Entero-ceptive system. — The organic sense is especially hard to deal with in isolation. That it plays an important part in the daily activities of most animals is apparent. There is something in animals which functions like a time sense. Nest shifts among the terns and other birds show oftentimes a marked and regular rhythm. Hunger is prob- ably the most compelling of all stimuli. Since the stimulus to hunger consists largely in the rhythmical movements of the muscles in the stomach, it would seem possible, by operative experiments, to eliminate the afferent impulses aroused by these movements. It would be exceedingly in- structive to study the formation of habits, functioning of certain instinctive acts, etc., in animals which had under- gone such operations.
Cutaneous system as a distance receptor. — Many prob- lems remain to be attacked in the cutaneous mechanisms, when they are considered as distance receptors. For some time we have felt the need of accurate information upon the sensitivity of animals to changes in temperature, i.e., the extent to which associations may be formed between cer- 42 SOME PROBLEMS ENUMERATED tain acts and certain temperature stimuli. The problem is surely not an insoluble one. Some preliminary experi- ments upon the sensitivity of squirrels to temperature changes have already been made. Further needed facts are those which relate to the animal's ability to detect differ- ences in the moisture contents of two air columns, differ- ences in pressure between two air columns, and the extent to which olfactory stimuli may be detected by the cutaneous receptors in the nasal cavities (experiments must be carried out upon'anosmic animals). Such experiments, in addition to the light' they will throw upon the function of the cutane- ous receptors, will guide us and serve as a means of con- trol in determining the animal's experimental range of sensitivity to olfactory stimuli. Furthermore, they will assist those of us who are engaged in the study of homing. As is well known, some of the homing theories maintain that the bird is guided back in its flight by responding to appropriate air currents, which may differ in their cutaneous and possibly in their olfactory values. Further- more, problems concerning the presence of a " common chemical sense " and those connected with the functioning of the lateral line organs lie in this field.
II. Instinctive Functions Introduction. — In the chapter devoted to the special con- sideration of instincts, we shall discuss more intimately the nature of instinct, its origin, source, etc. In the present connection we are concerned with some of the broader questions which must be raised concerning the relation of congenital response to acquired forms of behavior, and to the study of the sense-organ processes.
Some types of instinctive response. — In the first place, it is essential to know the stimuli which have native and unacquired biological significance: those to which, apart from training, the animal responds either positively or negatively or by the inhibition of overt response altogether. These types (and many others as well) may be more easily observed oftentimes in the lower forms of animal life, but PKOBLEMS IN INSTINCT 43 they are not lacking in the higher forms, as has been al- ready pointed out. In addition to these there are those types of stimuli which produce first inhibition and then negative response, as the call of the mother hen to the chick when a shadow passes over the field of vision; flickering lights; certain types of noises; and sudden changes in the intensity and character of lights and sounds. The necessity of knowing the instinctive tendencies of our ani- mals comes out clearly in experimentation. Suppose, e.g., we desire to test the retention of an association between a very weak light and a brighter light, or that existing be- tween two colors: we find from our preliminary work that the animal is positive to the one or the other, — shows an instinctive tendency to go to the lighter or darker of two lights, or to the one or the other of two monochromatic lights. If the animal is allowed to obtain its food with the light to which it shows a strong instinctive tendency to re- act positively, we shall, after the lapse of the desired time, not be testing retention at all: we shall be showing simply that the native positive tendency still persists.
The animal's instinctive mode of attack on problems. — Furthermore, we have to be concerned with the animal's method of dealing with objects and problems. The rat and the squirrel will claw, bite, hold small articles with the forepaws and nibble at them, sitting the while on their haunches. The chick, under similar circumstances, will peck at objects too large to seize and swallow whole, and will move them all around the room without attempting to secure them with the feet. Carnivorous birds will hold the booty with the feet while they rend and tear it with the beak. The monkey will pick up small objects and carry them to the mouth with one hand and bite off small portions much like the human being. The list of these native differences in reaction is almost endless. It is desirable to have accurate knowledge of them in order to shape our prob- lems for the study of these functions and of the ani- mal 's types of learning. The failure to take account of the instinctive capacities of animals has led certain naturalists and even scientific students of behavior to criticise many 44 SOME PROBLEMS ENUMERATED of the methods in use in the behavior laboratories on the ground of being unsuitable to a particular animal's mode of life. In some cases the criticism may be just.^ On the whole, we think the behaviorist has not shown a lack of ingenuity in adapting his studies to the different forms of animals.
Instinct and habit. — It is because of the intimate con- nection between instinct and habit that one most desires to get a clearer knowledge of the animal's repertoire of per- fected instinctive responses, his partial and incomplete adjustments, and even his tendencies towards adjustment before beginning controlled work upon habit formation. In the common acts of animals how much is hereditary and how much is ontogenetically acquired? The squirrel has the ability, apparently, to pick out a sound from a faulty nut. Some animals seem to have an instinctive endowment with respect to the choice of food. The everyday functions of drinking, catching, seizing, and swallowing prey or booty probably belong in the class of partially instinctive and partially acquired reactions. The chick, e.g., while ap- parently it can drink instinctively, needs habit to perfect the complex act of striking, seizing, and swallowing small objects. Only careful experimentations upon young ani- mals will enable us to tease habit from instinct. The extent to which we have gone in this work will be taken up in Chapter IV. The problems in instinct have been little more than hinted at here. Enough has been said, it is hoped, to show the necessity for a wide knowledge of animals' instinctive capacities. We have literally hundreds of books upon instinct. Most of them are worthless. If only the naturalists who made them had been taught the patient habit of stating just what they saw and not enlarg- ing upon it, what a mine such books would be to us! As it 2 Criticisms have been several times directed against the carrving out of experiments in the dark room. These objections are pointless: first, because the behavior of the animal in darkness is just as im- portant a part of our work as his behavior 'in light; and second, even those animals which ordinarily use daylight, work perfectly well in the dark room and remain in good physical condition throughout the course of the investigation.
HABIT FORMATION 45 is, we must begin all over again. Here, more than in any other place, we feel the need of careful field observations.
III. Learning Introduction. — On account of its bearing upon human training, learning in animals is probably the most impor- tant topic in the whole study of behavior. Entirely apart from this connection, this division contains the behaviorist's most important group of problems, since by means of habit formation he finds the most direct way of controlling ani- mal activity. The fundamental instincts cannot be easily changed, but habits which are just as efficient and just as economical in the matter of energy expenditure can be im- planted alongside of them. There seems to be almost no limit to the number of habits which the higher vertebrates may form, or to the complexity of such habits. The work of Pfungst on '' Clever Hans," and of Krall upon the Arabian steeds are examples. We cite also the training of Peter, Consul, and Roger, as well as a host of other less well known animals. But these animals have been trained for purposes of exhibition. Their acts are almost com- pletely without scientific significance except in so far as they show the behaviorist the possibilities in delicate methods of training and the limits within which training may take place. Before beginning upon the simplest problem in learning, it is necessary, as we have already shown, to have some knowledge at least of the instinctive modes of response of the animal and of the receptors to which we are making appeal. In behavior up to the present time, we have largely put the cart before the horse. In entire ignorance of in- stinctive capacity and sense-organ functions, we have plunged in medias res and attempted to do satisfactory work upon learning. We have tried to point out under I and II that work of this character can never be final. Most of the work has been undertaken for the purpose of gaining orientation. We have felt the necessity of making broad and general studies. Nearly all of the longer monographs have treated of the senses, instinctive equipment, and the 46 SOME PROBLE]\IS ENUMERATED capacity for learning of each animal studied. No one would attempt to study a new race of human beings in this whole- sale way. While it is highly essential to have pioneer studies it is equally important for us to remember that such studies are purely provisional and temporary. When we have finished our general survey, we are then prepared to begin really intensive work upon the animal.
Grouping of problems. — Problems of learning and reten- tion may be grouped under three headings: (1) the " per- severance " or fundamental method of acquiring control of skeletal muscles; (2) efficiency of training methods; and (3) complex forms of learning.
(i) The perseverance method. — The work on trial and error learning (better, '•' perseverance learning ") has been rather desultory. It may now be safely taken for granted that every animal form from the Paramecium to the primate can learn by this method. Any further work in this field ought to have a definite purpose: either that of establishing norms in a particular form for the learning of certain kinds of problems for comparing with similar re- sults from other animals, or, that of obtaining laws of plasticity with respect to age, sex, species, etc. It seems quite essential, if we are not to waste time, to devise a set of standard problems which can be learned under stated conditions. There is no agreement at present about the most serviceable kind of problems, but there are movements towards such a standardization. In the chapter on methods several problem boxes will be suggested which seem to have a wide range of usefulness and adaptability. In the work of establishing norms for inter-species compari- sons, similarity in the types of problems is all that can be hoped for, so different are animals in their structure, size, and action systems. In the intra-species work all of the separate groups (the separate groups differing only in age or sex, etc.) of animals should learn the same problems under identical training methods.. There has been great laxness in this respect. Many of the experimenters have not even felt the necessity of keeping the conditions started with rigidly the same throughout the course of the forma- HABIT FORMATION 47 tion of the habit. Intercomparability of records should be the chief desideratum. Given the learning curves of these simple problems, we should next, under conditions as nearly- identical as possible, determine the retention of the habit after varying intervals. It is unfortunate that this has never been done with care. To do it well requires a large expenditure of time and effort. It is also highly valuable to obtain records for the learning and retention of complex habits. This has already been partially attempted in the case of a few animals, — the monkey and the raccoon. The problem chosen was a complex puzzle-box consisting of a combination of latches, bolts, and levers (p. 98). This type of problem is, in a way, highly artificial and foreign to the daily life of the animals. However, it offers an interesting means of establishing a set of ' ' unit habits ' ' which may be found later to be of great service in tests on imitation (p. 277). It is hardly more foreign to the life of the animal than are nonsense syllables to the human being. All admit the value to the latter of taking material that is far re- moved from his daily sets of habits. Similar arguments hold for the necessity of choosing like material for the animals, since here, as in the case of man, we wish to study what might be called the crude or native learning and re- tentiveness (corresponding" to the experimental range of functions as opposed to the practical range of everyday life).
(2) Efficiency of training methods. — With standard records on the learning of simple problems before us, we are in a position to take up what we may call, broadly speak- ing, the distribution of effort in learning. What is the most advantageous way to teach a problem to an animal? Should we allow him to work at it continuously? Shall we give him one trial a day? Three trials a day? Five trials a day? Shall we allow our animal to form three habits abreast? Or is it more advantageous and more economical in every way to allow him to form his habits serially? What forms of incentives should be used, food or reward of some kind, punishment or a combination of re- ward and punishment? Furthermore, how shall we regu- 48 SOME PROBLEMS ENUMERATED late these incentives so as to obtain maximum but constant stimulating effect? We have been engaged in the study of these problems for some years now. That we are slowly gaining an insight into the efficiency of training methods may be seen from the summary of such work given on p. 228. It is at this point that the study of behavior has its closest contact with experimental pedagogy. On account of the numerous conventions which everywhere surround the human being, it has been next to impossible to intro- duce experimental methods into the teaching of children. We are content, as we have been in the past, to teach in the high schools four or five subjects a day and to spend thirty or forty minutes, as the case may be, on each of these subjects. Who knows whether we are right in this pro- cedure? Surely here, if anywhere, animal behavior can be of practical service to the experimental student of educa- tion. In Chapter I it was urged that the results of experi- mental pedagogy should be put in such terms that they will be directly comparable with those gathered from animal experimentation. It is hoped that the reasons for urging this are now clearer.
(3) Complex forms of learning. — The third division in learning is connected largely with the social life of the animal. To what extent is the animal influenced in his learning, (1) through imitation of a mate; (2) through imitation of the experimenter; (3) by being put through an act; and (4) by various forms of encouragement? Finally, do we find that animals learn by methods which are closely analogous to those employed by the human be- ing? A large amount of work has been done upon these various forms of complex learning. The problems are deli- cate. The results at hand are not harmonious. All of us have worked rather blindly. We have put our animals into strange and complex situations and have expected them to imitate acts of our own devising (which are oftentimes quite difficult), without being sure that the various simple coordinations C unit habit ") are really parts of their sensory-motor equipment. Our procedure is on a par with asking a man who has never played the piano to HABIT FORMATION 49 render " The Spring Song " on that instrument. It is useless to ask young children to imitate acts as wholes where the elementary coordinations are lacking or are ill- formed. There must be complete mastery of simple habits, — a readiness to respond to a difficult and complex environmental setting in a variety of different ways, — the ability to change responses ever so slightly to meet the slightest change in a heretofore well-known object. In order to do this our stock in trade of acts must be much more numerous than the objects to which we respond. Any one familiar with the early tentative imitations of chil- dren,— their failure to imitate at early ages the types of acts we call upon animals to imitate, — will not find fault with these statements. Apparently new coordinations are not established by imitation either in man or in animal. What is new is the combination or method of grouping. Where imitation appears there are found always groups of flexible responses to every object worked with. Further- more, one is able to change the background or setting of the object (coincident stimuli) without breaking down the habit. It often happens that after an animal has learned to do a certain act, open a problem box, e.g., under given conditions, a change in those conditions, — removal of the box to another place in the room, — will produce a breakdown in the habit. There seems to be a lack of readi- ness on the part of all animals to shift responses to meet slight changes in objects oft-reacted to. G-iven enough co- .ordinations of an elastic kind and it is hard to see how an animal can fail to imitate. Whether such conditions are realizable in the large majority of animals is the problem for experimentation to decide. The extent to which experi- mentation has thrown light upon such questions is discussed in Chapter VIII.
IV. Correlations Introduction. — Studies upon the structure of sense organs and upon the nervous system generally seem to be somewhat out of fashion at present. We have never, in 50 SOME PROBLEMS ENUMERATED this country, paid the attention to the histology of the sense organs and to the comparative anatomy of the central nervous system which these subjects deserve. This condi- tion of affairs is projiaid^'^^ggjf^iimorary. It is cer- tainly devoutly to JfJ^Soped that the^-sfu^ of sense-organ structure, integraf^^s in the nervous ^sVem, localization of function, etc/ will, icmjsci^ Ifiklwre JlwJng of the pen- dulum, once moik^ecome central with the anatomists and neuro-physiologis?lL It seems dis^uj^ing that neuro- logical conceptions a^^^tJ^fi^^rWe sadly need work- ing concepts of the nature of nervous impulses, of the anatomical and physiological nature of the sensory endings in the sense organs which are responsible for the different types of elementary functions and of the different chemical bodies set free by these highly organized structures: of fatigue products in the muscles, of the processes which make for the adaptation of sense organs and the like. In this day of advanced physiological and neurological technique surely the only difficulty in obtaining satis- factory answers to these questions is the lack of suffi- cient interest on the part of the men who are competent to carry out such researches. The interests of biological scientists are engaged chiefly in the study of evolution (variations, mutations, heredity, etc.), in the physiology of secretion, nutrition, and metabolism generally, and in ex- perimental embryology. The absorbing interest which these subjects afford leads to the neglect of topics which are not connected with the direct life-subserving functions of the body.
Lack of behavior data handicaps the neurologist. — The lack of accurate knowledge of the range and delicacy of sense-organ functions, and of the integrations among sen- sory responses, — data which the behaviorist must supply, — makes the more careful neurologists draw back from the detailed work upon the structural elements which must be correlated with particular functions. A history of correla- tion studies shows quite clearly that the functional side has not been worked up with sufficient care to furnish anato- mists and physiologists with the proper bases for correla- COERELATION 51 tion. There is little question but that a large number of correlations now adhered to between function and structure will have to be seriously modified or changed altogether. We call attention here to the work of Kalischer, Swift, and others upon the localization of the pitch center in dogs. The anatomical work was based upon the crudest kind of behavior work. At present.the human psychologist at- tempts to correlate sensibility to wave-length differences with the functioning of the cones, and the finer sensibility to differences in white light with the functioning of the rods (and cones). The cortical projection centers for both" sets of impulses are placed in the occipital region. The work of Franz, von Monakow, and others shows that we are not at all upon solid ground even with respect to the simple cor- relations which are used in psychology in everyday teaching. Simple working conceptions of what goes on in the nervous system when habits are formed will be very stimulating to the student of behavior. It would seem that the physical chemist and the physiological chemist might advance help- ful facts bearing upon these questions. The behaviorist is ready to assist in the work. He is in a position to force the formation of simple habits in whatever sets of sensorymotor arcs the neurologist or physiologist may specify. He is willing to take animals at infancy and establish such habits in neural systems not previously largely exercised as, e.g., in association tracts in the cortex. Whether such work will ever throw light upon what happens at the surfaces of separation (synapses) through habit formation remains to be decided by future work. It is worth a thor- ough trial in some well-equipped neuro-physiological labora- tory. It is quite possible that the increasing accuracy with which sense-organ functions are being isolated will give a very decided stimulus to structural w^ork in this country. Until the returns from this work begin to assume large proportions, one can hardly blame the more careful neuro- physiologists for interesting themselves along other lines. Donaldson for many years has been amassing a wealth of detail upon the growth processes and anatomical correla- tions in the rat. These studies have taken many different 52 SOME PROBLEMS ENUMERATED forms, — from determining the total body weight and weight of the central nervous system (relation between central nervous system weight and body length or weight) of the animal at different ages, to that of the analysis of the chem- ical constituents of the nervous system at different ages. The work upon behavior has just reached the point where it may begin to take advantage of such material.. C. J. Her- rick has suggested the advisability of centralizing behavior and anatomical work upon some one species of animal, e.g., upon the rabbit. While it is not advisable to limit behavior studies in any way, much good will come even from bear- ing such a suggestion in mind. The problems singled out should be definite and simple, and designed to emphasize the functions of particular organs or parts of organs. This would enable the anatomical work to proceed without wait- ing for complete returns.
General aim of behavior. — It is believed that the next few years will bring order out of chaos in the physico- chemical relations in the nervous system. That the organ- ism is a machine is taken for granted in our work. The only point we insist upon is that the machine be made not too simple to perform the multitudinous demands which the behaviorist must make upon it. There has been a strong tendency on the part of many biologists to assume that the mechanisms are exceedingly simple. '' Having found that the ovum of the sea-urchin can be fertilized without the stimulus of the spermatozoon, they fail even to give the egg the credit which rightfully belongs to it! " It is hoped that the chemical and physical relations which exist in the egg it- self can soon be so thoroughly understood that the egg too can be put together in the laboratory and then fertilized. The production of artificial protoplasm in the laboratory will in no way interfere with the study of behavior. Even after we have made it, it is highly probable that nature will still be allowed to take her course. As long as we have ani- mals in the world it is hard to see how the demand for a study of their methods of living, moving, and having their being can ever grow less. That such phj^sico-chemical stud-