Popular Science Monthly/Volume 59/July 1901/The Intelligence of Monkeys
| THE INTELLIGENCE OF MONKEYS. |
By professor EDWARD L. THORNDIKE,
TEACHERS COLLEGE, COLUMBIA UNIVERSITY.
A GOOD test of the intelligence of any animal is its ability to learn to do a thing by being shown it or by being put through the requisite movements. Human adults would learn readily in either of these ways, because we thus get ideas of what to do and how to do it and modify our actions in accordance with these ideas. If the reader had never seen a glass or a faucet, he would nevertheless learn how to get a drink by turning a faucet and holding the glass beneath it, if he saw some one else do it, or if some one took his hands and put them through the movements. The intelligence required in such cases is not of a very advanced sort; it is not the power of abstract reasoning or of seeing the relationships of facts, but is simply the capacity to have ideas and to progress from the idea of doing a thing to the act itself.
A study which I made four years ago of the mental powers of dogs, cats and chicks showed that these animals did not, at least not habitually, learn from this sort of tuition. They learned only in the following manner: If in any situation their own impulses led them to do something which brought desirable results, they would, when put in that situation again, do that particular thing rather than anything else. If for instance a kitten was shut up in a box from which it could escape only by turning around a button which held the door, it would claw and bite and pull and squeeze at random as its instinctive impulses led it to do. If by chance it made a pull at the button and so secured freedom and food, it would, the next time it was put in that box, be likely to make that particular movement earlier than in its first trial. After enough trials it would pull the button around as soon as put into the box. It had learned by the selection of one of its own impulses. If you showed it how to get out by putting it in the box and turning the button for it, thus letting it out, it learned no more quickly than when by itself. So also if you took its paw and with it pulled the button round. And any acts which it failed to learn by means of the selection of chance successes from its own impulsive activities, it could never be taught by example or by being put through the movements, scores of times.
During 1900 I was engaged in investigating the mental capacities of monkeys and included in my experiments a number bearing upon this question. The monkeys are quicker to learn and learn more things than do the dogs and cats. Considering this and also their close relationship physically to man, it seemed of special interest to discover whether they could learn from example or from being put through certain movements, whether they would manifest the capacity so evident in man and so lacking in the lower animals in general.
I had one monkey for over a year and two others for about four months. All three were of the genus Cebus. The general method of the experiments was as follows: Boxes about a foot square were made with small doors held by all sorts of contrivances. For instance, one was held by a hook, another by a bar, another by a bolt, another by a wire fastened firmly at one side and wound round a nail at the other. A bit of food was put inside such a box and the door of the box left open. The box was then put inside one of the large cages containing a monkey. It would come down, reach into the box and get the food. After this had been repeated a few times the box would be put into the cage, with its door shut. The monkey would try to get in as before. He might chance to operate the simple mechanism that held the door. If, however, he did not succeed of his own impulsive activity I would show him or put him through the movement a few times and then leave him to himself again to see if he had profited by the tuition.
The general result was that they did not profit by the tuition, that they did not gain and use ideas of how to open the doors, but learned only by a process of selection from their own impulses. The meaning and value of this general fact will ajpear in the details of the experiments.
In order that such experiments shall be valid tests of the workings of an animal's mind it is necessary that he surely desire to get into the box, that he be not disturbed by the surroundings in any way that will alter his mental efficiency, and that the experimenter be able to handle him easily without frightening him or taking his attention away from the box. In all cases it is further necessary to make sure that the monkey sees you perform the acts you expect him to imitate, and sees and feels himself make the movements you put him through. These desiderata were obtained by testing the monkeys when hungry and using bits of food of which they were especially fond as the attraction; by experimenting with them after they were quite used to their habitat and to my presence; by getting them into the habit of coming to me and enjoying being handled, having their paws taken, etc.; by showing them the act or putting them through it only when they were attending to the box.
A sample of one of the experiments on the influence of example is the following: The box used was arranged so that the door opened when a brass lover was depressed about an eighth of an inch. The monkey could reach this lever by putting his hand through a hole about an inch square at the right side of the front of the box. The door was at the left side of the front. On January 12th I put this box in No. 3's cage, the door of the box being open. I put a bit of food in the box. No. 3 reached in and took it. This was repeated three times. I then put in a bit of food and closed the door. No. 3 pulled and bit the box, turned it over, fingered and bit at the hole where the lever was, but did not succeed in getting the door open. After ten minutes I took the box out. The monkey having failed by his own impulsive efforts to depress the lever, I began the tuition. I took No. 3 out and let him sit on my knees (I sitting on the floor with the box in front of us). I would then put my hand out toward the box and when he was looking at it would insert my finger and depress the lever with as evident a movement as I could. The door, of course, opened, and No. 3 put his arm in and took the bit of food. I then put in another, closed the door and depressed the lever as before. No. 3 watched my hand pretty constantly, as all his experiences with me had made such watching profitable. After ten such trials he was put back in the cage and the box put in with a large piece of food in it and its door closed. No. 3 failed in the course of five minutes to get the door open. His behavior was just the same as it had been before he had seen me open the door ten times. He had not profited at all by my example. Later I showed him 15 times more and then tried him by himself. He failed as before.
The two monkeys. No. 1 and No. 3, were given a number of such chances to learn acts from seeing me. Other, boxes were used, the doors of which could be opened by pulling up a bolt, pulling out a plug, pushing a bar back into a slot, unwinding a wire and pulling a loop off from a nail. I had also certain pieces of apparatus arranged which would throw a bit of food down a chute into the cage when some simple mechanisms were operated; when for instance a nail was pulled out of a hole or a loop pulled off a nail or a bar pushed in. These could be set up outside the cages so that the monkeys could reach them through the wire netting and could easily see me operate them.
No. 1 had in all eight chances to learn from seeing me. In seven of the cases he failed utterly after seeing me operate the mechanisms 21, 5, 10, 4, 15, 40 and 15 times respectively. He did succeed in one case where the act required was to pull a wire loop off a nail. This must, I think, have been an accident. The other monkey failed utterly to learn to do the same thing though he had continued tuition.
No. 3 had seven chances to learn from seeing me. In five out of the seven he failed after seeing me operate the mechanisms 40, 30, 25, 5 and 30 times respectively. In the case of the other two, although he succeeded in getting the door open, it was not by doing as I had shown him. I opened a door 25 times by pulling a bolt up, but he opened it by pulling and pushing at the door itself until he worked the bolt up out of place. In the other case I pulled a hook out from a catch but he yanked at the bar to which the hook was attached and so jerked the latter free.
It might be that although the monkeys did not succeed after tuition where they had previously failed, yet they attempted acts which they had not previously attempted. This is not the case, however. There were no signs that the monkeys tried more after tuition to do the things they saw me do than they did before. Their behavior was unmodified by the tuition save that in general they tried less.
It may be objected that the acts I failed to teach the monkeys were not consistent with their make-up, that a monkey might be very intelligent and still not manifest his intelligence by depressing levers, unwinding wires or pulling off loops, that monkeys might be able to learn to do certain things from seeing them done and still be unable to learn the particular acts needed in these experiments. But as a matter of fact, these particular acts were quite natural for the monkeys, quite in accord with their interests and propensities. They learned by the typical animal method acts of the same general sort, e. g., to open boxes and operate the mechanisms throwing food into their cages by pulling bars around, unhooking hooks and pulling at strings. And often the very same act with which I tested one monkey in the experiments just described had been learned by another through the repetition and selection of a chance success. Thus No. 1 learned of himself to unwind a wire though No. 3 failed to do so after seeing me do it 30 times.
The systematic experiments designed to detect the presence of ability to learn from human beings are thus practically unanimous against it. So too was the general behavior of the monkeys, though I do not consider the failure of the animals to imitate common human acts as of much importance save as a rebuke to the story-tellers and casual observers. The following facts are samples: The door of No. 1's cage was closed by an iron hoop with a slit in it through which a staple passed, the door being held by a stick of wood thrust through the staple. No. 1 saw me open the door of his and other cages by taking out sticks hundreds of times, but though he escaped from his cage a dozen times in other ways he never took the stick out and to my knowledge never tried to. I myself and visitors smoked a good deal in the monkeys' presence but a cigar given to them was always treated like anything else.
The following is a sample of the tests of the monkey's ability to learn to do a thing from being made to do it: A box was arranged with its door held closed by a bar of wood held in position in a slot. When it was pushed back an inch and a half further into this slot the door could be opened. It was fastened so that it could not be pulled out from the slot altogether. The only way to get the door open was thus to push or pull the bar back. It worked very easily, a pressure of perhaps 15 grams being sufficient. On January 4, 1901, this box was put in No. I's cage. He failed to get in in 5 minutes, though he was active in trying to get in for about 4 minutes of the time and pulled and pushed the bar a great deal, though up and down and out instead of back. In his aimless pushings and pullings he nearly succeeded. He failed in 5 minutes in a second trial also. I then opened the door of the cage, sat down beside it, held out my hand, and when he came to me took his right paw and with it (he being held in front of the box) pushed the bar back (and pulled the door open in those cases when it did not fall open of itself). He reached in and took the food and went back to the top of his cage and ate it. I put him through the act thus 10 times. I then let him try alone. He failed to get in. In this and the two following days No. 1 was put through the act 80 times and given frequent opportunities to open the box himself. He never derived the slightest profit from the tuition.
No. 1 had eight such tests and No. 3 had six. Their behavior was in some cases ambiguous but the verdict would surely be that they had no general capacity to acquire these simple habits by seeing and feeling themselves make the movements and get food thereby.
The theoretical importance of the failure of the monkeys to learn from example or from being put through movements consists in the testimony it bears to their lack of a general fund of ideas. Adult human beings learn to do things by getting ideas of the circumstances and of the acts required and then proceeding to act upon these ideas. We think of where we are going, and so go; we have an idea of what we wish to do and so do it. Rarely if ever do monkeys learn in this way.
The behavior of the monkeys apart from these specific experiments seemed also to show their inability to acquire and use ideas of objects or acts. In getting them so that they would let themselves be handled, it was of almost no service to take them and feed them while holding them or otherwise make that state pleasant for them. By far the best way is to wait patiently till they do come near, then feed them; wait patiently till they do take hold of your arm, then feed them. If you do take them and hold them partly by force you must feed them only when they are comparatively still. In short in taming them one comes unconsciously to adopt the method of rewarding certain of their impulses rather than certain conditions which might be associated in their minds with ideas, had they such.
Monkey No. 1 apparently enjoyed scratching himself. Among the stimuli which served to set off this act of scratching was the irritation from tobacco smoke. If anyone blew smoke in No. I's face he would blink his eyes and scratch himself, principally in the back. After a time he got in the habit of coming to the front of his cage when anyone was smoking and making such movements and sounds as in his experience had attracted attention and caused the smoker to blow in his face. He was often given a lighted cigar or cigarette to test him for imitation. He formed the habit of rubbing it on his back. After doing so he would scratch himself with great vigor and zest. He came to do this always when the proper object was given him. I have recounted all this to show that the monkey enjoyed scratching himself. Yet he apparently never scratched himself except in response to some sensory stimulus. He did not with all his experiences of scratching ever get the idea of that act and use it to arouse the delightful act. He was apparently incapable of thinking 'scratch' and so doing. Yet the act was quite capable of association with circumstances with which as a matter of hereditary organization it had no connection. For by taking a certain well-defined position in front of his cage and feeding him whenever he did scratch himself I got him to scratch always within a few seconds after I took that position.
The fact that monkeys do not possess the human type of ideas must not be taken as evidence that they are no nearer relatives to us mentally than are the other lower animals. On the contrary they occupy an intermediate position in every main psychological feature between mammals in general and the human species.
The essentials in an inventory of an animal's mental capacities are its sense powers, the kinds of movements it can make and their delicacy, complexity and number, its instincts or the sum of those tendencies to feel and act which it has apart from experience or learning, and its methods of learning or of modifying its behavior to suit the multitudinous circumstances of life. In each of these respects the monkeys show kinship with man.
In point of sense powers they rely little on smell and much on vision. They possess the power of clear, detailed vision which is absent, for instance, in dogs and cats and is so important a possession of man. A monkey will notice a hair on your hand or a pin six feet off. He thus resembles man in what has been universally recognized as the most intellectual of the senses.
In their motor equipment monkeys possess first of all the muscular coordinations necessary to sustain an upright position and consequently the free use of the fore-limbs. The movements of these fore-limbs are more in number and suited to more complex and varied tasks than are those of lower animals. The attractiveness of the monkey cage in a zoological garden is largely due to the similarity of the monkeys' movements and our own. The monkey not only has a body like a man's, but he also uses it like a man.
Our native tendencies are so metamorphosed by the education of a civilized environment that in adult age they seldom appear in recognizable form. But if wc take human beings at from 6 months to 3 years of age or later, we find plenty of traits that appear in the monkeys. In fact the human instinct which is perhaps of prime importance in human mentality, the instinct which perhaps is the real cause of many of our most boasted powers, has its clear prototype and homologue in the monkey. I refer to the instinctive enjoyment of physical and mental activity in general, to the tendencies to act and feel as much as possible, regardless of any ulterior practical considerations, which we sometimes call destructiveness or constructiveness and curiosity.
Even the casual observer, if he has any psychological insight, will be struck by the general, aimless, intrinsically valuable (to the animal's feelings) physical activities of a monkey compared with the specialized, definitely aroused, utilitarian activities of a dog or cat. Watch the latter and he does but few things, does them in response to obvious sense presentations, does them with practical consequences of food, sex-indulgence, preparation for adult battles, etc. If nothing that appeals to his special organization comes up, he does nothing. Watch a monkey and you cannot enumerate the things he does, cannot discover the stimuli to which he reacts, cannot conceive the raison d'être of his pursuits. Everything appeals to him. He likes to be active for the sake of activity.
The observer who has proper opportunities and takes proper pains will find this intrinsic interest to hold true of mental activity as well. No. 1 happened to hit a projecting wire so as to make it vibrate. He repeated this act hundreds of times in the few days following. He could not eat, make love to or get preliminary practise for the serious battles of life out of that sound. But it did give him mental food, mental exercise. Monkeys seem to enjoy strange places; they revel, if I may be permitted an anthropomorphism, in novel objects. They like to have feelings as they do to make movements. The fact of mental life is to them its own reward.
Finally in their method of learning, although monkeys do not reach the human stage of a rich life of ideas, yet they carry the animal method of learning by the selection of impulses and association of them with different sense impressions, to a point beyond that reached by any other of the lower animals. In this, too,they resemble man; for he differs from the lower animals not only in the possession of a new sort of intelligence but also in the tremendous extension of that sort which he has in common with them. A fish learns slowly a few simple habits. Man learns quickly an infinitude of habits that may be highly complex. Dogs and cats learn more than the fish, while monkeys learn more than they. In the number of things he learns, the complex habits he can form, the variety of lines along which he can learn them, and in their permanence when once formed, the monkey justifies his inclusion with man in a separate mental genus.