Upon the Threshold of the Mind

I

‘So far as an exact science of animal conduct is possible, the experimentalist has the advantage over the free observer.’ This admission from Mr. Burroughs in an Atlantic article, last year,¹ is generous in view of the contempt in which he holds the effort to arrive at an exact understanding of the animal mind. ‘Anything like an exact science of animal behavior is, it seems to me, as impossible in the laboratory as out of it ’; and he begrudges ‘the time spent in learning’ what seem to him the trivial bits of detail about animals.

So do I, and so does every experimentalist. How we should like to plunge into the complex mental processes and say, ‘This is characteristic of the dog, and this is true of the cat. Here is an animal that is wholly controlled by tropisms, and here is one that is the victim of its instincts, and here is another that manifests intelligence.’ Take the question of the evolution of mentality: how fine it would be if we could say that here at one end of the scale is man, and there at the other is the single-celled bit: of protoplasmic substance called amœba, and then dispose the myriad forms of animal life each in its proper place in the ascending scale. Here belongs the dog and there the earthworm, and there the oyster and here the ape. But however much we may believe in the evolution of mind, the materials for the definite placing of any one of the more complex animals in such an evolutionary scale are absurdly insufficient.

One thing seems sure as the result of experimental studies, and that is that the mind did not evolve as a whole. This is most clearly seen in the case of the senses. If one were to arrange the animal kingdom along an ascending scale as regards the sense of smell, man would probably fall very far below many of the higher mammals and possibly some of the birds. Another scale based on the development of hearing would probably reveal another order, and the sense of sight would show a still different arrangement. Possibly only in the latter would man stand at the top, and some of the birds, such as the hawks, might displace him even there. He doubtless would show the keenest sense of color, but hardly of movement. The dog might excel him in the case of smell, and the house-fly in the case of temperature. The cat may be more sensitive to touch and the bee to direction. Where each of a dozen different animals would fall in such an evolutionary programme cannot be determined by ordinary observation. Only the most detailed and persistent experiment can settle any single point finally.

Nor is this irregularity of development confined to the several senses. There are many different stages in the evolution of the memory. The singlecelled animal has memory, and so has man. Both have good forgettories, however, and the balance between remembering and forgetting would by no means place the several animals at the same level of development at which the acuteness of their sense-powers would place them. In other words, the ability to remember has an evolutionary history of its own, and must be studied on its own account. The same is doubtless true of perception, of the ability to learn, of the evolution of thinking and self-control.

A corollary of these facts is that the animals would not fall to the same level in the scale of mental evolution as in the scale of structural evolution. It is generally accepted that monkeys stand nearer man than do the canine family. It is not at all established that their intelligence is of a higher sort than that of dogs and birds. If there is absolute correlation of structure with intelligence it is of structures that have not yet been made out. The fact is that in many cases we know more about the behavior of animals than we do about the minute structures upon which that behavior is supposed to depend. We must construct the tree of mental evolution out of the material which we can gather In many cases it appears that such a mental tree would cut at right angles the tree of morphological evolution.

II

Apropos of the comparative abilities of animals Dr. Hamilton has recently published a study of trial and error reactions in mammals. He used thirtyseven different animals, distributed among several species as follows: eight normal human beings, men, boys and girls; two defective human beings; five monkeys; sixteen dogs; five cats; and one horse. In the experiment the subject passed through an entrance door into a room from which it could escape by one of four doors at the opposite end of the compartment. All that was necessary was to push against the correct door. Before the actual tests were begun the animals were given thorough familiarity with the inside of the room by being fed there, and by a series of steps they were trained to open the exit doors and pass out. ‘The subject was considered “ trained ” as soon as he had learned to seek the exit doors for escape from the apparatus and to push against them without hesitation.' The ‘formal trials ’ were then begun. Three of the exit doors were locked on the outside by the experimenter without the subject’s knowledge. On the inside the doors all looked the same but only one of them could be opened. When the animal escaped he was rewarded with food. The unlocked door was then locked, another was unlocked and the animal was put back for a second test. Thus the unlocked door of the first trial was one of the locked doors of the second trial, and the animal could escape on this second trial only by pushing against one of the doors which had been locked on the first trial. Which door it might be there was nothing before him to indicate. To discover how the animal would go about finding this unlocked door was the point of the experiment. In the course of a hundred trials, which were given at the rate of ten a day for ten days, each door was unlocked twentyfive per cent of the time and locked seventy-five per cent of the time. No door was unlocked twice in succession.

The behavior of the animals was carefully observed and tabulated. On the basis of what they actually did, Dr. Hamilton describes five types of behavior which he designates A, B, C, D, and E respectively. In type A, the animal makes the fewest useless movements. For instance, he does not try the door through which he has just escaped, this door always having been locked in the succeeding trial. Each of the other doors he tries once each, and in the very nature of the situation he will find one of them unlocked. Each of the thirty-seven subjects behaved a part of the time in the type A manner. ‘This,’ says the writer, ‘is the most adequate possible type of classified reaction,’ and he refers to it as the ‘rational inference tendency.

As regards their ability to react in the type A manner, the several species fell into the following order. The figures indicate the percentage of their trials which fell into this highest group. Human, 68.18; monkey, 18.41; dog, 13.35; horse, 8.00; cat, 7.69. These figures, which group all the animals of a species together, do not, however, give a correct understanding of the several individuals within the species; and the apparently clear-cut distinctions between species are shown to be false by the records of individuals. Thus, while the horse ranks above the cats when the latter are taken as a whole, the mature cats rank higher than the horse when they alone are considered. The monkeys as a group exceeded the whole group of dogs by almost three per cent; but the three best dogs, all pups under three months, averaged better than the three best monkeys by the margin of one per cent. The whole group of human beings, including the youngest and defectives, averaged higher than any group of lower animals, and the older human beings, individually, had better records than any single animal. On the other hand the two-year-old boy, whom the experimenter describes as ‘ very quick to form new associations ’ and whose intelligence he would not be likely to underestimate since he was his own son, fell below four of the monkeys and seven of the dogs in his ability to behave in the ‘rational inference manner.’ One three-months-old pup of mixed breed — Boston terrier and English setter — almost doubled the record of the boy as well as that of his own species. If, instead of taking the highest form of behavior as the measure of ability, one takes the poorest form, the E type, then the boy falls below the whole group of monkeys, t he latter having an index of ability almost three times as great, as the boy.

These facts offer rather formidable propositions to the person who believes in a qualitative difference between the animal mind and the human. Superiority appears to lie with the four monkeys and seven dogs as against the two-year-old boy, on the one hand, and with the normal human beings as against the animals on the other. This is small ground upon which to affirm that the human mind is qualitatively superior to the animal. Nor is there much comfort in these experiments for the person who holds that animals have an invariable manner of reaction peculiar to their species, while man has the possibility of wide variation from the average of his race. In the eight normal human subjects the greatest variation from the average of the group was 9.6 per cent, whereas one dog varied 12.46 per cent from the average of the dogs, and another dog varied 6.65 per cent; one monkey varied 4.67 per cent from the average of the monkeys, and the variation of one cat was 4.55 per cent. The difference between the poorest dog and the best dog was 21.04 per cent, while the widest difference in the normal adult human group was only 16.48 per cent. Such figures as these show that animals present great variations of behavior within a narrowly defined group, and that it is never safe to conclude that the animal which you meet casually, and observe without previous study and selection, is a representative member of his species. It is always dangerous to conclude from the behavior of a single animal, even though it has been carefully observed, that such behavior is the fixed type for that ’level’ of animal life.

Such a study as this does not of course show that the animal mind is on a par with the human mind, nor is one really interested in showing that. It was only in the eight normal human adults and in one defective boy that a majority of the efforts showed the highest form of reaction. Of this same group, however, practically 20 per cent of the reactions were of type B, the level of behavior which holds a large percentage of monkey and dog reactions. In fact, one finds that the behavior of both animal and human subjects is distributed throughout the entire series of classified reactions. While none of the animals exhibit a large percentage of type A reactions, neither do any of the human beings confine their behavior to that mode. There is clearly much animal behavior in man, and some human behavior in dogs and monkeys. To unravel the complexities in each case is the task of the experimental method.

Everybody will admit that there is nothing final about Dr. Hamilton’s results. Methods and results are always subject to the criticism of later investigation. Any student can suggest a multitude of questions that remain unanswered. Dr. Hamilton himself points out certain limitations of his methods, and is only in part responsible for the inferences that I have seen fit to make here. The next step is to refine the method, to eliminate errors, to extend the tests to a much larger number of subjects within the species and to other species, to collect such data in quantity, and then let the conclusions be whatever they turn out to be. The important thing just now is that the investigator has struck his pick into an important problem with some show of success. If he has merely widened a crevice so that the next blow will sink deeper, that is something.

III

Mr. Burroughs’s denial of an exact science of the animal mind amounts to a denial of all real science whatsoever. ‘The uncontrolled animal going its free picturesque ways, solving its life problems as they come to it in the revolving seasons, using such mind as it has, without constraint or arbitrary direction, threading only the labyrinth which nature prepares for it, stimulated only by the sights and sounds and the odors of its natural habitat, perplexed by no puzzles but how to get its food, avoid its enemies, rear its young, hide its nest, or den, and get out of life what there is in it — how much more engaging and stimulating an animal under such conditions than the same creature being put through its paces under the controlled conditions of the laboratory.’ How fine! And how true also of the human animal! What a joy to see him in action, solving his life problems as they come to him in the revolving seasons, using such mind as he has, without constraint or arbitrary direction, discriminating only between the colors of nature, attending only to the sounds of the woods and the streets and the voices of his fellows, remembering only the days of his childhood and the experiences which have been vital to him, fatiguing himself only with the joy of living, dreaming only the hopes of his youth, flinging himself in the front of the conflict, marshaling all his powers in an efficient commercial life, controlling his body in behalf of his neighbors’ rights, loving with a beautiful passion, fearing moral and physical dangers, blazing in anger at injustice and following the instinct of curiosity into the woods and the laboratory — how much more engaging and stimulating a man under such conditions than the same creature testing his memory with a series of nonsense syllables, or studying his peripheral color vision with a perimeter, or testing his powers of action by a chronoscope, or exhibiting an emotion in such a manner as to show its influence on the blood pressure, or voluntarily submitting to strenuous labor in order to discover the laws of fatigue!

How the critics have laughed at the brass instruments and scorned the technique of the psychological laboratory! Even so recent and pretentious an authority as the new Encyclopœdia Britannica speaks of American psychology as ‘largely a matter of experiment and apparatus.’ It omits altogether to give us an article on either experimental or comparative psychology, and fails to class some of the bestknown students of mental life under the title of psychologist at all. We have been repeatedly told that experimental psychology is extremely limited in its possibilities. It gives us nothing but a psychology of sensation; the complex mental processes lie beyond its reach. Yet the simple fact remains that the last thirty years have been the most fruitfully productive years thus far in the history of psychological thought. The new psychology, to quote Professor Royce, has been one of the ‘two notable movements of world-wide significance’ which have characterized our time. This has been also the period of the development of experimental psychology, and no one doubts that the general movement, broader of course than ‘the special scientific study of mind by the modern methods,’is under important obligations to the psychological laboratory. And just now, when we have been told that the complex mental processes are beyond the reach of our methods, there is arising an experimental psychology of thinking and of will, a rebuke to the faithless critics of the experimental method.

Why should we not say the same to physics and chemistry, and to every other experimental science? You are interested in sound. Then let us go to the symphony concert, to the seashore or the summer woods. How thin and juiceless seems the single sound wave produced by a single electric spark under the controlled conditions of the laboratory! You want to know about light. There is the noon-day sun, the iridescent bow, the sun-splashed western sky. Why not the October woods and the glittering sheen of hill-set lakes, or the early peeping green of flowers and grass above the russet covering of the fields and woods? How stupid in contrast the attempt to determine the relative intensities of two tallow candles by the complicated apparatus and formulæ of the laboratory! How wasted seem the years spent making a machine to grind a micrometer screw to be used in grinding out a diffraction grating to be used in measuring the waves of light emitted from a sodium vapor flame! You want to know about electricity. Then take the storm when lightning tears the sky, crashes through the trees and licks up the foaming waves. In that you get the thrill beside which all the amperes, volts, and ohms of laboratory investigation seem dry rot, mere mental punk. Why not the grateful taste of luscious fruit, the odors of the fields and woods, the glitter of the polished gem, rather than the crucible, the test tube, and the chemist’s balance?

The complexity of nature? We cannot stop at that, even though it be the complexity of human and animal nature. It is just because of natural complexities that science is called into being. Confronted with the multitudinous shifting forms and processes in the world about him, man tries to understand and to control. This is far easier when he can simplify and reduce to laws; and in the world-wide, agelong struggle the experimental laboratory has come to be the chief instrument of progress. Physics, chemistry, astronomy, biology with all its subdivisions, psychology, education, government, war, —one almost says, marriage and religion, with eugenics and social service, — have accepted the experimental method in one or another form, and it is probably fair to say that the most settled faith our people now have is faith in the results and methods of experimental science. There is hardly a religious tenet that is accepted with a more unquestioning faith than are the most fundamental postulates of science. The law of gravitation, the atomic theory, the evolution of life, these have, at least, as general acceptance as the doctrine of personal immortality or the authority of the church.

On the tangible results of experimental science our whole present social organization is absolutely dependent. One cannot realize what it would mean to drop out of our industrial life the products of the laboratory; $9,000,000,000 invested in the results of one man’s work. We risk our lives with the machines science has given us; we trust our fortunes to a few wires and brass devices. Science is the servant who feeds and waters our urban populations; we stave off the spectre of death for days and years as the result of experimental science. We trust to science to cut out our viscera, to repair our brains, and to relieve our pain. It welds together the fragments of a divided personality, and lays down the laws for the propagation of a superior race of human beings. With the gifts of science we can now provide ourselves with the needs of existence and yet have leisure for the cultivation of life. As Professor Cattell has pointed out, we can now have leisure for all our people to a degree that has not hitherto been possible in the history of the world. Science thus becomes the real condition upon which we found our new democracy. No longer must culture be built upon the institution of human slavery. Our modern slave is science; from it we have the guarantee of that leisure and opportunity that will make possible a cultivated race.

Suppose the early heroes of the cult had said, ‘You can’t do it. Nature is too complex. You never can understand her as she is. If you take her into the experimental laboratory you will distort her so that you will have left but a phantom of the reality. You must stay in the open and wait until she reveals herself in her own capricious way.’ Suppose they had said that?

IV

It is fair to ask, however, whether the achievements of the animal laboratory are so meagre as they sometimes seem. The experimental study of animal behavior has been a matter of concern in this country, to say nothing of Europe, for a dozen years, the merest infancy in the life of a science. Think how our knowledge of animal structure had to wait on the perfection of the microscope as an instrument of morphological research! Professor Michaelson has shown that the development of the science of spectroscopy has been paralleled at every step by advances in astronomy, the former being the instrument of the latter. A decade ago the new interest in the behavior of animals had no such elaborate instruments of research at its command. The technique had to be created outright. It is no disparagement of the results already accomplished to say that most of it yet remains to be worked out. For very few of our animal problems are there as yet adequate methods of investigation. The immediate concern of every student of animal psychology must be the discovery and the perfection of methods. The finished science is a hope and not an accomplished fact.

Is it not something, however, that in the comparatively short time a movement should have been created? That there are many signs of such a movement no one acquainted with the facts will deny. Experimental data on the behavior of animals have been forthcoming from a score of psychological and zoölogical laboratories of the leading universities of America and Europe. Half a hundred well trained men and women in this country alone, together with their students, are more or less actively interested in the problems of animal behavior. Loeb, Thorndike, Jennings, Yerkes, Parker, Sanford, Wheeler, Watson, Porter, Washburn, — these are a few of the names which have been most prominent.

More than a thousand monographs now confront the beginner who would master the work accomplished. A halfdozen books intended to condense and interpret this scattered literature have appeared, and more are promised. The publications for the year 1910 alone number one hundred and fifty. That the complexities of nature do not repel investigation is evidenced by the fact that twenty-five of these studies were devoted to vertebrates, specifically to fish, frogs, reptiles, birds, dogs, cats, mice, monkeys, and apes. Nor were the problems the merely simple ones. Color and olfactory discrimination in the fish, distance orientation in birds, imitation in English sparrows and other birds, the instinct of kittens to kill mice, the speed of migration of the salmon in the Columbia River, instinct and intelligence in the dog, and the mental processes of the rhesus monkey, were among the things inquired after. The year 1911 moreover witnessed the establishment of the Journal of Animal Behavior, and of the Animal Behavior Monograph Series.

One of the most important recent advances in animal psychology is the attempt to study the subject of instinct experimentally. For many years it has been the custom to think of instinct and intelligence as set over against each other. The former represents the inherited reactions of the animal; the latter signifies those acts which the animal learns in the course of its individual life and its power to learn in this fashion. Instinctive acts are ‘perfect the very first time,’ while intelligent acts are slowly acquired. In so far as an animal is dependent on its instincts it is the victim of its ancestry and of its immediate environment. In so far as it is intelligent it can adapt its environment to itself, can rise above its immediate surroundings and act in the light of a distant end to be accomplished. Instinct is blind as to the outcome of its action; intelligence foresees and modifies its behavior in the light of its foresight. The supreme example of instinct has been the wasp going through an elaborate course of action to provide food for its yet unhatched offspring, seeking out a particular species of caterpillar, stinging every segment just enough to stun the animal but not kill it, and then depositing the worm at the bottom of its newly made nest where it has laid its egg, covering up the cell and then leaving, never to see the larvæ nor to live to do the same act another season. The supremely intelligent animal is man possessing the earth by the genius of his intellect.

Now no one doubts that the antithesis between instinct and intelligence points to very important phenomena in life. Just as certainly no one can read the literature of instinct without feeling that much of it is based on verbal and logical tangles rather than observed facts; and the more the latter have come to light the more certain has it become that current distinctions were unsatisfactory and barren of further insight into our problems. Evidence of the generally unsatisfactory nature of our understanding of the relation of instinct and intelligence is found in the several recent attempts to redefine these terms. Within a short time the effort has been made by the British Psychological Association, the Aristotelian Society, and the American Psychological Association, but with no very astonishing success in any case. The last-named body conducted a symposium at its Washington meeting on the relation of instinct and intelligence, and although it had at its hand the results of the British discussion, it is safe to say that not many persons present were gratified with the result. The difficulty seems to lie in the lack of concrete information regarding the part that instinct plays in the animal world; and it seems perfectly evident that further discussion of the subject will be but the mere juggling of words until we have a widespread accession to our knowledge of concrete facts.

In this unsatisfactory situation it is encouraging to know that the laboratory can do something to help us. Instincts are complex forms of behavior, and in many cases we have been merely throwing words at these complexes rather than analyzing them into their simpler elements. It is one thing to say that kittens have the instinct to feed, that birds have the instinct to fly, that children have the instinct to construct or to destroy. It is another thing to break these complex bits of animal behavior up into their simpler component bits of behavior and see the exact relations of these simpler elements to each other. The former method serves for a rough classification; only the latter will enable us to get at the roots of life-processes, to discover whether the ultimate explanation of animal conduct is to be in terms of the physical sciences, or whether it is explainable in any terms.

Several experimental attempts have been made to study animal instincts, but the one with most promise is the work of Professor Breed, who has studied the feeding instincts of newly born chicks. The pecking instinct in chicks is a complex bit of behavior made up of at least three elements, the striking, the seizing, and the swallowing reactions. By actual measurements Dr. Breed has shown not only that the pecking instinct is not perfect at birth, but that the three factors vary more or less, independently of each other. He shows that the pecking instinct improves with practice and with age, and that each of the three factors improves independently of the other two, at least to some extent. That is to say, the ability to seize the food does not improve exactly in coördination with the ability to swallow. His studies further show that the instinct varied from chick to chick and from group to group. And this with an animal where one might have expected to find, if anywhere, the supposedly invariable behavior. The important thing about this work, of course, is not what it shows regarding the instinct of the chick. The critics will doubtless say that they knew all this before. Far more important is it that an instinct has actually been submitted to measurement. If the evident indication of the work is followed there will be many studies of the instincts of the higher animals, and we shall be able to make mathematical statements about them. While it is exceedingly arduous, it is perfectly possible to make quantitative measurements of some of the instincts of the dog or the cat or practically any other animal which can be studied under controlled conditions.

Enthusiasm for the experimental method should not of course blind any one to such facts as appear without experimental search. There is no virtue in a fact merely because it has come through a laboratory, nor is a fact any the less a fact because it has been picked up in the woods. Nor should there finally be any opposition between the facts that have been gathered from the two sources. Possibly when we have pushed our work far enough we shall see that the two methods supplement each other. Certain it is that we can never dispense with the study of animals in their native environment if we would really understand them. Personally I have always thought that we should get our experimental problems from the observation of animals in their free life. Only by such observation can one really penetrate the strategic problems and invent the adequate experimental methods. Without a doubt we blunder in trying to apply the methods of human psychology directly to the animal mind, or in applying the method that has proved efficient in the case of one animal to an individual of a different species. For the correction of such blunders, and for the suggestion of important problems, we can never know too much of animals in the wild.

On the other hand, there is no longer any reasonable doubt that such knowledge as we get of free animals can be fruitfully supplemented by the exact information that we can get about animals in the laboratory. The elemental aspects of behavior that are covered up by the free-moving bird or mammal can be brought to light by the controlled conditions of experiment. How voluminously such investigations will add to our knowledge of the animal mind we have as yet only dimly realized. Animal senses, animal memory, animal instincts, animal learning, animal thinking, all have surrendered some of their secrets to the experimental method. What has already been achieved gives us sufficient reason to think that much more can be gained by the extension of the laboratory into this realm of nature; that, in fact, a new science will ultimately replace our at present inadequate ideas of the animal mind.