Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 11, Issue 3, September 1963

Ethology—The Zoologist's Approach to Behaviour — Part I

page 170

Ethology—The Zoologist's Approach to Behaviour — Part I


This short review is intended primarily as an introduction for students not acquainted with the literature of ethology. Within the space available it has been necessary to be selective; I have attempted at least to mention all the major aspects of the subject but it has not been possible to give equal attention to them. My personal interests and opinions have no doubt introduced some bias and placed emphasis where others would not put it. For this I make no apology.

I have chosen to treat the subject in a roughly historical fashion: firstly because in a short period ethology has undergone such rapid development that it already serves as a good example of how sciences progress; secondly because some of the older ideas of ethologists are still in current use in some quarters, although they have been superseded in the opinions of most present day workers. I think the newer ideas can be best presented by indicating how they grew out of the older.


Large scale study of animal behaviour by zoologists is a relatively new development; it lagged behind the study of animal behaviour by psychologists and physiologists.* Roughly speaking (there are many exceptions), these earlier students of behaviour had turned to animals as means to ends: psychologists had studied animal behaviour as a means to the better understanding of Mind, particularly the human mind (see Lehrman, 1962), physiologists as a means to the better understanding of the functioning of the nervous system. Zoologists, on the other hand, are interested in animals much more as ends in themselves. When they turned their attention to behaviour zoologists differed from their predecessors in this field by being more aware of, and laying

* This statement should perhaps be qualified on two counts: firstly, the distinction between zoologists, psychologists, and physiologists is a relatively modern one and, even now, it is not always easy to draw; secondly, students of animals from at least the time of Aristotle more than occasionally took note of behaviour — in the cases of such people as Charles Darwin (e.g. 1872), Lloyd Morgan (e.g. 1894). Jacques Loeb (e.g. 1918), H. S. Jennings (e.g. 1906), and Jacob von Uexküll (e.g. 1921 and in Schiller. 1957) we have extensively worked out approaches to animal behaviour.

page 171 more emphasis on, the significance of an animal's behaviour in the context of its environment or way of life.

By and large the psychologists had looked for behaviour, in animals, that conformed to the categories of learning patterns that had been worked out for humans, and had found little else*; or they had dismissed the possibility of a satisfactory analysis of animals other than humans because such animals cannot introspect and tell us about their motives (e.g. Bierens de Haan, 1947). The physiologists, working at the level of simple reflexes, had generalised their results to the point of saying that all behaviour can be reduced to description or explanation in terms of simple stimulus-response connections — nervous links between specific receptors and specific effectors (e.g. Pavlov, 1927).

It was largely a reaction to these teachings which established the existence of ethology. Konrad Lorenz (e.g. 1935, 1937 a & b, 1950), building on the work of such people as C. O. Whitman, Oscar Heinroth, Wallace Craig. Edmund, Selous, Eliot Howard, Julian Huxley and Jan Verwey, showed that the doctrines of Gestalt or ‘purposivist’ psychology (the European school that denied the possibility of behaviour analysis without introspection), and of behaviourism (the composite of learning theory and Pavlovian reflexology that flourished in America during the thirties) would not do when applied to the majority of animals in nature. The Gestalt people were convicted of vitalism, of retreating into mysticism before the limitations of a scientific analysis had been tested. The behaviourists were praised for their tough-mindedness but censured for their narrow-mindedness. If they had taken the trouble accurately to observe their animals, Lorenz claimed, they would have seen that much of the behaviour was spontaneous — not dependent on changes in the immediate external stimuli — and that reaction to a stimulus was rarely constant. The facts, for Lorenz, indicated a measure of internal control that is independent of the external stimuli of the moment. A further set of facts indicated that certain aspects of this internal control are independent of the stimuli of any moment, i.e. they are inborn rather than acquired by experience — innate rather than learned.

Properly to understand a piece of behaviour we have to appreciate its function in the life of the animal and its position in the whole behavioural repertoire of the animal. This necessitates study of the animal in its natural situation or in conditions that do not disguise the biological relevance of its behaviour. Studies of this

* There are, however, numerous exceptions, particularly among American psychologists. For example some of the early work of Watson (1908), Yerkes (1912), and Lashley (1915, 1938) treated the behaviour of sub-human animals as sui generis and, more recently, Schnierla and his associates have emphatically argued the case for keeping in mind the differences in behaviour between animals at different phylogenetic levels (e.g. Schnierla, 1949).

page 172
sort, by such people as Selous (e.g. 1905), Howard (e.g. 1929), Huxley (1914) and Verwey (1930), had shown that an animal's behaviour is as nicely adapted to its environment and way of life as are its structure and physiology. To a zoologist brought up on Darwinian principles, adaptation suggests natural selection. This line of thought was strengthened by the observations of Whitman (1899, 1919) and Heinroth (1910, 1930) that many of the acts of birds (most of the early observations of behaviour were made on birds I are stereotyped, are performed in exactly the same way by all members of the species, and can be recognised as homologous with similar acts in related species, the degree of similarity corresponding to taxonomic affinity*. For example the courtship behaviour of the surface-feeding ducks studied by Heinroth (1910) and Lorenz (1941) was found to consist of sequences of unit movements or postures; the complete set of units is present in the courtship of each species but each species differs from the others in details of the form or orientation of the units or in the arrangement of the units in the sequences. Not only the motor components of behaviour may be stereotyped and species specific; the stimuli which elicit such behaviour patterns are often limited to a small portion of the range which the animal is capable of perceiving and these ‘releasing stimuli’ are the same for each member of the species but differ in details from one species to another. In social behaviour, such as courtship and territorial behaviour, the behaviour of one animal often functions as the releasing stimulus for behaviour in another of the same species. For such a signal function to work efficiently it is necessary that the display or posture be ‘understood’ by all members of the species and it helps for it to be conspicuous. Usually such a display or posture indicates a readiness in the animal performing it to do something such as attack or copulate and for this to be efficiently conveyed it is necessary that the signal be unambiguous. Hence we find that social displays and responsiveness to such displays have evolved together — are reciprocally adapted; the displays are stereotyped in form and message content, and they are clearly marked off from one another and from other behaviour. In many cases they appear to have been derived from the elements of some ‘neutral’ pattern, like preening or feeding, which have been modified in the direction of exaggeration of conspicuous features

* The Shorter Oxford Dictionary lists three meanings for ethology the closest to the present connotaion being J. S. Mills' use of it for ‘the science of character’ (System of Logic, 1843). Lorenz and his followers adopted the word from Heinroth (1910) and this probably links with Mills' use through Heinroth's emphasis on species specific aspects of behaviour. At the Macy conference on Group Processes. 1954, Lorenz claimed that Heinroth's meaning was ‘the study of innate behaviour. Species-specific drive activities’. Tinbergen recommended that ‘ethology’ be understood as ‘the biological study of behaviour’ (ibid.: 77).

page 173 and stylisation of variable features (see Tinbergen, 1952; Morris, 1957). Such a modification of behaviour in the service of a signal function is referred to in the ethological literature as ritualisation. In social hostile contexts displays clearly are of selective value because they avoid the risk of physical injury that animals expose themselves to in actual fighting. The selection of species specific courtship patterns seems frequently to have been involved in the evolution of sexual isolation between diverging populations (see Mayr, 1942). These facts of adaptation and the taxonomic distribution of behavioural characteristics strengthened the case for saying that much of the variation of behaviour between individuals and between species corresponds to variation in the germ plasm.

Finally there were observations of animals performing complicated behaviour patterns perfectly at the first opportunity without previous experience of practice or imitation. For example Grohmann (1939) reared a group of pigeons in narrow tubes so that these birds were prevented from carrying out the flapping movements of the wings which young pigeons perform before they can fly. At the age when pigeons are normally able to fly these experimental birds were released and flew immediately as well as unconfined controls. A similar experiment was carried out by Spalding (1873, 1954) on young swallows. Carmichael (1926, 1927) raised a number of tadpole eggs in a solution of chloretone, a substance that produces anaesthesia of striped muscle but permits normal growth. This prevented the practice of swimming movements by the developing tadpoles, but when they were eventually placed in pure water they swam as prefectly as controls of the same age that had been reared in normal conditions. To this list could be added the cases where trial and error learning is ruled out because unless a response is performed perfectly at the first time of asking, the animal is killed. For instance, unless the courtship dance of a male salticid spider inhibits the feeding responses of the female, he will be killed and eaten at his first attempt at mating. These cases, Lorenz argued, could not be explained as instances of learning from experience in the life of the individual; they could be explained only in terms of the history of the species and its genetic endowment.

The writings of Lorenz stimulated field naturalists and zoologists to pay close attention to the behaviour of animals and to think of it in terms of biological function and evolutionary origin. Detailed descriptions were produced such as the ethograms of Makkink (1936, 1942), Tinbergen's studies of birds (1935, 1939), insects (Tinbergen et al, 1942), and fish (Ter Pelkwijk & Tinbergen, 1937), Baerend's work on digger wasps (1941) and cichlid fishes (Baerends & Baerends van Roon, 1950). The functional significance of such things as the bill colour of gulls (Tinbergen, 1949. Tinbergen & Perdeck, 1950), countershading in caterpillars (de Ruiter. 1955), the red breast of the Robin (Lack, 1943), the page 174 large cheliped of male fiddler crabs (Crane, 1941), and the eye-spot patterns on the hind wings of Lepidoptera (Blest, 1956) were demonstrated by observations and experiments on behaviour. Comparison between related species indicated ways in which behaviour evolves, e.g. Tinbergen (1952, 1954), Daanje (1950). Behavioural characters were found to be useful in working out taxonomic or phylogenetic relationships (e.g., Lorenz, 1941, Spieth, 1950) and were sometimes even superior to structural characters in this respect (e.g. Adriaanse, 1947, discovered, on the basis of clear differences in behaviour, that what had been judged as a single species of digger wasp, on morphological grounds, was, in fact, two distinct species). New information was accumulated about the sensory worlds of different kinds of animals and their roles in the control of behaviour patterns (e.g., see Tinbergen, 1951).

Though much of this new information could be explained in terms of ultimate causes — biological utility and phylogeny — there remained the questions of proximate causation — the factors and mechaninisms acting here and now which directly determine what an animal is doing. The emphasis Lorenz placed on the inateness of behaviour implied a degree of independence of behavioural control from the vagaries of the external world, and this was made explicit in the kind of mechanism that he suggested for this control. He claimed that, far from being a stimulus-bound reflex machine, an animal is a spontaneously active thing driven from within by endogenously generated energy. He started from Wallace Craig's (1918) observation that many behaviour patterns can be described as a chain of variable, striving, goal-directed responses (appetitive behaviour) which terminates in performance of a simple stereotyped response (the end act or consummatory act). The appetitive acts are oriented by external stimuli (releasers or sign stimuli) and, once released, runs its course without further mediation from external stimuli*. Thus a hawk will fly over the countryside in search of food; sight of the food will change its behaviour to chasing and catching the prey — a sequence that will be guided by the kind of prey involved and the efforts it makes to get away; finally the prey will be killed, torn apart, and eaten. The act of swallowing will constitute the consummatory act. If such a consummatory act has just been released once or a certain number of times, repeated presentation of the releasing stimulus fails to have any effect; as time passes the threshold for stimulation sufficient to release the pattern falls and, if the stimuli are withheld long enough, the pattern will be performed in their absence. As an

* Some end acts have been analysed into an externally oriented component — the taxis — and a component independent of external cues after release— the fixed action pattern, e.g. the egg-retrieving of the Grey-lag Goose, Lorenz & Tinbergen. 1938.

page 175 example of such a vacuum activity, as he called it, Lorenz (1937) cited a captive starling that he had and which he consistently fed by hand. This bird would perform a complete sequence of prey-catching and eating reactions although there were none of the normal releasing stimuli present. (Tinbergen, 1951: 61-62, mentions a number of similar examples).

As further support for his belief that much behaviour is the expression of endogenous co-ordination and energy fluctuations, independent of afferent input, Lorenz cited the work of von Holst, Weiss and W. R. Hess. Von Holst (1932, 1933) had shown that the isolated nerve cord of an earthworm, deprived of all afferent stimulation, continues to send vollies of impulses along its length and that the timing of these rhythmical vollies corresponds exactly to the contraction waves that pass down the segments in normal locomotion. A spinal eel with its nerve cord isolated from all proprioceptive input continues to perform perfectly co-ordinated swimming movements (von Holst. 1937). In experiments on the growth of nerve fibres in axolotls. Weiss (1941) contrived a transplated limb graft that received connections with motor nerve fibres from the nerve cord before it had received any connections with sensory fibres; such a limb graft began making perfectly co-ordinated walking movements as soon as the motor nerves made their connections. Hess (e.g., 1956) electrically stimulated the mid-brain of cats with implanted electrodes and found that it was possible, by this means. to produce fully co-ordinated behaviour patterns, including appetitive sequences terminating in consummatory act, identical with normal behaviour. In Lorenz's view such experiments as these could not be accounted for by a chain reflex theory of integration.

From the fact of the specificity of the stimuli releasing a response, Lorenz argued that there must be. in the animal, a releasing mechanism for each such response. which is selectively responsive to only a narrow range of external stimuli. This is referred to as the angeborene auslosende Schema (AAM) or, in Tinbergen's translation of the term, the innate release mechanism (IRM)*. From the variability of stimulus threshold for such response patterns Lorenz concluded that there must be a variable internal factor underlying each pattern and posited that this factor took the form of reaction specific energy: corresponding with each response pattern there is an internal source, generating energy, which activates the appetitive behaviour when it reaches a certain

* In a recent review of the concept. Schleidt (1962) has pointed out an interesting difference in the senses of Schema and Mechanismus in German and that Tinbergen's translation also shifted the meaning. Schema was von Uexküll's term and it signified simply a correlate or image of the releaser, that must be carried inside the animal. Tinbergen's IRM signified a kind of structural organisation linking a specific stimulus to a specific reaction.

page 176 level; if the appetitive behaviour achieves the situation where releasing stimuli trigger the IRM for the end act then the accumulated energy is used up (consumed) in the performance of the end act; if the final releasing stimulus is not attained the progressive accumulation of energy dammed up at the IRM may burst through to expression in the end act in the absence of the appropriate stimuli, or this energy may ‘spark over’ into another outlet and be expressed in a quite irrelevant action. The performance of oddly out-of-place actions like preening movements, nest-building movements and even sleeping, during hostile encounters over territorial boundaries or during courtship, had been recorded in a number of studies (e.g., Makkink, 1936, 1942; Kirkman, 1937). These now received explanation in terms of the diverting of reaction specific energies, or drives, as a consequence of thwarting or conflict. They were labelled Ubersprungbewegungen (Tinbergen, 1940) or displacement activities (Armstrong, 1947, 1950). According to this theory a response could thus be caused in two ways: it could be caused by ‘its own factors’ as when feeding behaviour is consequent on hunger and the presence of food; or it could be caused by the factors belonging to another behaviour pattern, in which case it was a displacement activity. Kortlandt (1940 a & b) coined the terms autochthonous and autochthonous to distinguish the two kinds of causation.

Lorenz's scheme was elaborated by Tinbergen (e.g. 1951). He gave a more neurophysiological ring to it by renaming the action specific energy as motivational impulses and by referring to the sources of these impulses as centres in the central nervous system. Tinbergen (1942, 1950) and Baerends (1941) also introduced the notion of hierarchy into the system. Behavioural functions can be classified in a hierarchical fashion. An act can be described as belonging to a series of progressively more comprehensive classes. For instance a particular movement might be labelled as ‘digging’; this, together with others such as carrying material, will be further classified as ‘nest building’; nest building, together with classes on the same level such as courtship, territorial fighting, care of offspring, can be classed together as reproductive behaviour. Tinbergen claimed that the course of many behavioural sequences is a descent through such a hierarchy of functional classes. To return to an earlier example, the hawk that flies over the country is showing appetitive feeding behaviour; the precise behaviour that this leads to will depend on the kind of stimuli that is discovered — if the hawk encounters a flock of small birds it will do one thing, if it encounters a lone pigeon it will do another, if it sees a young rabbit will do a third thing. Whatever the kind of prey encountered, the result will be a switch to a more restricted class of actions (‘starling catching’ behaviour for example). Once the prey is secured the new set of stimuli at page 177 the bird's disposal guide it into the still more restricted classes of acts involved in killing, plucking and dismembering, and this finally provides the stimuli which release the fixed action pattern of the consummatory acts of biting-off and swallowing. Tinbergen represented his scheme graphically as a set of centres each of which was controlled from above by supply of motivational impulses, and from below by an IRM which could be opened by the action of a small range of external stimuli. The sequence of releasing stimuli encountered progressively narrowed the choice of outlets for a stream of impulses as it descended from level to level down the hierarchy of centres — the animal proceeds to more and more restricted types of appetitive behaviour until the sequence terminates in the end act.

A prominent place was given to innate elements in these theories. It was believed that learning could affect appetitive parts of a behaviour pattern to some extents but the consummatory act was regarded as purely innate — its constancy of form in the life of the individual, and in each individual of the species, pointed to the stability of species genotype rather than the uncertainty of environmental influences. The releasing mechanisms were labelled as innate for the same reasons.

Lorenz emphasised that learning could, for particular behaviour patterns, be confined to crucial short periods during development. He (e.g. 1935) found that in many birds, such as ducks and jackdaws, if the young are exposed to certain stimuli during a critical period in early life, these stimuli become irreversibly linked to certain behaviour patterns. Thus a duckling can be made to treat a green box as if it were its mother, and a jackdaw can be made to direct all its courtship behaviour to a man. This phenomenon Lorenz called imprinting.

Ethology, then, in the early 1950s, could be identified with a school of animal behaviour students who approached behaviour from the direction of ecology, evolution, taxonomy and comparative anatomy; who consequently emphasised the roles of genetic components in the development and control of behaviour; who insisted on thorough study of the whole of an animal's behavioural repertoire, preferably in its natural setting, and developed a set of new technical terms for classifying and describing behaviour; and who interpreted their findings in terms of models of energy generation, flow, accumulation and exhaustion.

These models exerted considerable influence and still provide the conceptual basis for analysing and thinking about behaviour in some quarters. They had the virtue of introducing order into a wide range of otherwise unconnected facts; their elegance and comprehensiveness carried considerable appeal of a sort that might he called aesthetic.

(To be continued)