Animal Behaviour

I.—The Nature of Intelligent Behaviour

Such an animal as a newly hatched bird or an insect just set free from the chrysalis is a going concern, a living creature. It is the bearer of wonderfully complex automatic machinery, capable, under the initiating influence of stimuli, of performing instinctive acts. But if this were all we should have no more than a cunningly wrought and self-developing automatic machine. What the creature does instinctively at first it would do always, perhaps a little more smoothly as the organic mechanism settled down to its work—just as a steam-engine goes more smoothly when it has been running for a while; but otherwise the action would continue unchanged. Instinctive behaviour would remain unmodified throughout life. The chick, however, or the imago insect is something more than this. It affords evidence of the accommodation of behaviour to varying circumstances. It so acts as to lead us to infer that there are centres of intelligent control through the action of which the automatic behaviour can be modified in accordance with the results of experience. When, for example, a young chick walks towards and pecks at a ladybird, the like of which he has never before seen, the behaviour may be purely instinctive; and so, too, when he similarly seizes a wasp-larva. Even when he rejects the ladybird or swallows the larva, this may be directly due to unpleasant stimulation in the one case, and pleasant in the other. But when, after a few trials, the chick leaves ladybirds unmolested while he seizes wasp-larvæ with increased energy, he affords evidence of selection based on individual experience. And such selection implies intelligence in almost its simplest expression. We may say, therefore, that, whereas instinctive behaviour is prior to individual experience, intelligent behaviour is the outcome and product of such experience. This distinction is presumably clear enough; and it is one that is based on the facts of observation. But we must not fail to notice that, though the logical distinction is quite clear, the acquired modifications of behaviour, which we speak of as intelligent, presuppose congenital modes of response which are guided to finer issues. We may say, then, that where these congenital modes of response take the form of instinctive behaviour, there is supplied a general plan of action which intelligence particularizes in such a manner as to produce accommodation to the conditions of existence.

We have already frankly admitted that, in the present state of our knowledge, we do not know with any definiteness how intelligent modification of behaviour is effected. But it seems probable that from all parts of the automatically working organic machine messages come in to the centres of conscious control, and that in accordance with the net result of all these messages, and the past experience which they recall, other messages go out to the automatic centres, and, by checking their action here and enforcing it there, give new direction to the resulting behaviour. If this be so, the consciousness associated with the control-centres is like the person who sits in a central office and guides the working of some organized system in accordance with the information he is constantly receiving; who sends messages to check activity in certain directions and to render it more efficient and vigorous in others.

It may be said, however, that intelligent guidance is, at any rate in such simple cases as the selection of a palatable grub and the rejection of a nauseous ladybird, itself determined by instinctive likes and dislikes. All young chicks apparently find wasp-larvæ palatable and ladybirds the reverse; and this is just as much the outcome of heredity as the instinctive act of pecking. Since, therefore, heredity determines what shall be selected and what rejected—since the likes and dislikes are themselves instinctive—any essential difference between congenital and acquired behaviour seems to be evanescent.

Now, if we apply to the affective qualities of mental states—the pleasurable tone or its opposite which characterize such states—the term “instinctive,” we do so in reference to the broader psychological conception of instinct, rather than in accordance with the narrower biological acceptation of the term. For the likes and dislikes constitute part of the conditions under which the behaviour occurs, and not elements in the co-ordinated response as such. Hence it is preferable to apply to these hereditary qualities the term innate, rather than the term instinctive. But, waiving this distinction, it is true that such pleasant or unpleasant qualities of the sensory results of stimuli are part of the animal’s hereditary outfit, and are not acquired in the course of individual experience. What, then, is acquired? What part does experience play in the development of intelligent behaviour? Let us consider the case of the chick and the ladybird, and see whether it helps us to answer these questions. The chick is stimulated to the instinctive pecking response by a small moving object. That is the first scene of the little drama. In the second scene the ladybird is seized, sensory centres are unpleasantly stimulated, and the insect is dropped or thrown on one side with signs of disgust. Let us grant that this aversion with its characteristic response is also instinctive. There is no hereditary connection between scene 1 and scene 2. After an interval the curtain rises on act ii. The characters are the same as in the first scene of the previous act. But the action of the drama is different. The chick does not seize the ladybird. Why? Because there is an acquired connection between scenes 1 and 2 of the previous act. The chick has gained experience of the nauseous character of the insect, and this experience influences and modifies his behaviour. The essentially new feature, therefore, is the establishment of a connection which is not provided through inheritance. To put the distinction in a brief form, we may say that instinct depends on how the nervous system is built through heredity; while intelligence depends upon how the nervous system is developed through use.

Assuming that an animal is capable of gaining experience and of acquiring new nervous connections in the course of individual experience, it follows that, as has already been indicated, instinctive behaviour in its logical purity is only presented in the first performance of any given co-ordinated act. For after this the animal has gained experience of its performance; and this can no longer conform to a definition of instinct, according to which it is characterized as “prior to experience.” On the other hand, intelligent behaviour cannot be presented on the first occurrence of any action, since there is no prior experience thereof in the light of which it may be guided. This logical distinction may be expressed by saying that instinctive behaviour is always prior to experience, while intelligent behaviour is always subsequent to experience. When, however, instinctive procedure continues throughout life practically unmodified or but little modified, we may still class it under instinct, since the hereditary connections are still the predominant factors. And where the latter part of an instinctive sequence is modified by the experience gained in the former part, we may still term the modification intelligent, however small may be the time-interval implied in the word “subsequent.” Sharp as the logical distinction is, the behaviour of animals is in the main a joint-product, and whether we term it instinctive or intelligent depends upon whether the hereditary or the acquired factor predominates.

Passing on now to consider some further characteristics of intelligent behaviour, we may first notice what Dr. Charles Mercier, in his work on “The Nervous System and the Mind,” terms the four criteria of intelligence. Intelligence is manifested, he says, first in the novelty of the adjustments to external circumstances; secondly, in the complexity; thirdly in the precision; and fourthly, in dealing with the circumstances in such a way as to extract from them the maximum of benefit.

If, however, we are to regard these severally as criteria of intelligence, each should serve to differentiate intelligent from instinctive behaviour. But this is not the case. The precision of the adjustment cannot be regarded as a criterion of intelligence, for many instinctive acts are remarkably precise. No grocer’s assistant rolls a paper funnel with more precision than is displayed by the birch-weevil (Rhynchites betulæ) in constructing the leaf-case in which her eggs are laid. Curved incisions of constant form are made on either side of the midrib, and are “of just the right shape to make the overlaps in the rolling, and to retain them rolled up with the least tendency to spring back,”[49] while the tip of the leaf is rolled into a second smaller funnel, which is tucked in to close the opening of the first, after the eggs have been deposited. “The eggs hatch in their dark place, each giving rise to an eyeless maggot, which ultimately leaves the funnel for the earth.... Hence the beetle cannot be considered to have ever seen a funnel, and certainly has never witnessed the construction of one, though, when disclosed, it almost immediately sets to work to make funnels on the complex and perfect system” characteristic of the species. This is but one example of instinctive precision out of the many which could be cited. We may say, then, that though, when an act is otherwise shown to be intelligent, the precision is a criterion of the level attained by the intelligence, still it cannot be said to be a criterion which serves to distinguish intelligent from instinctive behaviour.

Nor can we regard apparent prevision (which is sometimes advanced as a criterion of intelligence) as specially distinctive of intelligent acts regarded objectively in the study of animal behaviour. For, as we have had occasion to show, there are many instincts which display an astonishing amount of what may be termed “blind prevision”—instance the instinctive regard for the welfare of unborn offspring which the mother will never see, and the instinctive preparation for an unknown future existence in the case of insect larvæ.

Nor, again, is the complexity of the adjustment distinctive of intelligence as contrasted with instinct. We have cited examples which afford evidence of much complexity in instinctive behaviour. The construction and storage of the nest among solitary wasps, and their methods of capturing and conveying the insects or spiders on which they prey, are sufficiently complex. So, too, is the behaviour of the Sitaris larva which attaches itself to the male bee, passes to the female, and then slips on to the eggs she lays; and so, again, is that of the Yucca moth, which collects pollen from the anthers, conveys it to the stigma, and then lays her eggs among the ovules. These cases show, too, that the circumstances may be dealt with in such a way as to extract from them the maximum of benefit. It would be difficult intelligently to improve upon the manner of dealing with the circumstances displayed in many familiar modes of instinctive procedure.

There remain the novelty of the adjustment and the individuality displayed. And here we seem to have valid criteria of intelligent behaviour. The ability to perform acts in special adaptation to new circumstances, and the individuality manifested in dealing with the complex conditions of a variable environment,—these seem to be distinctive features of intelligence. On the other hand, in instinctive behaviour there seems to be no choice; the animal is impelled to their stereotyped performance through impulse, as by a stern necessity; they are so far from novel that they are performed by every like individual of the species, and have been so performed by their ancestors for generations; and in performing the instinctive act, the animal seems to have no more individuality or originality than a piece of adequately wound clockwork.

Granting, then, that behaviour is shown to be intelligent by the fact that there is evidence of profiting by experience, we may say that the level attained by the intelligence is indicated by the complexity of the adjustment, its precision, the individuality shown, the amount of prevision disclosed, and in its being such as to extract from the circumstances the maximum of benefit. Many of these points, however, serve equally well to mark the level of instinctive procedure.

II.—Intelligent Behaviour in Insects

It is, as we have seen, among the higher invertebrates, especially in insects, that some of the most remarkable and complex instincts may be found. There is,[50] however, a tendency to ascribe the behaviour of insects entirely to instinct, without sufficient evidence that neither imitation, instruction, nor intelligent learning play any part. This is, perhaps, a survival of the old-fashioned view that all the acts of the lower animals are performed from instinct, whereas those of human beings are to be regarded as rational or intelligent. In popular writings and lectures, for example, we frequently find some or all of the following activities of ant-life ascribed to instinct: recognition of members of the same nest; powers of communication; keeping aphides for the sake of their sweet secretion; collection of aphid eggs in October, hatching them out in the nest, and taking them in the spring to the daisies, on which they feed, for pasture; slave-making and slave-keeping, which, in some cases, is so ancient a habit that the enslavers are unable even to feed themselves; keeping insects as beasts of burden, e.g. a kind of plant-bug to carry leaves; keeping beetles, etc., as domestic pets; habits of personal cleanliness, one ant giving another a brush-up, and being brushed-up in return; habits of play and recreation; habits of burying the dead; the storage of grain and nipping the budding rootlet to prevent further germination; the habits described by Dr. Lincecum, and to a large extent confirmed by Dr. McCook, that Texan ants prepare a clearing around their nest, and six months later harvest the ant-rice, a kind of grass of which they are particularly fond, even, according to Lincecum, seeking and sowing the grain which shall yield this harvest; the collection by other ants of grass to manure the soil on which there subsequently grows a species of fungus upon which they feed; the military organization of the ecitons of Central America; and so forth. Now, the description of the habits of ants forms one of the most interesting chapters in natural history. But to class them all as illustrations of instinct is a survival of an old-fashioned method of treatment.

To put the matter in another way. Suppose that an intelligent ant were to make observations on human behaviour as displayed in one of our great cities or in an agricultural district. Seeing so great an amount of routine work going on around him, might he not be in danger of regarding all this as evidence of hereditary instinct? Might he not find it difficult to obtain satisfactory evidence of the establishment of our habits, of the fact that this routine work has to some extent to be learnt? Might he not say (perhaps not wholly without truth), “I can see nothing whatever in the training of the children of these men to fit them for their life-work. The training of their children has no more apparent bearing upon the activities of their after-life than the feeding of our grubs has on the duties of ant-life. And although we must remember,” he might continue, “that these large animals do not have the advantage which we possess of awaking suddenly, as by a new birth, to their full faculties, still, as they grow older, now one and now another of their deferred instincts is unfolded and manifested. They fall into the routine of life with little or no training as the period proper to the various instincts arrives. If learning thereof there be, it has at present escaped our observation. And such intelligence as their activities evince (and many of them do show remarkable adaptation to uniform conditions of life) would seem to be rather ancestral than of the present time; as is shown by the fact that many of the adaptations are directed rather to past conditions of life than to those which now hold good. In the presence of new emergencies to which their instincts have not fitted them, these poor men are often completely at a loss. We cannot but conclude, therefore, that, although shown under somewhat different and less favourable conditions, instinct occupies fully as large a space in the psychology of man as it does in that of the ant, while their intelligence is far less unerring and, therefore, markedly inferior to our own.”

Of course, the views here attributed to the ant are very absurd. But are they much more absurd than the views of those who, on the evidence which we at present possess, attribute all the varied activities of ant-life to instinct? Take the case of the ecitons, or military ants, or the harvesting ants, or the ants that are said to keep draught-bugs as beasts of burden: have we sufficient evidence to enable us to affirm that these modes of behaviour are purely instinctive and not intelligent; that all the varied manœuvres of the military ants, for example, are displayed to the full without any learning or imitation, without teaching and without intelligence on the part of every individual in the army.

That in some cases there is something very like a training or education of the ant when it emerges from the pupa condition is rendered probable by the observations of M. Forel. As Romanes says,[51] “The young ant does not appear to come into the world with a full instinctive knowledge of all its duties as a member of a social community. It is led about the nest and ‘trained to a knowledge of domestic duties, especially in the case of larvæ.’ Later on, the young ants are taught to distinguish between friends and foes.”

We have only to weigh the evidence brought forward by such observers as Fabre and Dr. Peckham to see that among the solitary wasps and mason bees the behaviour, though founded on instinct, is in large degree modified by intelligence. The care with which a site for the tunnelled nest in the ground is selected, betokens something more than instinct. The following is a slightly condensed statement of Dr. and Mrs. Peckham’s observations on one of the solitary wasps (Aporus fasciatus).[52] “We were working one day in the melon-field when we saw one of these little wasps going backwards and dragging a spider. She twice left it on the ground while she circled about for a moment, but soon carried it up on to one of the large melon leaves, and left it there while she made a long and careful study of the locality, skimming close to the ground in and out among the vines; at length she went under a leaf close to the ground, and began to dig. After her head was well down in the ground we broke off the leaf that we might see her method of work. She went on for ten minutes without noticing the change, and then, without any circling, flew off to visit her spider. When she tried to return to her hole it was evident that some landmark was missing. Again and again she zig-zagged from the spider to the nesting-place, going by a sort of path among the vines from leaf to leaf, and from blossom to blossom, but when she reached the spot she did not recognize it. At last we laid the leaf back in its place over the opening, when she at once went in and resumed her work, keeping at it steadily for ten minutes longer. At this point she suddenly reversed her operations, and began to fill in the hole that she had made. She then glanced at the spider, selected a new place, and began to dig again. This hole was also filled in; she looked once more at the spider, and started a nest in a new place. This, in turn, was soon abandoned, as was a fourth. The fifth beginning was made under a leaf that lay close to the ground, but after twenty minutes’ work this place also was abandoned and a sixth started. This, however, was the final choice, and after forty-five minutes spent in digging it was completed.”

This description shows an amount of apparent fastidiousness which is quite irreconcilable with the hypothesis that the behaviour is merely instinctive. Not less fastidious are some wasps in the temporary closure of the hole with a stone or pellet of earth, the operation being repeated several times with different covers before the insect seems to be satisfied; while in other cases the hole is hidden by bringing earth in such quantity as to render the place indistinguishable from the rest of the field. But in one case observed by Dr. Peckham, intelligent procedure was carried so far as apparently to involve the use of a tool, the same behaviour having been independently observed in the same genus (Ammophila) by Dr. S. W. Williston of Kansas University. “Just here,” writes Dr. Peckham,[53] “must be told the story of one little wasp whose individuality stands out in our minds more distinctly than that of any of the others. In filling up her nest she put her head down into it and bit away the loose earth from the sides, letting it fall to the bottom of the burrow, and then, after a quantity had accumulated, jammed it down with her head. Earth was then brought from the outside and pressed in, and then more was bitten from the sides. When, at last, the filling was level with the ground, she brought a quantity of fine grains of dirt to the spot, and, picking up a small pebble in her mandibles, used it as a hammer in pounding them down with rapid strokes, thus making this spot as hard and firm as the surrounding surface. Before we could recover from our astonishment at this performance she had dropped her stone and was bringing more earth, and in a moment we saw her pick up the pebble and again pound the earth into place with it. Once more the whole process was repeated, and then the little creature flew away.”

Here we have intelligent behaviour rising to a level to which some would apply the term rational. For the act may be held to afford evidence of the perception of the relation of the means employed to an end to be attained, and some general conception of purpose. In this section, which deals with description of behaviour based on observation, the psychological explanation cannot be discussed. Similar indications of deliberate action may be held to be afforded by the sometimes elaborate “locality studies” which these insects seem to make,—by the “care that is taken by wasps to acquaint themselves with the surroundings of their nests.” A Sphex, for example, which had partially made and then abandoned several nests, left them without any locality study; but when she had completed a nest in a suitable spot she made “a most thorough and systematic study of the surroundings. She flew in and out among the plants, first in narrow circles near the surface of the ground, and then in wider and wider ones as she rose higher in the air, until at last she took a straight line and disappeared in the distance.” Another species (Cerceris deserta) “has the habit of making a number of half circles in front of the nest, and then, after rising a little higher, of flying several times completely round it.” The method of procedure is, it seems, so normal to the species that it is probably founded on an instinctive basis. Dr. and Mrs. Peckham, in commenting on their observations, say:[54] “If the examination of the objects about the nest makes no impression upon the wasp, or if it is not remembered, she ought not to be inconvenienced nor thrown off her track when weeds and stones are removed and the surface of the ground is smoothed over; but this is just what happens.” For convenience of observation they “sometimes gently moved intercepting objects to one side, but even such a slight change threw the wasp out of her bearings, and made it difficult for her to recover her treasure.” Where wasps form a number of nests in a small plot of ground, as in the case of Bembex, each knows and returns to its own hole, as was proved by Dr. Peckham, who marked the insects and their nests with paint.

So, too, with regard to prey. In the course of his observations on Pompilus, Fabre removed the spider which the wasp had deposited on a tuft of vegetation before she made her nest. As she was at work beneath the surface she could not see what went on above ground or where the spider had been redeposited some twenty inches from its former position. On emerging from the nest the wasp went straight to the original spot, searched round it for some time, then made further excursions, and discovered the spider. After slightly altering its position, and placing it on another tuft of vegetation, she returned to her subterranean labours, giving the observer an opportunity of again moving the spider. Five times did Fabre repeat the operation, and every time the wasp returned to the spot where she had last deposited her prey.

The same observer records some interesting experiments with the mason bee, Chalicodoma. The mud nests of the species investigated were affixed to stones on the banks of the Rhone. When a nest was partially constructed, the bee having flown off for more material, Fabre moved the stone to a new position, near at hand and easily visible from the original site. The bee went straight to the place where the nest had been, searched the immediate neighbourhood, flew off, and returned to the same spot to continue the search. If she came across her own nest in its new position she did not recognize it as hers, but left it after examination. But if a stone with the nest of another bee in about the same stage of construction was placed in the position occupied by her own, she adopted it. And when two nests near together, both half built, were transposed, each bee unhesitatingly adopted the nest which occupied the position where its own nest had been. It may well seem strange that, the general locality-memory being so well marked, the recognition of the particular stone and nest should be deficient. This may be due to the fact that the so-called compound eyes are the organs concerned in locality vision, while the ocelli deal with details at very close range, and that the former alone afford the requisite data for recognition; by their instrumentality alone arises the conscious situation which affords guidance in behaviour. And in that situation slight changes which for us make it “still the same but with a difference” render it no longer the same for a being of more limited intelligence—one probably incapable of analyzing the situation and seeing that the sameness preponderates over the difference. Be this as it may, the failure of a bee to recognize its own nest under circumstances so foreign to its experience as removal to a new spot may be paralleled with what I have observed in the case of sticklebacks. A nest had been built in a round glass bell jar which stood near a window. Some aquatic vegetation grew in the tank, and the nest was built on the window side. An experiment was made by turning the large bell jar through a right angle. The male stickleback searched for its nest in the old direction on the window side—that is to say, the same position in reference to the incidence of the light. The search was, of course, fruitless, and a new nest was begun in this position. Presently the old nest was discovered, and was then vigorously destroyed in just the same way as the nest of a rival is pulled to pieces and scattered. Here a new incidence of light and new direction of shadows seemed to have completely transformed the visual situation.

To return to insects, it is probable that the homing faculty is not the result of an inborn mysterious instinct dependent on some sense of direction of which we have no knowledge, but is based upon experience gained during their flight hither and thither—that, in a word, it is intelligent and not instinctive. Experiments of Fabre at first seemed to suggest some magnetic influence to which bees were sensitive; for when a minute magnet was fixed to a bee as it started on its return journey, the insect was at fault; but as a check experiment he affixed a piece of straw instead of a magnet, with similar results. Some of Fabre’s observations and those of Dr. Bethe[55] are difficult to reconcile with the hypothesis that, in the homing, guidance is due to acquired acquaintance with the locality. But, on the other hand, the experiments of Lord Avebury and of Romanes seem to favour this view. Romanes found that when bees were taken inland from their hive near the seaboard, and then liberated, they returned from considerable distances, the whole locality being familiar; but taken to the seashore, where the objects around them were unfamiliar (since the seashore is not the place where flowers and nectar are to be found), the bees, though not far distant from the hive, were nonplussed and lost their way. Dr. and Mrs. Peckham, as the result of their extremely careful observations, seem to have no doubt that the homing of solitary wasps is due to locality-experience; and of the social wasp, Polistes fusca, they say:[56] “We have seen the young workers make repeated locality studies when they first began to venture away from home, but as they occupy the same nest all summer they, of course, grow more and more familiar with their surroundings, until they become so thoroughly acquainted with them that they can find their way without the least difficulty. We have no doubt that with them, as with the solitary wasps, the faculty is not instinctive, but is the direct outcome of individual experience.”

In the interesting pages of the works in which Dr. and Mrs. Peckham describe their investigations, there are many observations which show that wasps are capable of intelligently profiting by the experience which their instinctive behaviour places them in a position to acquire. The inherited tendencies and aptitudes pave the way for acquired modification and accommodation of behaviour. To catch and paralyze spiders, to dig and prepare a tunnelled nest, and to carry the prey to the nest, all this affords the instinctive basis; but when the observers tell us that they “have several times seen wasps enlarge their holes when a trial had demonstrated that a spider would not go in,” and even on one occasion without trial when an unusually bulky spider was brought, there is something beyond instinct; there is intelligent adjustment to special circumstances given in experience. Presumably intelligent is the habit frequently observed in one species of Pompilus, and occasionally in another, of hanging the paralyzed spider in a crotch of a branching stem, usually of bean or sorrel, where it will be safe from the depredation of ants. On one occasion Dr. Peckham, desirous of seeing the exact manner in which the victim was stung, substituted an unhurt spider for that which the wasp had paralyzed.[57] “According to the habit of its species when danger threatens, it kept perfectly quiet, and when the wasp returned it was hanging there as motionless as a piece of dead matter; but she would not touch it; she hunted all over that plant and then over several others near to it, returning continually to look again at the right spot. After five minutes she flew off in the direction of the woods to catch another spider. Why did she go to the woods? Why did she not take the one that hung there in plain view? It could not have been due to the fact that we had handled the spider, since when, on other occasions, we took one that had been paralyzed, examined it, and then returned it to the wasp, she accepted it without hesitation.... In forty minutes she came back with another spider, but, instead of taking it into the nest, she hung it upon a bean plant near by, and then proceeded to dig a new hole a few inches distant from the first. Foolish little wasp, what a waste of labour! Truly, if you are endowed with energy beyond your fellows, you are but meagrely furnished with reason.”

Here we have the routine of instinct—the normal mode of hunting and capturing prey, the normal procedure of bringing the spider, and then making the nest, predominating over any tendency to initiate intelligent improvements. This, however, should not surprise us, in whom the force of habit is often so strong. Nor should we feel surprise at the apparently stupid tolerance some of these wasps display in presence of parasites. Bembex, which does not store and close its cell, but brings continual supplies of food to its larvæ, is not disturbed by the presence in the cell of the grubs of the parasitic fly Miltogramma. She could, we think, easily free her nest of these intruders, but she continues to bring supplies, though the parasites may absorb it all and leave her own larvæ to perish. She adapts her procedure to the new conditions, being incapable of knowing that she is feeding the enemies of her race.

Enough has now been said to show the extent and the limitations of the intelligence of such insects as the solitary wasps. It will be noticed that the acquired modifications of behaviour occur in close connection with the inherited ground-plan of instinctive procedure. We shall have occasion to note the same connection in our discussion of social behaviour in the next chapter. And we shall consider the influence of intelligence on instinct before we bring this chapter to a close.

III.—Some Results of Experiment

It is unnecessary to give a resumé of entertaining anecdotes illustrative of intelligent behaviour in the higher animals. Such anecdotes are too often the outcome of casual observations by untrained observers; and the interpretation put upon the facts frequently shows a want of psychological discrimination. Such is not the material of which science is constituted. What is needed is systematic observation conducted, so far as possible, under controlled conditions. Two things are necessary: first, to distinguish instinctive behaviour, inherited as such, from the acquired modifications or new departures due to intelligence; and secondly, to determine the method and range of intelligent procedure. These problems can only be solved in their entirety by a complete knowledge of the life-history of the animal concerned. But they may be attacked in detail by a systematic study of particular modes of behaviour, and by an investigation into their manner of origin. That this may be done with some approach to accuracy, resort must be had to experiment, which permits of observation under controlled conditions.

To ascertain, for example, how far nest-building is instinctive in birds, Mr. John S. Budgett hatched a hen greenfinch under a canary. In the following autumn he bought a caged bird, a cock, probably of the same year, and in the succeeding spring turned the pair into a large aviary, supplying such material as twigs, rootlets, dried grass, moss, feathers, sheep’s wool and horsehair. The hen soon began to build, the cock bird taking no share in the work, and finished her nest in a few days. On careful comparison it was found to resemble that of a wild greenfinch in every particular, being made of wool, roots, and moss, lined with horsehair. A second nest the aviary greenfinch built was also quite normal.

In the case of a bullfinch which Mr. Budgett reared, having obtained it when a few weeks’ old, the first nest was composed of dried grass with a little wool and hair, but without either rootlets or twigs. A second which she built was, however, quite typical, made of fine twigs and roots, and lined with horsehair; as was also a third nest.

It is just possible, though most improbable, that the bullfinch utilized its three weeks’ experience gained in the nest from which it was taken. But Mr. Jenner Weir describes[58] observations on canaries in which this source of experience is excluded. “It is usual,” he says, “with canary fanciers to take out the nest constructed by the parent birds, and to place a felt nest in its place, and when her young are hatched, and old enough to be handled, to place a second clean nest, also of felt, in the box, removing the other. But I never knew that canaries so reared failed to make a nest when the breeding time arrived. I have, on the other hand, marvelled to see how like a wild bird’s the nests are constructed. It is customary to supply them with a small set of materials, such as moss and hair. They use the moss for the foundation, and line with the finer materials, just as a wild goldfinch would do.”

Experiment seems, therefore, to show in a way, and with a clearness impossible of attainment by observation under natural uncontrolled conditions, that nest-building in birds is instinctive. That the manner and method of procedure is often modified in accordance with special conditions—that the instinctive outline of nidification receives its final touches through individual experience—is sometimes seen under nature, and more often under the semi-experimental conditions of domestication. Thus three pairs of pigeons in the Wilson Tower of Clifton College made their nests in 1898, as I am informed by Mr. H. C. Playne, of galvanized iron wire, pieces of which were left in a corner at the top of the tower, thus affording a parallel to the behaviour of the unconventional crow of Calcutta, mentioned by Mr. F. W. Headley,[59] which made its nest of soda-water bottle wires, which it picked up in a back yard. But even in this matter experiment serves to bring out clearly the selective influence which is exercised by intelligence. Bolton,[60] in 1792, observed a pair of goldfinches beginning to build their nest in his garden. They formed the ground-work of moss, grass, etc., as usual; but on his scattering small parcels of wool in different parts of the garden they, in great measure, left off the use of their own stuff and used the wool. Afterwards he gave them cotton, and they then used this instead of the wool; then he supplied fine down, and they finished their work with this, leaving the wool and cotton.

In studying the behaviour of wild animals under natural conditions, it must always be difficult to distinguish the congenital basis from the acquired elements; for both tend to bring about a working adjustment to the conditions of life, and we can seldom have opportunities of tracing the interplay of the factors which produce the instinct-habits of adult life. But under domestication we seek to bring about a new working adjustment to conditions imposed by man. The skilful trainer utilizes the natural instinctive tendencies as a basis; and, by a system of rewards and punishments, leads the intelligent modifications of behaviour along lines directed by his deliberate purpose. The conditions are largely those of experiment, and they bring out the play and range of intelligence in a way which would otherwise elude our observation. The training of falcons for the chase affords a good illustration, since they cannot be bred in confinement, and the effects of training cannot therefore be hereditary. The falconer’s object is to modify the congenital instinctive behaviour of a bird of prey for the purposes of sport. She is trained to the lure at first at short distances, and step by step through longer flights; she is taught by snatching away the lure to stoop at it repeatedly as often as it is jerked aside; and then she is trained on living quarry, at first under easy conditions, till eventually she can be flown at a wild bird. And as a result a well-trained falcon will follow her master from field to field, regulating her flight by his movements, always ready for a stoop when the quarry is sprung. The fact that she can be thus educated for her work shows that her behaviour is plastic, and can be moulded by intelligence. Experimental conditions reveal the fact; but under nature the moulding influence of intelligence is presumably not less important, though it is more directly in line with the congenital instinctive tendencies.

That much of the behaviour of the higher animals is guided by experience similar to that which plays so large a part in their training under the experimental conditions of domestication is generally admitted. But what are the range and limits of animal intelligence, and whether it attains the level of rational conduct, in the restricted sense of the term “rational,” are questions open to discussion, to which answers are more likely to be obtained through experiment than by chance observations.

Before giving some of the results of such experiment it will be well to revert to the distinction, which was drawn in the second chapter, between the lower or intelligent stage of mental development and the higher or rational stage. It will be remembered that rational processes were characterized by the fact that the situations contain the products of reflective thought, presumably absent in the earlier stages of development; that they were further characterized by a new purpose or end of consciousness, namely, to explain the situations which at an earlier stage are merely accepted as they are given in presentation or re-presentation; they require deliberate attention to the relationships which hold good among the several elements of successive situations; and they involve, so far as behaviour is concerned, the intentional application of an ideal scheme with the object of rational guidance.

On the other hand, the animal at the stage of intelligent behaviour deals with the circumstances of his comparatively simple life by making use of the particular situations which have been presented to consciousness in the course of his practical experience. If such an animal be placed in the midst of new circumstances he has to find out by a process of trial and error how they are to be met. After a longer or shorter period of trial, guided only by particular experiences, he chances to hit upon a mode of procedure which is successful. The successful act is then incorporated in a new situation; at first, perhaps, only incompletely. The association is eventually established by repetition, through which is acquired the habit of doing the right thing in the appropriate manner. Why he does this and not something else, in so far as he is intelligent and not rational, he probably neither knows nor has the wit to consider. The satisfaction of success suffices for intelligence as such. If the circumstances be so modified as to render the particular mode of meeting them ineffectual, after trying again and again in the old way, he will sometimes stumble upon the proper mode of overcoming the difficulty, and after doing so two or three times a new conscious situation involving the requisite associations will be established, and the appropriate behaviour will become habitual. But why this new mode of procedure rather than any other is adopted, intelligence as such does not know, because it does not analyze the situation and disentangle the essential relationships. The satisfaction of success again suffices. In a word, such an animal in the perceptual stage of mental development seems wanting in the power of reflection. He does not appear to show evidence of framing anything like a general scheme of knowledge which he can apply to the solution of particular problems, of a practical nature, involving difficulties and obstacles.

The method of intelligence—in the sense in which we are using this term—the method of varied trial and error with the utilization of chance success, is a lengthy and somewhat clumsy process; but it suffices. Now contrast it with the procedure of a rational animal, such as man is or may be. When he is confronted by a difficulty he is not content to meet it by trying this way, and that way, and another way, anyhow, and trusting to chance to bring success, but he considers the problem in all its relations with a view to ascertaining the essential nature of the difficulty. For each attempted mode of meeting the case he has a definite reason. He knows why he does this and not that. He has a plan or scheme which he puts into execution. And if it fail, he is not content till he finds out wherein the failure lay. This enables him to plan a better scheme. He sees why it is better; and if at last he be successful by a happy hit, as in the chance procedure of intelligence, he looks for the reason of it. And seeing why this fortunate attempt, unlike his previous efforts, just meets the case, he repeats it because he perceives that herein lies the essential solution of the difficulty. Both in the case of intelligence and in that of reason, as here distinguished, present procedure is based upon past experience; but reason has built upon the foundations thus laid an orderly scheme, and knows its whys and wherefores, while intelligence is at the mercy of chance associations. The reason for success it has not the wit to assign.

The essential difference between the two cases may be put in another way by saying that the intelligent being forms sensory impressions and sensory images linked together by bonds of association, combining and coalescing to constitute a conscious situation effective in behaviour under the guiding influence of pleasure and pain; while the rational being not only does all this, but goes further. He fixes his attention on the way in which the elements in the situation are connected and related; he builds an ideal framework on which the sensory impressions are set or move in an orderly manner. And it is this scheme, fashioned by reason and transforming the situation, which he utilizes in dealing with difficulties.

Yet another way of putting the same essential distinction is to say that intelligence deals with pictures, either directly presented to the senses or called up in re-presentation. If we state the matter thus, however, we must remember that the “pictures” may be painted in colours supplied by any of the senses; and that smells, tastes, sounds, touches, pressures, limb-movements, and so forth, are elements in the pictured product. Bearing this in mind, we may say that intelligence deals with sensory impressions and their revived images in concrete and particular situations; while reason analyzes the pictures, and extracts from them general notions in terms of which the pictures may be explained. For example, we picture a stone falling to the earth; but we explain it by the general notion of gravitative attraction. The conception forms part of our ideal scheme of knowledge, which is not itself picturable, though this or that example of its action may be presented or re-presented in sensory imagery.

Once more we may say—and this way of looking at the question arises naturally out of what has gone before—that intelligence deals with concrete examples, and does not rise to the abstract and general rule. The ideal scheme of reason is the result of abstraction and generalization. It is a framework of conceptions which can be applied to the particular facts which fall under observation to see whether it fits and meets the case. Intelligence has to deal with the facts as they present themselves, without the aid of an organized system of knowledge built up into an ideal scheme.

Enough has now been said to indicate the distinction between the method of intelligence and that of reason. It may, no doubt, be said that the terminology used is open to criticism; for, on the one hand, the words “intelligence” and “intelligent” are frequently used as synonymous with “reason” and “rational;” and, on the other hand, acts requiring neither abstraction, generalization, nor the application of any scheme of knowledge are frequently spoken of as “rational.” Hence there is, it may be urged, some danger of misunderstanding. This may be granted. And unless some such restriction of meaning under suitable terms be accepted by psychologists, misunderstanding will continue. More essential, however, than the distinctive terms we are to use is the distinction of method which underlies them. That, I trust, is sound. Dr. Lindley, in an interesting paper on “A Study of Puzzles,”[61] has utilized the distinction in his investigation of the mental development of children, and has found that the procedure of young children is predominantly of the “sense-trial and error” order which has above been termed intelligent; and he expresses the opinion that “most of the adaptations of animals are on this sense-trial and error level.”

Such certainly seems to be the conclusion to be drawn from my own experimental observations on dogs. It has frequently been asserted that the behaviour of a dog with a stick in his mouth, when he comes to a narrow gap, shows that he at once perceives the nature of the difficulty, and meets it in a rational manner by adopting the appropriate plan of action. He pulls the stick through by one end. But experiments, which I have elsewhere described,[62] showed that a fox terrier, fourteen months old, seemed to be incapable of perceiving the nature of the difficulty which vertical iron railings presented to his passage with a stick in his mouth, and only imperfectly learnt to overcome it after many ineffectual trials and many failures. The results obtained on the first afternoon may be quoted to indicate the nature of the evidence. The dog was sent after a short stick into a field, and had to pass through vertical rails about six inches apart. On his return the stick caught at the ends. I whistled and turned as if to leave; and the dog pushed and struggled vigorously. He then retired into the field, lay down, and began gnawing the stick, but, when called, came slowly up to the railings and stuck again. After some efforts he put his head on one side, and brought the stick, a short one, through. After patting and encouraging him, I sent him after it again. On his return he came up to the railings with more confidence, but, holding the stick by the middle, found his passage barred. After some struggles he dropped it and came through without it. Sent after it again, he put his head through the railings, seized the stick by the middle, and then pulled with all his might, dancing up and down in his endeavours to effect a passage. Turning his head in his efforts, he at last brought the stick through. A third time he was again foiled; again dropped the stick; and again seizing it by the middle tried to pull it through. I then placed the stick so that he could easily seize it by one end and draw it through the opening between the rails. But when I sent him after it, he went through into the field, picked up the stick by the middle, and tried to push his way between the railings, succeeding, after many abortive attempts, by holding his head on one side.

Subsequent trials on many occasions yielded similar results. But the following summer, when I resumed the experiments, I was able with some guidance to teach him to bring a long stick to the railings, drop it, and then draw the stick through by one end; though even then, if he had dropped it so that one end just caught a rail, he often failed, shaking his head vigorously, dropping the stick and seizing it again, and repeating this behaviour until it chanced to fall in a more favourable position. He did not apparently perceive that by gently moving the stick a little one way or other the difficulty could be simply overcome with little effort. Nor when given a crooked stick, which caught in a rail, did he show any sign of perceiving that by pushing the stick and freeing the crook he could pull the stick through. Each time the crook caught he pulled with all his strength, seizing the stick now at the end, now in the middle, and now near the crook. At length he seized the crook itself, and with a wrench broke it off. A man who was passing, and who had paused for a couple of minutes to watch the proceedings, said, “Clever dog that, sir; he knows where the hitch do lie.” The remark was the characteristic outcome of two minutes’ chance observation. During the half hour or more during which I had watched the dog he had tried nearly every possible way of holding and tugging at the stick. Such is the mode of behaviour based on intelligence—continued trial and failure, until a happy effect is reached, not by methodically planning, but by chance.

Two of my friends criticized these results, and said that they only showed how stupid my dog was. Their dogs would have acted very differently. I suggested that the question could easily be put to the test of experiment. The behaviour of the dog was in each case—the one a very intelligent Yorkshire terrier, the other an English terrier—similar to that above described. The owner of the latter was somewhat annoyed, used forcible language, and told the dog that he could do it perfectly well if he tried.

In experimenting with my fox terrier on the method adopted in seizing and carrying differently balanced objects, I used (1) a straight stick, the centre of gravity of which was at the middle; (2) a Kaffir knob-kerrie, the centre of gravity of which was about six inches from the knob; (3) a light geological hammer; and (4) a heavier hammer. In the last, the centre of balance was close to the hammer head. The net result of the observations was that the best place for seizing and holding the object was hit upon in each case after indefinite trials; that after three or four days’ continuous experience with one (say the knob-kerrie), another (say the stick) was at first seized nearer one end, showing the influence of the more recent association; and that there was little indication of the dog’s seizing any one of the four at once in the right place, that is to say, the point of seizure was not clearly differentiated in accordance with the look of the object. I tied a piece of string, in later trials, round the centre of balance, but this, at the time of the dog’s death, had not served as a sure guide to his experience.

The way in which my dog learnt to lift the latch of the garden gate, and thus let himself out, affords a good example of intelligent behaviour. The iron gate outside my house is held to by a latch, but swings open by its own weight if the latch be lifted. Whenever he wanted to go out the fox terrier raised the latch with the back of his head, and thus released the gate, which swung open. Now the question in any such case is: How did he learn the trick? In this particular case the question can be answered, because he was carefully watched. When he was put outside the door, he naturally wanted to get out into the road, where there was much to tempt him—the chance of a run, other dogs to sniff at, possibly cats to be worried. He gazed eagerly out through the railings on the low parapet wall shown in the illustration; and in due time chanced to gaze out under the latch, lifting it with his head. He withdrew his head and looked out elsewhere; but the gate had swung open. Here was the fortunate occurrence arising out of natural tendencies in a dog. But the association between looking out just there and the open gate with a free passage into the road is somewhat indirect. The coalescence of the presentative and re-presentative elements into a conscious situation effective for the guidance of behaviour was not effected at once. After some ten or twelve experiences, in each of which the exit was more rapidly effected with less gazing out at wrong places, the fox terrier had learnt to go straight and without hesitation to the right spot. In this case the lifting of the latch was unquestionably hit on by accident, and the trick was only rendered habitual by repeated association in the same situation of the chance act and the happy escape. Once firmly established, however, the behaviour remained constant throughout the remainder of the dog’s life, some five or six years.