende Tiere, Beitrage zur Tierseelenkunde auf Grund eigener Versuche. Leipzig, Engelmann, 1912. Menegaux, a., " L'Education des Chevaux pensants d'Elberfeld," Bull, de VInst. Gen. Psych., 1913, XIII, 111. Pfungst, 0., Clever Bans, the Horse of Mr. Von Osten. New York, Holt, 1911. (Translation by Karl Pahn.) Sanford, E. C, "Psychic Research in the Animal Field: der Kluge Hans and the Elberfeld Horses," Am. Jour. Psych., 1914, XXV, 1. Watson, John B., " The Need of an Experimental Station for the Study of Certain Problems in Animal Behavior," Psych. Bull., Witmeb, Lightner, "A Monkey with a Mind," Psych. Clinic, 1909, CHAPTER X MAN AND BEAST Introduction. — Man and animal. — The lack of language habits dif- ferentiates brute from man. — Instinctive basis of language. — The nature of language habits. — Early predominance of language habits. — Forms of language habits. — Futility of the search for reasoning, imagery, and the like in animals. — Division of labor between human and animal behaviorists.
Introduction. — We have now considered several forms of habit, from the simple ones we find in the rat to the more complex ones we find in the chimpanzee. We saw that in the former we could, with a measure of success, reduce habit to the functioning of a relatively simple set of arcs. In the complex coordinations of the higher ani- mals we found analysis difficult, yet even in the most complicated cases we saw no reason to lose faith in our methods. It is probable that a more careful study of such acts will enable us ultimately to reduce them to their lowest terms and to state the laws of their genesis and of their combination. Stated in other words, our science does not feel the need of changing its methods and prin- ciples in passing from simple forms of behavior to more complex ones. It is probably the search for a new prin- ciple of explanation that has created so much diversity of viewpoint in the study of behavior. Strange to say, it has been the biologists and not the psychologists who have introduced this confusion, although a great many of the psychologists have followed their lead. We find among biologists generally the tendency to treat simple reflexes and habits in a perfectly objective way, but suddenly when the reactions begin to get complex we find them introducing the concept of the psychic. It is introduced apparently as a deus ex machina to account for complexi- ties in response. Even Loeb has not escaped the tendency.
318 MAN AND BEAST Certainly his earlier work makes a wide distinction between tropism and associative memory. Throughout this work we have tried to be consistent. If tropism or reflex is the ultimate unit of analysis of behavior in the lower vertebrates, it serves the same purpose in the higher or- ganisms, including man. No new principle is needed in passing from the unicellular organisms to man. As we pass from the responses of the simple organisms to the more complex ones of the higher animals we find (1) a greater number of units, and (2) more complex forms of combinations of these units. Jennings has stated that if the amoeba were the size of an elephant, we should feel pretty much about the amoeba as we do about the elephant. We doubt if he means to be taken literally in this matter. If we knew as much in proportion about the elephant as his studies have revealed about the amoeba, the disparity in behavior would seem even greater than it does at pres- ent. Jennings' point, though, is well taken, and means that the amoeba's responses are usually very complex when carefully scrutinized; probably fully as complex as the elephant's responses appear to be when not closely analyzed. (By this we do not mean to affirm that behavior becomes complex in proportion to the mass of the animal which displays it! ) Man and animal. — This leads to a consideration of the question of the dividing line between the behavior of man and that of brute. Hitherto we have not found any pecu- liarly subtle differences among the brutes themselves. We have found innumerable differences among the different phyla with respect to the presence or absence of certain types of action systems, in the range of sensitivity of re- ceptors, the delicacy of the functioning of these receptors (amount of energy of stimulus necessary to arouse them to action), in the development, range, combinations, effec- tiveness, etc., of the muscular systems. Our studies on habit and instinct have brought out such points and con- tinued study will make them increasingly apparent. It is evident that we find in the primates, at least in rudi- mentary form, all types of reactions that we find in man GENERAL CONSIDERATIONS 319 (with one exception, which we shall discuss later), such as complex forms of response to complex objects, imitation, use of tools, etc. (We say nothing of reasoning since we do not admit this as a genuine type of human behavior except as a special form of language habit.) Appeal to the nervous system yields us no better results. Man has neither the greatest absolute brain weight nor the greatest brain weight in relation to body length. While it is true that the complexity of the convolutions are greater, yet many animals have an encephalic surface which is respect- ably convoluted (especially the anthropoid apes). It is quite improbable that convolution of surface means anything per se. It is well within the bounds of possibility that had man appeared with a smooth brain three times as large as at present but having the same number of conductors and the same number of functional connections as at present, he would have shown the same superiority over his nearest competitor, the anthropoid ape, as he does now. Certainly the idea that convolutions mean anything sp-ecific in the way of differentiation would not arise under such circum- stances. Neurologists have told us faithfully enough that convolutions are simply indicative of greater cortical surface, hence of an increased number of long and short reflex arcs. Furthermore, the development of the frontal lobe has often been used as another neurological instru- ment for sounding the superiority of man over animal. It is in this region that Wundt has assumed that the apper- ception centers reside. It must be admitted again that the increased development of the frontal lobes is a genuine fact, but that any special mechanism resides in it, i.e., in the so-called silent areas, is most improbable. While it is generally admitted that complex fiber tracts connect this area with the sensory projection centers, with motor centers (and more recently admitted) with the basal ganglia, this does not furnish any proof that it exerts anything special in the way of ''control " over the projec- tion centers, as is often assumed in theories of ' ' atten- tion." Since it was the last brain tissue put on in evolu- tion, and is to be found chiefly in man, we have hastened 320 MAN AND BEAST to assign to its care all those functions in which man is thought chiefly to excel the brute. Its presence does un- questionably increase the number of neural elements and the number of synaptical connections. It is quite thinkable, though, that the same effect would have been produced if these increased connections had been placed in the parietal or occipital lobes or even in the basal ganglia. Of course to those who are wedded to the Darwinian theory and who be- lieve that simply because we have the tissue it must have a life and death significance, this argument will not appeal. But as a matter of fact we have always been skeptical of this attempt to put " something not found elsewhere " in the frontal lobes. It is extremely questionable whether (aside from its purely motor function) it has anything like the significance of the silent areas in the parietal and occipital lobes. How, then, shall we account for the notion that man is something and has something which the brute is not and has not? The feeling that a cataclysmal difference exists has been strong through all the centuries, and is as firmly fixed in the popular mind today as ever. Among scientific men the conviction that the gap is not so wide as was formerly supposed is growing; yet we find scientifically-minded men still searching for the '' missing link. ' ' ^ The feeling that a gap exists is shown all through the behavior work from studies on amoeba to those on the anthropoid ape. This feeling has manifested itself in various ways: in the search for (1) imitation, (2) rea- soning, and (3) imagery in animals. Since man was sup- posed to possess such functions and animals to be lacking in these respects, what more natural than to scrutinize closely the behavior of animals in order to discover them at least in ovo and thus to bridge the gap? We will dis- cuss (2) and (3) on p. 333. (1) Imitation we have already ^ E.g., Witmer states: "Speculation has at times arisen concern- ing the existence of a ' missing link ' between the higher apes and man. Such a ' missing link ' ought to present an anatomical form intermediate between the highest anthropoid ape and the lowest type of man. This gap (italics ours) has been partly filled by the dis- covery of the Neanderthal man and the ape-like remains found in Java."
GENERAL CONSIDERATIONS 321 treated at length on p. 277. If the analysis given there is correct, it is clear that the absence of any high degree of imitative capacity in the anthropoid ape should not be looked upon as any special form of break since apparently all that is amiss in the Simile is, on the one hand, too little complexity in their coordinations, and on the other hand, too great rigidity in habits once they are formed. The differences between man and ape in these respects are closely similar to the differences we find along the same lines between the ape and the squirrel.
In thus arguing against a fundamental difference be- tween the behavior of man and brute it must not be sup- posed that we are trying to support the continuity theory of the Darwinians. We have just seen that we have had to admit characteristic differences in the receptors and action systems in different forms. Historically, however, the existence of the ' ' break ' ' has been urged most strongly in connection with those animals which have bodily struc- tures essentially like our own. It has often been said that where there is similarity in structure, as an observed fact we find similarity in function. When we are observing the behavior of animals lower than man we are willing enough to abide by our dogma. Yet when we reach man we find a certain hesitation in applying it. Logically we must apply the principle consistently. If we find man doing something which the animal does not do, it is due to one of two things: (a) the animal does not possess the structure, or (h) he does not possess it in a highly enough developed form. If we examine man and animal from the standpoint of {a) and (&), we find that there is a differ- ence in structure which at once accounts for the popular and the scientific feeling that a break exists between man and animal. We refer to the lack of well-developed speech mechanisms in animals and to the consequent lack of language habits. So far no language habits have ever been found in animals, nor has any one succeeded in developing such habits in them.
The lack of language habits forever differentiates brute from man. — In taking the position that the chief 322 MAN AND BEAST difference between the behavior of man and animals flows out of man's ability to form language habits, we shall neglect the plainly observable difference in man's more finished use of the skeletal muscles and in the greater delicacy of his receptors, etc. It is possible, however, easily to overrate these differences. In many cases it is clearly observable that the difference between Peter's motor and sensory equipment and that of the guinea pig is probably more marked than that between Peter and the ordinary man. In taking this position we hasten to add that it is not fair to us to suppose that we believe that the putting on of simple language habits in the gibbon, e.g., would make him a fit subject to compete with a cultured Euro- pean from the standpoint of complexity of behavior. We do mean that if we could estahlish in the anthropoid even a primitive language and were then to compare him with the primitive Australian bushman, we should at once lose the feeling that there is some qualitative difference between his behavior and our own, i.e., we should lose the feeling at once that he is a brute.
Instinctive basis of language. — In the general discus- sion of habit on p. 186 we spoke of the fact that before habit can be formed the stimulus must call out a large group of reflexes. Unless the stimulus does this no habit can arise. Examination of the child shows that such reflexes are early^ called out with such abundance as to foreshadow the fact that the speech mechanism is eminently suitable to the formation of habits. The stimuli which may lead to these reflex movements in the vocal apparatus are exceedingly varied. Stimulation of any appropriate receptive surface by bright objects, oft reacted to objects, as the face of nurse and mother, sound of the human voice, loud noises, etc., will call them out. In addition to these random reflex movements, we have in the case of the throat mechanism something similar to that which we found in the general bodily mechanism, viz., there exist rather distinct concatenations of reflexes which we designate as instinctive cries and sounds. These in mammals at least are racially typical and must be con- INSTINCTIVE BASIS OF LANGUAGE 323 sidered definite instincts. These instinctive acts, executed with the vocal apparatus, have, just as is the case with instinctive acts executed with the trunk, arm, and leg muscles, more or less definite stimuli which call them forth (e.g., the mouse calling out certain responses in the cat). The animals, as a rule, possess a fairly definite and fixed group of these. This has given rise to the misguided efforts of Garner and others to show that the primates have a language. What the primates have is a much larger number of such vocal instincts than the other mammals, and a much larger number of stimuli can call them out, e.g., injury to bodily tissue calls out one group; hunger calls out a certain group; sex stimuli (mate, etc.) another; and similarly, cold, swiftly moving objects, tones, strange animals, call out others. When attachments are formed between the female and her offspring, another large group is called into action. There is no evi- dence to show in the case of mammals that these vocal instincts are modified by the sounds of other animals. We have brought up dogs, rats, guinea pigs, rhesus mon- keys, Cebus, etc., in the same room. We have shown elsewhere that although songs and call notes develop in a bird reared in isolation, they are not necessarily the ones that would have developed had the bird been reared with the adults of its own or of a different species. These throat habits may be cultivated to such an extent in birds that we may get an approximation, more or less complete, to a few such habits possessed by the human being. Such throat habits, however, are not language habits, as will be clear from our later discussion. It is pretty clear that the mutant man, when thrown off from the primate stock, sprang forth with a vocal apparatus different from that of the parent stock, and possessing abundant richness in reflexes, even far surpassing that found in the bird. It is interesting to observe too in this connection that within the narrow space occupied by the vocal apparatus we have a system of muscular mechanisms which has within it, looking at it now as a whole, the same possibilities of habit formation that we find in the remaining portion of the