— (a) Instinctive repertoire of sounds. — (6) Modification of vocal sounds through social influences. — (c) Influence of such sounds upon mates. — 3. Olfaction and gustation. — Lack of defi- nite knowledge about smell functions. — Problems in olfaction. — Function of olfaction in habit formation. — Problems in gustation. — 4. Cutaneous, proprio- and entero-ceptive systems. — Difficulties in the way of making satisfactory tests. — Proprio-ceptive system. — Cutaneous system as a distance receptor. 11. Instinctive func- tions.— Introduction. — Some types of instinctive response. — The animal's instinctive mode of attack on problems. — Instinct and habit. III. Learning. — Introduction. — Grouping of problems. — (1) Perseverance method. — (2) Efficiency of training methods.
— (3) Complex forms of learning. IV. Correlations. — Introduc- tion.— Lack of behavior data handicaps the neurologist. — Gen- eral aim of behavior. — Summary.
No lack of unity in the problems of behavior. — In the preceding discussion of the relation of behavior to psy- chology several problems which face the behaviorist were touched upon. In the present chapter the attempt is made to develop them in such a way that their unity may be traced. At the present time behavior appears to the casual observer to consist of a large number of rather isolated bits of research which may be classified here and there under sciences already well recognized. This is a serious mistake and one which will handicap advance in this subject in the years to come. The sketch of the problems given below may be looked upon as a program for unified and sys- 29 30 SOME PROBLEMS ENUMERATED tematic work rather than as a complete expression of the scope and aim of behavior. Even such an immature and hasty sketch of the problems confronting us will, it is hoped, offer convincing evidence that the work of the behaviorist, while closely related to that of the zoologist and the physi- ologist, is, nevertheless, independent.
Field versus laboratory studies. — Before presenting the special groups of problems of behavior, it is necessary to face one criticism often urged against the young science, viz., that of the narrowness of its problems. In the last few years behavior has become mainly a laboratory science. This has led a large number of men who have devoted their lives to the field activity of animals, such as Burroughs, Wesley Mills, and a wide group of naturalists, to say that the laboratory is not a suitable place in which to study behavior. According to them only highly specialized prob- lems can be attacked in the laboratories. True and un- trammeled expressions of habit and instinct and of the uses of the senses must be sought for in the;field activity of animals. Unquestionably it is a mistake to neglect field work. It requires no lengthy argument to show that gen- eral orientation with respect to the daily routine of adjustments of animals and an accurate knowledge of the environmental conditions under which animals live can come only through field observation. No one who has ever used monkeys as subjects can help feeling how handicapped we are at the present time in our laboratory studies of simian life through lack of systematic knowledge of their life in the open. What is true in the case of the primates is true with respect to nearly every other animal form. It is highly improbable that any of us could describe in a really helpful way the daily routine of the domestic fowl or the dog. We are even less familiar with the seasonal routine of animals, such as hibernation, migration, etc. On the other hand, it can hardly be claimed that mere observa- tion of field activity, even when made by competent students, can ever hope to answer in any scientific way the basal questions which must be asked about the mechanics of stimulus and response. Even the most superficial ob- FIELD OBSERVATION 31 servation of field activity by the trained student raises at once a host of questions, the answers to which must be sought in the laboratory — is the cat visually stimulated by a moving object in steadily decreasing intensity of light for a longer time than the human being, or does this animal stalk its prey after it is lost to sight through a highly de- veloped sense of smell? Do the birds which feed upon decaying animal matter sense it through smell or sight? A few minutes ' observation will show that while many of our problems are raised in the field, the scientific answers to them come through the laboratories. But granting the in- dispensableness of the laboratory, it is well, after finishing with our animal, to observe him yet again in the field. One can readily conceive of some such situation as the following: — after years of analytical study upon the tem- perature, visual, olfactory, and auditory senses of a given species of bird, one might predict the utter absurdity of that bird's being able to get back to its home when carried out to sea for a distance of one thousand miles. Yet on specific test we find the bird able to do this. Of course our laboratory study was incomplete, or we should have been able to predict what actually happened. Yet the incompleteness of the laboratory study would be discovered only when we, so to speak, began to put the bird together again! "Without developing the subject further, it would seem obvious that there is no conflict between field work and laboratory work. The field is both the source of problems and the place where the laboratory solutions of these prob- lems are tested.
Grouping of problems. — The vast majority of the prob- lems in both human and animal behavior may be grouped under one or another of three divisions: I. Se7ise organ functions..II. Instinctive functions. III. Habit forma- tion. In addition to these large divisions in which the subjects for research lie, there remains the work of, IV. Correlation: first, among behavior data — giving both an ontogeny and a phylogeny of behavior; second, of behavior with structure; and finally the correlation of behavior and structure with physico-chemical processes. "We shall take 32 SOME PROBLEMS ENUMERATED up these divisions separately and attempt to show in some detail what special questions gather around them. Later we shall discuss the apparatus and methods which are employed in attempting to return answers to them.
Introduction. — The study of sense functions should come prior, logically, to the study of either instinct or learning, since neither instinctive action nor learning can be thor- oughly understood until we have definite knowledge about the sense processes of the animal under experimentation. In actual practice, work upon instinct, learning, and the sense organs has been carried forward simultaneously. It is very difficult to study any one of these subjects by itself. In the learning of mazes and puzzle boxes both by animals and human beings, many facts appear which are of impor- tance to a study of instincts and senses. On the other hand, our knowledge of sense-organ processes is obtained in two ways: (1) by forcing the animal to form sensory habits (p. 187); and (2) by noting the inherited modes of response to controlled stimuli. In the study of the spectral sensibility of the chick (limits) one finds that the chick is positive to light, i.e., will go towards a lighted compart- ment from the first, without learning. The relations among these three divisions seem at first sight to be too complicated for analytical work. It will be seen from the chapters that follow that the difficulties are more apparent than real.
1. Vision 1. Vision Necessity for consideration of anatomical structure. — Before making an extensive study of vision in any par- ticular animal form, it is essential to study carefully the structure of its eye and the visual conduction systems; and to have at least some knowledge of the animal's general anatomy. Questions as to the presence or absence of a fovea; whether both rods and cones are present, and their distribution; whether the animal has the essential structure PROBLEMS IN VISION 33 for binocular vision; whether there are accommodation, convergence, and divergence — are met at every turn in our work. Not to have orientation with respect to them means a lack of thoroughness in the setting of problems for the study of vision. It is equally essential (where possible) to make a study of the animal's visual environ- ment and of the field habits and instincts which seem to depend upon visual stimuli.
Reflex responses to light. — In the observations of the movements of lower forms of plants and animals, one is often struck by the remarkable sensitivity of some of the organisms — Euglena, Stentor, — and by the equally notice- able lack of sensitivity in other forms, e.g., Paramecium. The general responsiveness to light is equally observable in the higher animals as well. The tern remains absolutely quiescent in total darkness. The chick's behavior on the sudden elimination of light is equally remarkable. It re- mains at first quite still, as though stunned, and then be- gins aimless movements, — pecking, turning, running against objects, etc. The rats, and probably other rodents as well, require separate laboratory work in order to tell whether vision is functional at all. Vision seems to func- tion in some animals (nocturnal animals) only in very weakened intensity of light, strong light producing cessa- tion of activity equally as well as absolute darkness.
General divisions of problems in vision. — The problems in vision proper may be considered under {a) white light vision; (6) monochromatic light vision; (c) the role of vision in daily life (mutual relations among the senses).
White light vision. — («) One of the first problems in white light vision is the determination of the delicacy of the mechanism. What absolute intensity of light can the animal respond to with dark adapted eye, and with light adapted eye? What difference in intensity between two lights is it necessary to maintain in order to give a basis for a differential response? (Weber's law.) Are re- sponses to light positive or negative? Under what condi- tions may a positive tendency be changed to a negative?
34 SOME PROBLEMS ENUMERATED Can it be so changed in any of the vertebrates without in- volving habit formation? We are led over almost at once into a consideration of darkness and light adaptation in animals with cone retinae and in those possessing rod retinee. The difference in the response to white light be- tween animals with image forming eyes and those with eye structures too undeveloped to form images. Another for- ward step in the work on vision is the determination of the animal's behavior with respect to form and size, vertical and horizontal lines, moving and stationary stimuli, pat- terns, etc.
Response to monochromatic light. — The difficulty of gaining accurate knowledge of animals' monochromatic light responses is very great. Heretofore the work has been carried out by methods and apparatus which can never yield accurate results. The first step in the study of any animal's color responses should be the determination of the limits of its spectral sensitivity. The second step should be the determination of the energy of stimulus necessary to yield threshold responses at various points in the spectrum. Enough places should be chosen to enable the experimenter to plot the sensitivity curve throughout the spectrum. This threshold curve will serve in several connections. In the first place it enables us to say immediately whether both lights used in the work on sensory habits lie above the animal 's threshold, and con- sequently whether both are effective stimuli. We have learned the necessity for this after the loss of several months' work. In the second place it will largely increase our knowledge of the variations of intensity which it is nec- essary to make the monochromatic lights undergo in order to test for sensitivity to wave-length difference (color vision). Let us illustrate: suppose we have found that even at our standard intensity the red has very little stimulating effect upon the animal, i.e., that it is only slightly above its threshold. Now if one is confronting the animal with red and green it is perfectly evident that if we cut down the intensity of the red in only a small degree, it will cease to stimulate the visual receptors. This is only one of the PROBLEMS IN VISION 35 features which make it desirable to obtain the limits of the spectrum and the sensibility curve for the animal whose vision we wish to study at length. The next step, logically, in the monochromatic light work, is the careful working through of the Weber law for intensity at several points in the spectrum. Our interest is not psychophysical here but methodological. We need to know for subsequent con- trol of behavior, what difference in intensity one has to maintain between two reds of the same wave-length in order to afford a basis for a difference in response. The problem is simple enough in technique, as will be shown on p. 357. One allows a given bundle, red, for purposes of illustration, to come from the spectrum. This is then divided by means of a double image prism into two beams equal in intensity. The intensities are then controlled separately by means of smoked wedges (or rotating sectors). The tests on the animal are carried forward in exactly the same way as in tests on sensitivity to wave-length differ- ence. The tests should naturally be repeated with different absolute intensities. The Weber-Fechner law may or may not hold. Regardless of the bearing of the work upon that law we will gain a clearer insight than we now have of the relative stimulating value of the different colors at low, intermediate, and high intensities. The results of these experiments upon differences in intensity should aid us in testing the possibility of response to differences in wave- lengths, which is the most difficult problem in animal vision. Since this problem is already well known, we leave a further discussion of it to p. 356 ff., where the results of tests on color vision are considered.
Finally, as a means of controlling visual response it is essential to test the relative effects of decreasing the in- tensity of light by the four different methods usually em- ployed: (1) distancing the source of the light; (2) thinning the beam by the use of diaphragms; (3) by the use of the rotating sector; (4) absorption by means of smoked (neu- tral) wedges. As is well known, careful experiments have shown that in the human being it is a matter of indiffer- ence (Talbot's law) which method is used. This cannot 36 SOME PEOBLEMS ENUMERATED be assumed to be true in the case of animals. This work should best be undertaken with monochromatic light.
2. Audition Grouping of problems. — The problems in sound may be grouped under three general headings: (1) response to pendular vibrations; (2) to a-periodic vibrations; (3) to the ordinary sounds in the animal's environment (mates, other animals, etc.).
(i) Response to pendular vibrations. — As in the case of vision, so in audition, field observations ought to guide us in making our audition tests. We can do little more here than to enumerate the various problems which lie be- fore us.
(a) Range of sensitivity. — The animal's range of sen- sitivity to simple pendular vibrations is probably the most fundamental problem in this field. The work can be car- ried out with good forks by forcing the formation of sen- sory habits. The method of making these tests is described on p. 81. The tests should be combined with threshold tests and with others designed to bring out differential sensitivity both with respect to pitch and to intensity. While there is no convenient standard of intensity at pres- ent, it is possible for the work to be carried out upon differ- ent animals in the same laboratory with the same apparatus. We can obtain in this way values which are strictly com- parable for the different animals. Such work is, however, not reproducible in other laboratories unless some system of exchanging apparatus is adopted.
(&) Localization. — The localization of sound stimuli (in the case of bodies which give periodic and those which give a-periodic vibrations) is important both for obtaining knowledge of the experimental range of auditory sensitivity and for tests of the function of - sound in the daily life of the animal.
(c) Response to clangs. — The question as to the ability of the animal to respond to differences in the timbre of sound complexes (analysis of chords) has m interesting PROBLEMS IN AUDITION 37 bearing upon the nature of structural relations in the inner ear (complexity demanded in organ of Corti).
(2) Response to a-periodic vibrations: (a) Stimulus threshold. — The falling ball (some standard apparatus should be chosen) gives us the best means of testing stimulus thresholds. The tests would be very easy to carry out. It is surprising that this work has never been attempted.
(b) Difference threshold. — The differential sensitivity for sound intensity should likewise be worked out with a similar instrument, but in this case the problem is more difficult, since two stimuli must be employed. The tests of response to periodic and a-periodic vibrations are designed to give us the experimental range of auditory functions, and not the practical range employed in daily life. The problems which follow are connected with the practical range.
(3) Response to the ordinary sounds in its environment (mates, other animals, etc.): (a) Instinctive repertoire of sounds. — The instinctive repertoire of vocalization is im- portant, if for no other reason than to guard ourselves against those enthusiastic but untrained investigators who would tell us that animals have a language. It is possible to obtain this instinctive repertoire only by isolating the animal until the native vocalizations have all been recorded, and then comparing the vocalizations of the isolated animal with those of an animal brought up in company with its fellows.
(&) Modification of vocal sounds through social influ- ence.— The extent to which such vocalizations are modifi- able by social surroundings may be illustrated by the work of Conradi, who reared English sparrows with canaries, — certainly with no hardship to the voice of the sparrows (p. 143). As a part of (&) we include the problem of find- ing to what extent animals can be made to respond cor- rectly to articulate sounds made by the human being (note the behavior of the dog, Jasper; the chimpanzee, Peter, and of other highly trained animals, p. 299), (c) Influence of such sounds upon activity of mates. — The extent to which animals are induced to engage in com- 38 SOME PROBLEMS ENUMERATED mon activities by the voice of the mate, parent or com- panion (sex activity, feeding, flight, etc.)- While these problems possibly may have the flavor of the ^' manufac- tured " or laboratory variety, a little field observation will quickly show that they are really fundamental. Hodge would have us believe that the deer has an almost unbeliev- ably acute sense of hearing. How can we dispute it or confirm it in the deer or any other animal until we have made threshold tests under standard conditions? Kalischer tells us of the absolute pitch memory of dogs. Who can prove it or disprove it except by the laboratory type of test? Until these exact and rigorous tests are made and repeatedly confirmed by several investigators it is impos- sible to carry forward research upon the localization of pitch centers, experimentation upon the functions of the cochlea, and the like.
3. Olfaction and Gustation Lack of definite knowledge of smell functions. — It is singular that the functions of the organ of smell have never been investigated in any complete way in any ver- tebrate. Several investigations have been carried out upon smell in birds and in fishes. These tests have had as their object the determination of the fact whether those animals use that receptor. Of the range and complexity of its uses we have nothing. That enormous differences exist among the smell functions of different animals there is little room for doubt. In addition to the problems con- nected with the general functioning of the organ of smell, there are many concerning the instinctive life of the animal which have never been explored. We have in mind here the positive reactions made to certain " nauseous " smell stimuli, and to the apparent lack of sensitivity to the odor of flowers and perfumes generally, etc. The human being seems to be the only animal which responds negatively to the class of odors which Zwaardemaker calls nauseous.^ ^ K. S. Lashley finds that an Amazon parrot in his possession will vomit at the smell of an old pipe (Class VI).
OLFACTION AND GUSTATION 39 How much of this is due to social training and how much to fundamental biological tendencies is not known. From the study of primitive Australian tribes and from bio- logical studies generally the conclusions seem to follow that social conventions and training are responsible for the negative tendencies which are so apparent in man. Still, the evidence is not completely decisive. Certain smells in the cultivated European are connected with definite re- flexes: nausea and even vomiting being produced. That our ignorance of smell functions in general is almost colossal comes out clearly when we try to compare even in thought the delicacy of the average human being's sense of smell with that of certain varieties of dogs. Hunting dogs, blood- hounds, etc., must have an almost unbelievably delicate sense of smell. Yet when defective human beings have been forced to depend upon this sense, they have surprised us by the delicacy with which they use this organ. It is diffi- cult even in field work to detect the actual uses to which smell is put by animals. In the case of the monkey it is quite clear that objects are rejected by this sense long before they reach the mouth. That olfactory stimuli start seeking movements in many vertebrate forms is also clear. The specific problems in smell are much like those in the other sense fields. There is need of field observations to guide us in setting problems for determining the interrelations of smell with other sense functions, and the role smell plays in the daily life of the animal. As in vision and in audition we need to test the animal's range of sensitivity. For this problem the best plan of attack would probably be that of taking several odors from each of Zwaardemaker's nine classes and testing in order whether the animal is positive or negative to them. Either form of response, provided proper controls were introduced, would show that that odor was an effective stimulus. If no such simple inherited mode of response were present we should attempt to force the formation of a sensory habit which would enable us to determine whether the odor lies within the animal's range. Extended experiments would be necessary for the determi- nation of threshold intensities of different olfactory stimuli.
40 SOME PROBLEMS ENUMERATED We shall sketch a method farther on (p. 89) by means of which such tests may be started at least with some hopes of results.
Function of olfaction in habit formation. — The extent to which smell stimuli influence the formation of habits in the daily life of most of the vertebrates is at present un- known. It is generally believed that smell forms the center of reference for most animals. What tests have been made do not confirm this. That smell data are utilized in habit would seem to be clear from the anecdotal literature, and yet in many of the tests on the formation of motor habits we have been enabled to eliminate this sense without decreasing the rapidity with which learning takes place.
Problems in gustation. — Gustation offers as its main problems: (a) the ability of animals to react differently to sapid substances either instinctively or through habit; (h) the delicacy of the organ in detecting taste substances in solution (and at a distance), and (c) the localization of the taste organs. The sense of taste offers a more inviting field to students of lower vertebrate and invertebrate be- havior than to those engaged in work upon birds and mammals.
Difficulties in the way of making satisfactory tests. — We experience great difficulty in making tests upon the tac- tual, kinesthetic, organic, and other internal systems of receptors. Some interesting work has been done in all of these fields, but we have not gone far enough in such in- vestigations to talk very intelligently about the problems. Work here calls almost instantly for operative technique.
Proprio-ceptive system. — The maze experiments and those on the so-called delayed reaction (pp. 99 and 104) have been the most effective ones so far in yielding results. There still remains a very large number of problems in con- nection with the maze. The disturbance which results from rotating the maze after the animal has learned it is a case in point (p. 219). Many observers have noted this phenom- INTERNAL RECEPTORS 41 enon. While it may possibly be explained in terms of the change in the balance which exists among the processes in the distance receptors, it may be due, on the other hand, to a change in the functioning of some of the internal receptors, possibly those lying in the semicircular canals and vestibule. The work which Carr and his students are doing upon the responses of animals to discrete impressions (delayed reac- tion)— e.g., exposing for one second a light in one of three places and then waiting a definite period, one, two, three, or five seconds, then releasing the animal to see whether it will go to the place where the light was flashed, — is showing us the extent to which kinsesthetic and organic receptors may function in ordinary situations (p. 224). The homing sense in birds is another problem which may lie within this field. No one can say definitely to the contrary yet. Our experience already noted in getting the terns to come back home from long distances over territory wholly unknown should make us hesitate at least in explaining the homing sense in terms of the visual landmark theory.