Consider her ways and be wise:
Which having no guide, overseer, or ruler,
Provideth her meat in the summer,
And gathereth her food in the harvest.
FROM Solomon to Roosevelt and John Burroughs the human race has displayed an interest in the habits of animals. The most versatile of the Greeks foreshadowed the course of all later natural histories. The Historia Animalium, written about 345 B. C. is bubbling over with the same sort of facts that one finds in the books of natural history to-day. ’It is the instinct of the hedgehog,’ wrote Aristotle, ‘to alter the entrance to his burrow when the wind changes from the north to the south, or to change from wall to wall at the approach of weather changes’; ‘the woodpecker has been known to place an almond in a crack of the tree to prepare it for a blow of his bill, and in its hunt for worms in the bark of trees, it hollows them out so much as to throw them down’; ‘the disposition of sheep is foolish and without sense, but many animals in their mode of life appear to imitate mankind ’; ‘ because the cuckoo is conscious of its own timidity, it lays its eggs in the nests of other birds that its young may be cared for’; and the philosopher’s pigeons ‘can distinguish ten different varieties of hawks.’ The reader can but marvel at the wealth of material which Aristotle gathered together, though he may often be amused at the naivete of the interpretations.
Historically, animal psychology falls into three main divisions: the naturalhistory period, from Aristotle to Darwin; the critical period, including Darwin, Romanes, and Lloyd Morgan; and the experimental period, which, beginning with Lloyd Morgan, is now in full career. The Darwinian period differs from all that went before chiefly in a more scientific scrutiny of the anecdotal material, and a careful arrangement of this material with a view to substantiating a psychological theory: with Darwin and Romanes, the continuity of mental life throughout the animal race including man; with Morgan the dominance of instinctive behavior and accidental learning. The experimental period, taking its cue from Morgan, was at first dominated by Morgan’s bias, but is now freeing itself from all presuppositions except that it is worth while to know what animals do, and what psychological processes they have.
The recent interest in the behavior of animals has arisen from interest in two other sciences. Psychologists, stimulated largely by the writings of William James, have shown an increasing desire to know the genesis of the human mind. Two possible avenues of approach present themselves: the study of the child, and the study of the mind as it appears in the animal world. So for a number of years we have had genetic psychology in the schools, ontogenetic psychology, and phylogenetic psychology. In approaching either of these fields, however, it was found that the most one could do was to speculate on the basis of a meagre collection of facts. This was particularly true of phylogenetic psychology, because all the material available consisted of the anecdotes collected from widely scattered sources. That some of this material was authentic no one doubted. Some of it had been gathered by such accurate scientists as Darwin and Romanes. But some of it also came from the hand of such good story-tellers as Buffon and Brehm. So much of it gave evidence of being colored by the reporter’s own illusions that to separate the true from the imagined was an impossible task. Students in the field realized that if we were to have a phylogenesis of mind that was in the least degree reliable we must have new data collected under conditions that were accurately known. That the collection of such data was to be a slow task was evident from the start; the work could only be done by men trained in the methods of science who could devote large amounts of time to the work.
The movement began from the psychological end with the publication in 1898 of a monograph by Thorndike on Animal Intelligence, the important part of the paper being the report of a series of experiments on chickens, cats, and dogs. This was followed by a paper by Dr. Small on the mental processes of the white rat. Other papers followed from both the Columbia and Clark laboratories, and before long a number of American universities were conducting research along similar lines.
Almost contemporaneously with this movement downward along the phylogenetic scale, biological science took a new departure. Attention had long been given to morphological and structural science, but for a time at least it shifted to a study of the processes of nature. This movement has recently been characterized by Professor Jennings in these words: ‘A new spirit has permeated biological science in every division, — in brief, the desire to see the processes of nature occurring, and to modify and control these processes, — not merely to judge what processes must have occurred. In the words of the young Clerk Maxwell, we wish “to see the particular go” of the processes of nature.... In the new spirit of work the desire is to see the things happening, not to conclude what must have happened. We wish to see the processes themselves, not merely the results of the processes.’ An early result of this new biological spirit was the study of the behavior of simple organisms. It was a study, not of what organs an animal has, how many it has, and where they are located, so much as a study of how an animal behaves under changing environmental conditions.
Naturally, the genetic psychologist had much in common with this new biological spirit, and the two sciences have met in common territory. The outcome has been a collection and grouping of facts that may well lay claim to being called a new science, a science which in its present intention, at least, is essentially experimental, and which we may call the science of animal behavior.
When one speaks of studying an animal experimentally it must not be understood that the animal is to be sliced, to be tortured, to be put into cramped conditions, to be placed at a disadvantage. To experiment means to know and to control the conditions under which the animal behaves. To draw one’s finger across the path of an ant with a view to seeing how the behavior of the ant is changed is to experiment. The animal must be free to do its best, it must be kept in health and free from fear. It must be given a square deal, and be allowed to display every atom of sense-power or intelligence that it can muster. On the other hand, it is absolutely necessary that all the modifying conditions that play any part in the animal’s behavior must be known, and that in successive experiments they must be individually varied so that the exact effect of each will be discovered.
The following will make clear what I mean. In the Harvard Psychological Laboratory the writer was studying a dog’s power of visual discrimination. We never can know much about the dog’s intelligence until we know something more about his senses. Can a dog see colors? Does he recognize persons by sight or smell? To what extent can he discriminate between two forms ? How accurately does he distinguish varying shades of brightness?
To test certain of these matters a device of this sort was used: The stimuli for reaction were two circles of flashed opal glass through each of which a twenty-five watt tungsten lamp sent its rays. The circles were separated by a wooden partition and the dog must pass down a four-foot board alley and select one of the two circles by going to one or the other side of the partition. In case the small circle was selected a trap in the bottom of the alley was opened by a sliding door and the dog, a cocker spaniel, was allowed to get food. In order that the animal might not be guided by smell, similar food cups were placed on either side of the partition, and in each of them were placed pieces of food of the same size and kind. That she should not rely on the position of the smaller disc, the circles of glass were arranged in an aluminum slide which could be shifted from right to left and back. The smaller disc thus appeared irregularly on the right and left side of the partition. To prevent the dog choosing by the brightness of the disc, the lights were fastened to lamp carriages which were mounted on tracks. The lamps could thus be moved far away from the glass or brought close to it, thereby altering the relative brightness of the two discs. To minimize the difference in the amount of heat coming from the lamps at unequal distances, water cells for the absorption of heat were placed back of the glass. Further to eliminate differences in light, the whole apparatus was painted a dead black and used in a dark room.
By thus ruling out smell, regularity of position, and differences in shape, light, and heat, it was intended to force discrimination by a single visual factor, namely, size. In later experiments the sense for shape, position, heat, light, and color could be made, and finally we could arrive at an accurate knowledge of a dog’s power of vision. Each of the factors could be varied independently and the part played by each accurately determined.
My experience with the first dog tried shows how difficult it is to keep tab on all the factors involved. My method was to give the dog from ten to fifty trials a day until she learned to choose a three-centimeter circle in preference to a six-centimeter circle at least eighty or ninety per cent of the time. When this act had been learned, a three-and-a-half-centimeter circle was substituted for the smaller one and the tests were repeated. I found no difficulty in getting the animal to go forward, and when she chose correctly I opened the slide door and she got food. It was not so easy, however, to induce the animal to come back to the starting place and I was compelled to put a leash on her. This I allowed to hang loosely, barely missing the floor. After a very large number of trials, the dog, whose name was Dolcy, began to choose the smaller circle, and soon her learning was progressing rapidly. She discriminated a three, a three-and-ahalf, a four, and a four-and-a-half-centimeter circle from a six-centimeter circle in rapid succession. In the latter case she learned the act in fifty trials, finally discriminating correctly one hundred per cent of the time. An important feature of her behavior was her apparent comparison of the two lighted discs. She would go straight to one disc, thrust her head into the apartment, stop a moment, step back, look into the other apartment while standing with an uplifted forefoot, look again into the first side, back to the other and again to the first, finally choosing the smaller circle, the experimenter all the while in interested suspense.
It can well be imagined that after the painstaking work necessary to bring such a study to fruition, the experimenter would be much gratified at the clear results, and the reader may possibly imagine the chagrin when he found that an unsuspected error had crept into the work. One day when the leash was removed during the experiments, the dog was unable to find the circle, the choosing of which had always brought her food. Repeatedly she essayed to choose and at least half the time failed. She hit upon the plan of going to one side all the time, and from the irregular plan of shifting the circles from one side to the other, she got food a number of times. Then she was refused food until she went to the other side, and she resorted to going to the brighter light. In short, she had to learn all over again; she had not at any time discriminated between the two circles. Instead of being guided by stimuli in front of her she had relied upon stimuli from another direction.
Much as I disliked to admit it, I could find no other explanation for the animal’s unexpected behavior than that I, myself, had unconsciously given her the clue to choice. The leash was the source of trouble; holding the strap in my hand and interestedly watching the animal’s movements I had unintentionally changed the tension of the leash. How delicate must have been the dog’s muscular sense will be realized when you recall that the leash all but dragged on the floor of the alley. Surely a good case of muscle-reading! That this is the probable explanation is evidenced by two facts. First, by the fact that when the leash was not used the dog quit making what had seemed to be comparisons of the two stimuli. She no longer looked from one side to the other, but went directly to one of the two circles. The other bit of circumstantial evidence was that when the leash was again put on, the dog had no difficulty in selecting the correct circle.
One turns with a good deal of skepticism from the deception of a rigidly controlled experiment like this to the wholly uncontrolled observations of the naturalist, especially when the naturalist attempts a psychological explanation of what he supposes himself to have seen. To sit back on one’s front porch and watch a downy woodpecker hollowing out his cup in the top of a chestnut post, think that ‘it may have been the first cavity of the kind it has ever made,’ and then conclude that the bird is controlled solely by instinct, is to be content with a crumb of doubtful fact when a little ingenuity and a willingness to try might give the whole fact. However engrossing such observation may be to the naturalist himself, and however entertaining the anecdotes may be to the popular reader, the science of animal behavior and comparative psychology must be founded on something more analytic and more verifiable. What, in detail, does the naturalist know of the downy woodpecker’s past experience? Has he ever seen this bird before? Will he ever see it again and note how differently it may work at another time? How old is it? What does he know in detail of the bird’s various sense-powers? How well can it smell or see? With what senses is it endowed with which the naturalist has no first-hand acquaintance? How often has it tried this same act and failed? Is it an average bird of the species, or an unusual one? What fortunate circumstances enabled it to invent a new plan of action? How do the various powers of the animal develop? How stupid has the pecker been in circumstances overflowing with opportunity for intelligent action? These and a thousand other questions the mere observer will not answer in a long, long time, and until they are answered we can never have a scientific study of the animal mind.
No experimental student of animal behavior would deny the value of well-authenticated anecdotes of the doings of animals, or the unspeakably precious contributions of naturalists of all time to our knowledge of the habits of the wild folk. But the necessarily fragmentary character of such material will always leave the animal mind a region of myth into which the would-be comparative psychologist can project the fanciful conceptions of his own mind; conceptions which serve not nearly so much to illuminate the field as the actual discovery of some small power of sense-perception or the exact part imitation plays in animal learning.
It is to find an answer to such questions as the naturalist cannot answer that the experimental method has come into being. Besides the discrimination method already set forth, investigators have used three principal modes of procedure: the puzzle-box method, the labyrinth method, and the method of the salivary reflex. The simplest of these is the labyrinth method. Usually some form of the Hampton Court Maze is used. The animal is placed at the outer end of the intricate network of alleys, and it must find its way about, past openings which lead into blind alleys to the centre. Interest centres in the manner in which the animal learns to avoid bypaths and to hasten its journey to the end. Small first used this device on the white rat, and numerous investigators have since employed it.
The puzzle-box method, which requires the manipulation of a lock or fastening in order to get food, is illustrated by an experiment the writer performed with monkeys. The monkey was confined in a cage approximately four feet square and six feet high. In the back of this cage near the floor an opening was made. This opening was closed by a glass door through which the monkey was allowed to see bananas suspended by a cord. The glass door could be opened by a string which passed from the door down under the cage and up a corner post on the front of the cage. The end of this string was fastened to a wooden plug put into the corner post halfway up on the inside. If the monkey could learn to climb this post and pull out the plug, he could then get the banana by going back down to the door.
The method of the salivary reflex has been used chiefly in the Physiological Institute at St. Petersburg. A fistula is formed by making an incision in the lip of the dog to the salivary duct, and diverting this duct from the inside to the outside of the mouth. The operation is easily made, the wound quickly heals, and the animal is apparently not disturbed by the event. The training tests are then begun. The dog is shown colors, and while looking at one color, say red, he is given a piece of meat, but when looking at other colors he is not fed. In this way an association is formed between the red color and the food. The experiment proper is then begun by showing the animal various colors in succession. In the training tests, red had come to call forth the reflexes connected with the getting of food, and now when red appears in the series, the reflexes occur even though no food is present. One of the most important of these reflexes is the secretion of saliva. The amount and quality of the saliva secreted indicates whether the dog can discriminate a red color from another color of the same brightness. The dog’s sense of sight is thus tested by a chemical and physical examination of its saliva. In this way the dog’s power of discriminating sounds, odors, and colors has been tested.
In a large part of the work so far done, investigators have relied upon hunger as a motive to induce animals to work. It was supposed that with regulated feeding you have here a motive of fairly constant intensity. My work with the dog indicates that food is an unreliable motive in the work with that animal. The dog will fare well on a small amount of food and, in the case of a very difficult task, his hunger is not sufficient to make him endure repeated failure. The daintiness of Dolcy’s appetite and the fiction of hunger being a constant stimulus became evident in my experiments on size-discrimination. After each successful choice the animal was getting a small cube of corned beef. The dog did very well, but one morning was greatly at sea in her choices; she went to the large and small discs indiscriminately and failed so often that she finally gave up all effort and sat still in the alley. The situation was perplexing and I was about to replace the small disc by one still smaller when I thought of offering some of the fragments of roast lamb that I had brought along that morning. The instant the lamb was unwrapped Dolcy became active and could hardly be kept inside her cage. When given a chance she went directly to the proper place and continued to make correct choices for some time. Such is the direct effect of roast lamb on animal intelligence!
The unsatisfactory character of the food stimulus caused Yerkes to resort to punishment for wrong choices instead of rewarding correct choices. He covered the bottom of the discrimination box, in which he was testing dancing-mice, with small copper wire, and when the animal went the wrong way it was given a slight shock. It has been found that animals under these circumstances learn much more quickly than when prompted by hunger alone.
The results of ten years’ work in the experimental study of the animal mind may be stated as a widening of our knowledge in two directions. We know far more than we ever did about the sensory experiences of animals, and we know far more than before about their methods of learning, with all the collateral processes that go along with learning. In the former field the lower animals have been more widely explored; in the latter the higher animals have received most attention.
Our knowledge of the sensory experiences of animals has developed both by way of limitation and by way of expansion. We cannot conclude from the mere presence of a sense-organ that the animal sees, hears, smells, or tastes in the same way as other animals having these same organs, and certainly not as the human being does. Research has also revealed the presence of sensory reactions in animals, as in the amoeba, in which there are no specific sense-organs. In other animals there have been discovered sensory reactions to which there is nothing analogous in the human species, indicating the presence of an entirely new sense.
A good example of how experimental work alters our understanding of these matters is Watson’s investigation on the white rat. The normal man, seeing the rat endowed with all the senseorgans of man, concludes that they rely upon their sense-organs in a way similar to the ways of man. Experimental evidence points in a contrary direction. Watson worked with rats that were blind, rats that were deaf, rats that could not smell, rats whose vibrissœ had been cut off and the soles of whose feet had been anaesthetized. Not the absence of vision nor of hearing nor of smell nor of tactual sensation seemed to affect the rat’s ability to learn a labyrinth, or to run a maze which had been learned before the loss of the sense in question. The animals seemed guided by some sense whose organ is not apparent to normal observation, and Watson concludes that the process of correct turning in the maze is not controlled by extra-organic sensations, but by something that goes on in the body of the animal during the experience of learning: muscular sensations, changes in the bodily organs due to upright position, bodily balance, freedom of movement, etc.
Unexpected results of this sort have made students experimenting on the animal mind hesitate to accept popular beliefs about animal senses as true until the supposed facts have been given experimental verification. The work of Pawlow and his students indicates that the Russian wolf hound is color-blind. This raises a very pertinent question in regard to all other species of dog. On the other hand, the nocturnal raccoon, to which the color-sense must certainly be of much less value than to the dog, discriminates colors with considerable accuracy. Sparrows, cowbirds, and monkeys seem to fall in with the raccoon in this matter of color-vision, as do also certain kinds of fish and amphibians. The frog seems able to recognize the light waves, not only through the eye, but also by means of the skin. In many of the experiments, however, the apparatus used has not been such as certainly to separate the color-stimulus from the stimulus to the sensations of light and dark. It may be, therefore, that what has seemed in some animals a response to color is nothing more than the brightness of vision of colorblind human beings. The question has been raised for the whole animal world; from the standpoint of science we are on the verge of an undiscovered country, and we are not likely to accept the claims of mere casual observers or to rest content in our present ignorance.
With the other senses the case is somewhat the same. Yerkes found that the dancing-mouse is deaf, but birds, dogs, and raccoons are capable of fine discriminations of sound, while crayfish hear but little, if at all. The earthworm has a chemical sense analogous to the sense of taste and smell in the human family; the ants detect various kinds of odor with the several joints of their antennae, and Jennings has shown that the naked bit of protoplasm called amœba reacts to all classes of stimuli to which higher animals react. But what we know is small in view of the great unproved riches of animal sensations that lie before us.
In the field of learning the first and most important result of the critical and experimental work on the higher animals was to reveal the general poverty of these animals in higher intellectual processes. Cats, dogs, chickens, and monkeys do not reason out things, they do not learn by being put through acts, nor do they learn to the extent it is generally supposed they do by imitation. They learn new acts by accidentally happening upon modes of behavior that bring them pleasurable experiences. The pleasure of these accidental happenings stamps in an association between a sense-impression and the successful act, and thus the act tends to be repeated. This explanation calls for no ideas, no memories, no images even, apart from immediate sense-impressions. This explanation demands, of course, that the animal be endowed with the tendency to make movements of various sorts, the most stereotyped ones of which may be called instincts. Successive experimentation has shown that this form of learning is widespread. White rats, rhesus monkeys, crayfish, sparrows, and raccoons, all modify their inherited tendencies to action in the same way.
That the experimentalists, in the enthusiasm of their new discoveries, swept away too much of the popular faith in the mental powers of animals is evidenced by more recent studies on cats, monkeys, and raccoons. Imitation, and imitation of an advanced type, does play some part in the learning of cats and monkeys. It is the writer’s opinion that further refinement of experimental procedure and a more comprehensive study of individual species will be decidedly to the animals’ advantage. First attempts at experimentation were crude, and the animals’ reputation for intelligence suffered. One cannot set problems for animallearning that will adequately lay bare the animal’s possibilities without an extended analytic study of the free movements of the animal in question. That many of the conditions of early experimentation fell short in one or another respect is no more than was to be expected in the first incursions into a new field. A juster appreciation of animal intelligence is bound to come when laboratory men have had the time and insight to invent tests that will more adequately unravel the intricacies of animal behavior. It is the spirit, however, of current investigation to proceed with extreme caution, to allow to the animal mind no attribution of intelligence, the possession of which has not been demonstrated by rigidly-controlled experimentation. No present-day laboratory man will ever give credence to the once common absurdities of mere observation.
In the field of learning there has been an interesting though indirect confirmation of the continuity of the Darwinian hypothesis of mental life throughout the animal scale up to and including man. The impassable gulf between man and the beasts is an illusion, as Darwin thought it was. The confirmation of the doctrine, however, has not come about by demonstrating the presence in animals of clear-cut intellectual processes, but by showing that the sort of learning that does hold in animals is the very root of all that is developed in the mind of man. The lowest man, of course, rises above the highest animal in many ways, but the highest man has as the central core of all his mental and bodily life the fund of habits that he first learned in the trial-and-error fashion of the world below him. The modern psychology of human thinking gives no encouragement to the older belief that a man’s thinking processes go on after the fashion of Aristotelian syllogisms. The normal man is not gifted with any such clear-cut manipulations as was at one time supposed. His mind is a more or less confused mass of sensations of sight, memories of sound, imagined odors, perceived forms, impulses to move, frights, hopes, tastes of food, feelings of objects without and bodily changes within, pleasures and pains, hereditary tendencies to action, and the images of longed-for goods, the whole mass moving restlessly in the individual’s effort to live well in the midst of a changing environment; moving now slowly against stubborn difficulties, and now shooting forward with electric rapidity; moving now all together, as a mass, and now the larger part lying inhibited, while a fraction shoots off at the prompting of temporary attention; nothing in it certain but its imprisonment within the walls of sense, and its slavish conformity to habit; all its entrance into undiscovered country, which alone deserves the right to be called thinking, determined by its past history and its present interest, foredoomed to ceaseless activity by the imperative demands of breathing, of eating, of thinking, of loving, of hoping. The pure thought of the older metaphysical psychology is not the sort of thing that modern research brings to light. The concrete thinking of our work-a-day mind is something less pure, a little less ethereal, something more nearly akin to the animal from which we sprang. The same story repeats itself in every level of the race, — many trials, many errors, and possibly one happy accidental success, which, becoming stamped in by the pleasure of the result, constitutes learning.
I am aware that the reader is ready to ask the value of all this anxious work, for the experimental study of animal behavior is now a serious enterprise calling for the devotion of trained men and the expenditure of large sums of money. If the movement will successfully cope with the problems before it, there will be three rewards, any one of which is a sufficient justification.
First of all, there is the satisfaction of the great human instinct of curiosity. It is this instinct that makes the naturelover observe the facts of the world about him. It is this that has brought all our pure science into being, and in this body of science the study of animal behavior seeks to find a place.
Secondly, if this study fulfills adequately the motives that brought it into being, it will reflect valuable knowledge on both biology and psychology. The results already attained justify the devotion which the study has received, and the further scientific conquest of the field is bound to repay the older sciences for their labor.
Finally, it is the hope of at least certain investigators that the new science may do something toward putting education on a scientific foundation. It cannot, of course, perform the whole task, but if with our animals we can work out the laws of the modification of behavior in living organisms, that is, discover their methods of learning, there is no doubt that we shall contribute thereby to the fields of schoolorganization and school-instruction. Just as the bacteriologist and the pharmacologist work out their facts by experimenting on animals and then apply the results to the care and the cure of the human body, so the animal psychologist may in the future become a most important ally of the educator.