Chapter 5

Acquired Behavior
 

It does not appear difficult, at first glance, to draw a sharp distinction between innate and acquired behavior. The difference is clear from their very designations alone, or so one would think But, as this chapter aims to show, in practice it is extremely difficult to draw such a distinction.
We have already seen that innate behavior is controlled by nerve structures whose composition is determined by the hereditary formula inherent in a creature's chromosomes. The chromosomes do not, therefore, exercise direct control over hereditarily fixed behavior but merely create the mechanisms which contribute to such control. This is not the case with acquired behavior. This, too, is operated from the central nervous system but by nerve structures which the organism builds up as a result of personal confrontations with its environment. The controlling structures may be similarly constituted in each case – and there is much evidence to suggest that they are quite similar – but they come into being in quite different ways.
In the case of acquired behavior, we call the process by which they come into being learning. A distinction is drawn between various forms of learning, although they do not permit clearcut differentiation. First, there is a relatively passive process the accumulation of experience – which depends upon the formation of- conditioned reflexes (associations). Then there is active learning of the sort known as trial and error, in which the forming of associations also plays an important part. A third

(original book page 52)

alternative is learning by imitation, a process which can be linked either with deliberate demonstration or with teaching through the medium of verbal communication.
Each case presupposes a basic aptitude The central nervous system must somehow be in a position to store experiences. Particular environmental impressions must leave behind traces of some kind. An organism must somehow be able to note certain details of its struggle with environment. For this purpose it needs a faculty which, when more highly developed, we call memory.
This faculty, which must not automatically be equated with deliberate human recollection, has been tested in animals by means of training experiments. The degree to which unicellular organisms possess memory is still in dispute, but such a faculty has been clearly demonstrated in very primitive flatworms (planarians). Experiments with a cuttlefish proved that its memory retained an impression for 27 days. In the case of a trout, memory survived for 150 days, of a rat for fifteen months, and of a carp for as long as twenty months. In each of these cases a particular incident had left traces in the brain of the creature in question which affected its behavior for a given period.
Much controversy surrounds the question of how we ought to conceive of these memory traces (engrams) in practical terms -in other words, how the central nervous system stores such experiences. The original theory was that memory depends on morphological or chemical changes inside the nerve cells. According to Eccles, on the other hand, memory is based on electrical oscillatory cycles which are interrupted by a particular stimulus – i.e., by a specific experience – and then resume their course.
The theory of chemical anchorage (molecular hypothesis) is supported by experiments conducted with planarians. These small worms were trained to perform a certain task (they are capable of such an achievement) and then cut in half. The regenerative capacity of the planarian is such that the forepart grows a new tail and the hind part a new head. Ensuing experiments seemed to show that both new individuals – the one with the regenerated head included – could accomplish the task in

(original book page 53)

question. This implied that changes effected by learning were not only of a material nature but distributed throughout the body. A still more astonishing experiment was conducted with rats. They were also schooled in a task and then killed. An extract from their brains was injected into the abdominal cavity of other rats, whereupon the latter apparently produced a higher success ratio when performing the task in question than before being so inoculated. Although these findings are in dispute, it now seems likely that memory is stored in special molecules probably those of deoxyribonucleic acid (DNA molecules). This would be particularly interesting because these molecules also carry hereditary coordinations, which would confirm a conjecture made as long ago as 1870 by Hering to the effect that common features exist between memory and heredity, which he termed organic memory.
Further experiments indicated the existence of two forms of memory, short term and long term. That totally different phenomena are involved became clear from experiments with cuttlefish, in which the two faculties are located in different areas of the brain. In the case of goldfish, it was possible to prove that their short-term memory changes into long-term memory within an hour, and that the latter definitely depends upon the formation of protein. It is conceivable, therefore, that both theories of memory are correct. Short-term memory might depend upon an electrochemical oscillatory process within the nerve cells, and these oscillations could lead to the forming of a physical substance in which the memory trace remains anchored for a considerable period.
Let us now turn to the first form of learning: learning by the formation of conditioned reflexes. Such learning can result in entirely new reactions, but it often promotes changes or refinements in behavior which is already innate. This is how the toad, having initially snapped at small moving objects, learns to avoid insects which taste bad or sting. The unpleasant experience associates itself with the memory of such creatures' special characteristics, and the toad refrains from snapping at insects of similar appearance. The same applies to the young European polecat, which will initially chase only creatures that move. Only experience teaches it to recognize the motionless

(original book page 54)

mouse as well. The learning of special routes requires the correct collation of numerous landmarks. When the digger wasp first leaves its hole in the ground it flies in a circle several times so as to imprint the neighborhood on its memory and thus learns its return route. Bumblebees afford an even better illustration of how this form of direction finding depends upon imprinted landmarks. If they encounter a conspicuous flower and return immediately to their nest, they can still find the same flower again readily. If, however, they meet a less readily visible flower, they make several circuits before returning home so as to plot the bloom's exact position in relation to certain landmarks. In each case learning is based on an association of distinguishing features. The innate mechanisms for the recognition of key stimuli thus become more selective, and the behavior of the individual becomes better adapted to the particular features of its environment.
Learning by trial and error plays a major role in the acquisition of physical aptitudes. In birds, for instance, the motor coordination of flying is hereditarily fixed – but only in basic outline. Real skill in navigation – especially the difficult art of landing can be acquired only by practice. Young mammals learn many of their adult aptitudes in the course of play, of which we shall have more to say later. They gauge the potentialities of their own bodies by repeated experimentation, and so build up cerebral control formulas which later stand them in good stead. Innate behavior patterns are often refined and improved in the process. For example, the "killing bite" is innate in the European polecat, but the animal must first learn the proper method of applying it to the neck of its quarry (e.g., a rat). This it does while playing with youthful contemporaries. Orientation problems, too, are solved by trial and error. This process has been exhaustively studied in mice, in artificial mazes in which food can be reached by a single specific route. After several fruitless attempts, the rodents stumble on the route by accident and finally, after repeated successes, register it. The guidance formula built up within their brain is then based on a whole system of distinguishing features-on acquired recognition of key stimuli which, when present in a certain sequence, elicit certain reactions.

(original book page 55)

Learning by imitation clearly requires special mental capacities, since this process is demonstrable only among the higher vertebrates. As soon as lion cubs are old enough to accompany their mother, they watch her hunting and thus learn how to stalk prey, keep to leeward, and perform outflanking maneuvers. It is well known that rats quickly learn to avoid poisoned bait. In this case, knowledge passes from one animal to another because the inexperienced take their cue from the experienced. Traditions can grow up among animals in this way. In England, titmice have learned how to use their beaks to open milk bottles left on doorsteps. This avian discovery was first observed at Swaythling, Hampshire, in 1921 and spread to many other parts of the British Isles – Scotland and Ireland included – in the twenty-six years that followed. The growth of a similar tradition was traced from individual to individual among macaco monkeys on the Japanese island of Koshima. The animals were fed by scattering grain on the seashore, which meant that the grains of wheat had to be picked out of the sand. One of the monkeys discovered that it was easier to separate grain from sand by throwing a handful of sand, complete with grain, into the water, where the components separated because the sand sank more quickly. This discovery, which took place completely free of human influence, was subsequently copied by other monkeys of the same community, and the expedient was adopted by no less than nineteen of them over a period of twelve years. In each of these cases the creatures in question had managed to introduce other individuals' motor coordinations into their own – a process which may seem simple to us but actually represents an extremely complex linking of sensory impressions with personal experiments in motion.
Learning by demonstration, an even rarer phenomenon in the animal world, is a faculty that seems to be confined to creatures of the highest intelligence. At Basel Zoo, Schenkel watched a mother gorilla lead her newborn baby to the bars of the cage, encourage it to climb by means of appropriate movements, and even assist it by guiding its paws. She was thus teaching it by encouraging some movements and inhibiting others – another process which seems obvious to us because

(original book page 56)

we are conversant with it, but which, in the female gorilla, presupposes a very advanced and complicated feat of intelligence.
What we mean by intelligence or understanding is easier to illustrate by means of examples than it is to define theoretically. If a chicken sees an inaccessible cache of food behind a fence, it runs up and down behind the barrier and achieves nothing. A dog in similar circumstances soon "realizes" that this is pointless. It gains access to the food by examining the fence to see if it has an opening or can be bypassed. What we call intelligent behavior generally depends upon a better recognition of related factors. We are still ignorant of the particular cerebral processes which form the basis of this ability. In essence, however, they probably entail the evaluation of past experiences, acquired independently of one another, in such a way as to master a new problem as it arises.
An attempt was made to analyze such performances in specially constructed cages. The objects of research – mainly rats and mice, but also doves, cats, and monkeys – were encouraged to perform assignments by pressing keys or manipulating other contrivances, success being associated with a suitable reward. It became clear that initial successes achieved by random experimentation usually failed to make a lasting impression. A graphic evaluation of results shows that the learning curve rises very gently at first. After further successes, however, the number of correct actions increases notably as though the creature has suddenly grasped the nature of the task confronting it. In many intelligence tests the animals did not proceed at random even during preliminary experimentation, but in such a way as to suggest that they were working on the basis of a hypothesis. Depending on the fruits of success in earlier experiments, they would – when confronted by an intersection in a maze, for instance – either favor one of two turnings for a while or choose them alternately.
Numerous experiments have demonstrated that animals, too, have the ability to abstract. They are capable of recognizing the essential features common to different phenomena and thus by abstracting certain relevant characteristics – arrive at concepts. These, however, differ from human concepts in that they

(original book page 57)

are averbal, or not crystallized in the form of verbal definitions. Experiments conducted by Rensch and Dücker with a civet cat revealed a considerable ability to sift various sensory impressions for certain characteristics essential to the significance of the whole (generalizing abstraction). The animal was trained to distinguish between two parallel semicircles (meaning "food") and two straight lines (meaning "punishment"). It was then presented with increasingly complicated patterns in which these two recurred in modified form. The cat showed that it could eventually distinguish between the concepts "bent" and "straight." It also, in similar fashion, formed the twin concepts "equal" and "unequal."
When we remember the innate ability to select certain key stimuli from a plurality of stimuli and recognize particular objects by their distinctive features, the analogy with concept formation becomes obvious. What occurs through an innate

(original book page 58)

mechanism in one instance is effected in the other by a nerve structure founded on experience. Recognition of the common element by means of certain typical features is what is involved in each case.
Koehler recorded outstanding examples of generalizing abstraction while testing the abilities of various creatures to count. Doves, parrots, ravens, and squirrels learned to pick out a quite specific, prescribed number of seeds or morsels of food from a larger number of the same. In the case of a gray parrot the experimental procedure was rendered still more difficult by obliging it to take the seeds from several covered food bowls containing different numbers of seeds (and in one instance none at all). The bird proved equal to the task: It uncovered each bowl in turn and stopped after finding and eating the preordained number of seeds. More than that, it understood its orders, just as well when the prescribed number was conveyed by means of light and sound signals. In other words, it had transferred the concept of number from light to sound.
Where such achievements are concerned, related factors are embodied in a form of complex. This particular faculty was exhaustively tested in maze experiments with rats and mice, which revealed an astonishing ability to transpose. For example, once mice had committed all the passages in a maze to memory, they could pick their way through another maze which resembled the first except that its passages met at an angle of 45 degrees, say, rather than at right angles. And they found their way around the new maze even when all distances were doubled-indeed, they succeeded even when the new maze represented a mirror image of the old.
The ability to grasp related factors is particularly marked in chimpanzees. These creatures can solve the problem posed by a suspended banana by stacking two boxes and mounting them armed with a stick. They are also capable of lengthening a stick by fitting two sections together. In Tanganyika Jane Goodall watched wild chimpanzees extracting termites from their nests with thin twigs or blades of grass. Using their fingers to open one of the exits used by termites at swarming time, they insert the twigs. This causes a number of termites to bite, whereupon The chimpanzees withdraw them on the end of the twigs and

(original book page 59)
 
 

eat them. Another feat of intelligence Jane Goodall also observed enables chimpanzees to get at water too deeply secreted in hollow tree trunks to be reached with the mouth. Just as we would employ a sponge, so they take a handful of leaves, thrust their arms into the hole, dunk the leaves in the water, and thus convey the liquid to their mouths. All these cases exemplify a use of tools based on intelligent behavior.
The ability to form new and individual patterns of action which are not hereditarily determined can thus be traced in animals as it develops; beginning with slight modifications in innate forms of reaction and ending with genuine feats of in-

(original book page 60)

telligence which approximate our own. We can also trace the continuous expansion and differentiation of the central nervous system – in other words, the increasing refinement of the organ responsible for such feats. Quantitative as well as qualitative differences appear to be involved here. Rensch succeeded in establishing the existence of a relationship between vertebrates' performances in the field of learning and their absolute brain capacity. It would seem, therefore, that a part is played not only by the particular architectonics within the brain but also, and to a very large extent, by the absolute number of ganglion cells available.
As already mentioned during our allusion to animal concept formation, there are certain parallels between acquired and innate behavior. This becomes even clearer in the realm of movement. With hereditary coordinations we saw that the cells which provide their anchorage are characterized by a spontaneous generation of excitation, and that there is a growth of appetencies which compel the organism to perform certain movements. The same applies in this respect to coordinations acquired by learning, which may be termed acquired coordinations. Once engraved on the brain by frequent repetition, these become habits. And as one can plainly see in animals, habits are linked with appetencies, one expression of this being the restiveness which afflicts animals after the expiration of the time at which they are accustomed to perform an action. A dog which is used to retrieving, and which is given no opportunity to do so, plainly shows it. Here, too, there is a lowering of the stimulus threshold-and when the action cannot be carried out, displacement movements occur. The excitation caused by the blocking of a habit leads to the performance of other actions of some kind.
Although innate and acquired behavior may come into being in different ways, they do betray certain similarities. Distinguishing between them can be difficult, however. To take conditioned reactions as an illustration of this, it is innate in dogs that their salivary glands begin to function as soon as they perceive a certain taste. Ring a bell before feeding (this was Pavlov's classic experiment designed to prove the formation of (conditioned reflexes) and before long the sound of the bell

(original book page 61)

alone suffices to cause the secretion of saliva, even without the
perception of taste. What is acquired and what is innate in this instance? The new nerve connection (association) is clearly acquired, but the new reflex also makes partial use of the old innate nerve track because no new nerve connection to the salivary glands has been formed. Thus, strictly speaking, only part of this new behavior is acquired
In the case of complex movements the problem becomes even more difficult. Many hereditarily fixed behavior patterns are still incomplete at birth and do not mature until later, so that an animal may seem to have formed certain behavior as a result of some learning process or other. Instances of delayed maturation have been demonstrated in both the motor and sensory fields. One classic case is Hess' experiment with chicks. Soon after hatching, these begin to peck at small objects with their beaks, but in the first few days they usually miss the mark. Give them a target consisting of a slab of soft clay with a nailhead in the center, and one can clearly see from the imprint of their beaks how wide of the mark they are – i.e., the extent of their "scatter." Their aim improves on the second and third days, and by the fourth day the imprints are close to the nailhead. It seems in this case that improved aim is the result of practice – in other words, acquired. Hess succeeded in showing that this is not so, however, and that what is involved is a late maturing of the directional mechanism. He provided newly hatched chicks with prismatic spectacles which gave their vision a slight bias toward the right. The chicks' beak imprints were displaced accordingly – that is to say, their scatter was centered not on the nailhead but some distance to the right of it. On the fourth day they were more concentrated but still to the right of the nailhead. The chicks had certainly not learned – in fact, the nailhead had not been struck once-but the scatter had become smaller. This clearly showed that their aiming mechanism had improved as a result of late maturing, not of a learning process.
The distinction between innate and acquired is further obscured in many forms of behavior by the fact that innate and acquired components are often closely conjoined or, as Lorenz puts it, entwined. For instance, the motivation for learning –

(original book page 62)

i.e., the reason why an animal troubles to learn at all – derives largely from its various instincts. This or that member gains predominance in the parliament of instincts, whereupon the animal exerts all its faculties in order to attain the goal dictated by its prevailing impulse. This form of endeavor is one of the chief reasons why an animal learns at all. What it learns is undoubtedly acquired, therefore, but the motive power is innate.
In addition, there is another special instinct which aims at learning for its own stake. This is the play or curiosity instinct which prompts the young of the higher vertebrates actively to explore their environment and test their physical ability by carrying out every conceivable type of movement. For the most part, hereditary coordinations are innately fixed in these animals only in the form of quite short series of movements-components which are subsequently built, by means of learning and practice, into more highly integrated and complicated motor patterns. The animals thus benefit from an ability to adapt themselves to environmental conditions far better than they could if dependent solely upon chains of action determined by heredity. Is behavior that comes into being in this way acquired, or is it innate? The answer is: both.
Experimenting with squirrels, Eibl ascertained the precise extent to which the concealment and opening of nuts may be ascribed to innate or acquired motor control. The act of concealment is entirely determined by heredity, whereas the technique of opening nuts comprises both acquired and innate components. The movements of gnawing and cracking are already present as hereditary coordinations, but the squirrel learns the best method of putting them into effect by experimentation. An expert can tell by the marks of gnawing on opened nuts whether an experienced or inexperienced squirrel has been at work. Inexperienced animals begin by gnawing crisscross grooves at random until the shell breaks open at some point. Experienced animals, by contrast, gnaw only one groove, insert their lower incisors in it for better purchase, and crack the shell open. Eibl noted that inexperienced squirrels also try to exert a leverage effect, but this yields results only when the groove is correctly aligned.
A further entwinement of innate and acquired behavior oc-

(original book page 63)

curs in the so-called learning dispositions, which represent a kind of foreknowledge of what ought to be learned. Chaffinches, for example, have a song with an innately fixed length and number of syllables, but its characteristic division into three strophes must be learned by imitating adult members of the species. If young chaffinches reared in isolation are played recordings of other species of birds, they will accept their song as a model, but only if it resembles that of the chaffinch in tonal quality and strophic form. If they are played various songs including that of their own species, they will recognize the latter and give it preference as a model. In this instance, as in numerous others, the ability to learn is not entirely flexible but innately slanted in one particular direction. The creature has a prescribed curriculum, as it were-in other words, an innate knowledge of what it should learn.
Acquired behavior becomes still more firmly rooted as a result of imprinting, a phenomenon discovered by Lorenz. Here, learning dispositions make their appearance at a quite specific sensitive or critical period, and learning of this kind results in patterns of behavior which cannot subsequently be changed. Goslings, for example, become imprinted by what they see immediately after hatching out and follow it around from then on. Normally this is their mother, but if they catch sight of a human being or a balloon, they will follow only a human being or a balloon. Not even reunion with their mother will alter their behavior. They are henceforward imprinted in favor of another object, and their instinct to follow can be aroused only by that object. In this case the act of following is present as a hereditary coordination, but the nature of what is followed depends upon sensory impressions received during the early days of existence. Instinctive behavior is thus in part flexible. The motor component is complete, whereas the sensory component acquired its shape from a particular impression.
Hess determined the exact duration of this critical period in ducklings. He mounted a dummy drake on a disk which was rotated slowly while a loudspeaker concealed in the dummy emitted artificial calls. Having been kept in total darkness for varying periods, each of the ducklings was allowed to _follow the dummy in a circle for one hour. Depending upon the

(original book page 64)

strength of their impulse to follow, they faithfully scurried along after the dummy. By this means, Hess was able to determine that the critical period in ducks occurs between the thirteenth and sixteenth hours after hatching. Ducklings which followed the dummy during this period were henceforth imprinted in favor of its particular characteristics. They continued to prefer drakes, although ducklings are normally led by the mother duck, which has different coloring.
In many creatures, sexual behavior is also determined during sensitive periods. A male duckling which is made to associate exclusively with male ducks during the crucial period will behave homosexually for the rest of its life, even when females are in the vicinity. Similarly, a young cockerel in its sensitive phase can be imprinted in favor of ducks and will later wade into the water in order to court them. A jackdaw which is reared by humans until fully fledged and prevented from seeing other jackdaws remains sexually imprinted in favor of human beings. It may consort with other jackdaws, but when the mating season arrives a year later, it will only court human beings – even if there are other jackdaws around. In all these instances, imprinting must occur long before the creatures exhibit any sexual behavior. Budgerigars imprinted in favor of human beings can be induced to mate and breed in a covered cage. One sight of a human being, however, and both birds start courting the latter, pair formation is disrupted, and the brood neglected. Such examples are clearly illustrative of the power inherent in this process.
In many creatures even brood-tending behavior is influenced by imprinting. The cichlid Hemichromis bimaculatus, for instance, can distinguish between the young of its own species and those of another. It protects its own and eats the others. If alien eggs are put beneath it at first breeding, however, it will protect the young fish that hatch from them and also prefer the young of the same species to its own in future breeding periods.
Imprinting determines motor behavior as well as sensory. Heinroth discovered this in some nightingales which he reared. The young nightingales were within earshot while he was recording the song of some blackcaps; some months later, when

(original book page 65)

their own song commenced in springtime, Heinroth was surprised to hear that they sang exactly like blackcaps, faultlessly and without omission. Similarly, if zebra finches hear nothing but young gulls during their first thirty-five days, they become imprinted by the gulls' totally alien call. Even if they consort continuously with their own kind thereafter, they still emit calls like young gulls and continue to do so.
It is interesting to note, finally, that serious behavioral disorders can occur in a creature if it is deprived in youth of certain environmental stimuli which are necessary to its normal development. Female rhesus monkeys, for example, proved to be poor mothers and extremely aggressive if reared apart from their own mothers. Their contact with other members of the same group was also impaired. Again, it is particularly important for young rhesus males to have playmates. If reared in isolation, they fail to copulate with females when adult because they grip them incorrectly. Significantly, they are incapable of learning the technique in later life. We speak here of learning processes which resemble imprinting. In this case, too, the formation of an innate behavior pattern is linked with the timely appearance of particular environmental stimuli.
Thus the question "Innate or acquired?" can be decided only on the basis of very careful investigation. Even acquired behavior may be largely influenced by a hereditary formula.
 

Back to the Table of Contents

Continue to And What of Man