Social Behavior, Animal

views updated

Social Behavior, Animal








Social behavior may be defined as any behavior which is stimulated by or has an effect upon another animal of the same species. So defined, almost all behavior may have some degree of sociality—and it is therefore proper to speak of behavior as highly social or only minimally so— but there is very little behavior which can be called truly asocial. Social behavior is therefore very nearly equivalent to behavior in general, as the same behavior may be social or not, depending upon its source of stimulation and its object. To qualify the above definition further, there are occasions, particularly in cases of domestication or hybridization, in which social behavior can be elicited by a member of another species.

Behavior patterns. A behavior pattern may be defined as a segment of behavior having a definite adaptive function, and therefore, as a basic functional unit of behavior. For example, a chicken approaching a pan of water, lowering its bill into the water, and then raising its head is exhibiting a pattern of behavior which has the function of providing the animal with water.

Distribution of social behavior. The parame-cium, so commonly studied in elementary biology courses, is a highly developed protozoan which exhibits a variety of behavior patterns. One of these is a simple form of sexual behavior in which the paramecia fasten themselves together and afterward exchange nuclei and divide. Another pattern occurs under unfavorable environmental conditions. When a drop of water containing several paramecia is allowed to dry slowly, the animals will huddle together and thus protect each other against drying out. Such behavior has the function of shelter or comfort seeking. In addition, this one-celled animal shows other patterns of behavior which are only incidentally social. One of these is an investigatory or exploratory pattern of behavior in which the animal drives itself ahead with its cilia in a spiral path and thus comes in contact with various objects. Depending on the nature of the objects, it may remain in contact or back off and start out in another direction. Finally, its pattern of ingestive behavior is extremely simple, consisting of driving water laden with food particles toward its gullet by means of cilia, and it involves only starting or stopping the beating cilia.

More complex patterns of behavior having basically the same functions are present in the lower invertebrates, but only in the arthropods and vertebrates can we find new functions. Male crayfish, for example, will fight with each other during the breeding season, an example of agonistic behavior. Certain insects show highly developed patterns of behavior for the care of the young, thus exhibiting care-giving, or epirneletic, behavior. The young larvae, by their movements, may signal for care and attention, exhibiting care-soliciting, or et-epimeletic, behavior. Likewise, ingestive behavior assumes a highly social function, especially among ants, which feed each other in a relationship that Wheeler (1923) called trophallaxis.

All of these behavioral functions are highly developed in vertebrates. In addition, superior eyesight permits vertebrates to develop patterns of social behavior which involve mutual imitation; this allelomimetic behavior is commonly seen in schools of fish, flocks of birds, and herds of mammals, where each animal follows the movements of those next to him, resulting in highly coordinated group movements (Scott 1958).

Systems of behavior . As behavior becomes more complex in the animal kingdom, a given species develops several alternate patterns of behavior for each general function. A paramecium, for example, has only one simple pattern of ingestive behavior and no patterns of agonistic behavior. In contrast, a mouse attacked by another mouse may fight back, run away, assume a defensive posture, or become limp and passive. Each of these patterns of behavior has the function of adapting to the same stimulus, and the attacked mouse usually goes through all or part of this behavior repertory until one response is finally selected on the basis of its effect. Such a group of behavior patterns having a common general function, and usually a common physiological basis, is a behavioral system. Behavioral systems are quite similar to physiological systems, and many of them are directly related, as for example, sexual behavior and the reproductive system. Like physiological systems, behavioral systems do not function entirely independently of each other, and in various species they may be organized on a somewhat different basis. In the dog and other members of the family Canidae, sexual and eliminative behavior are closely related to each other, whereas in birds the two are unrelated.

The nature of behavior patterns and their organization into a system is largely a result of the special heredity of the species concerned. A primary task in studying the social behavior of any species is to describe its special behavior patterns and their organization into behavioral systems. This behavioral inventory, or ethogram, as Tinbergen has named it (1951), provides the essential raw material for understanding social organization or conducting experimental work.

Development of behavioral systems . In most of the higher animals, behavioral systems are incompletely developed at birth or hatching. In young mammals the ingestive system is present in an immature form, the newborn animal taking its food only by nursing. The development of behavior patterns for ingesting solid food comes later. Some systems of behavior may be completely absent, and all of them tend to be loosely organized together as et-epimeletic behavior. As the animal grows older, both behavior patterns and behavioral systems become differentiated from each other, partly as the result of learning and partly because of the processes of growth and maturation.

Social organization

Social relationships

A social relationship may be defined as regular and predictable behavior between two individuals, usually of the same species. In some animals these relationships may be organized largely as the result of heredity, but in higher animals relationships develop as a combined result of hereditary factors and learning. When two chickens meet for the first time, each presents a problem of adaptation to the other, and one solution is to attack. The result is a fight, and usually one chicken wins and the other loses. When they meet again, there is a tendency to repeat the behavior, so that in the long run one individual forms a strong habit of winning and the other of losing. Since one of the principles of learning is the “law of effect” (which might also be called the law of least effort), the winning chicken reduces its attacks as time goes on, and the whole relationship is reduced to one of threat and avoidance behavior rather than actual fighting. This is an example of a typical dominance-subordination relationship. [SeeLearning, articles onInstrumental Learning and Reinforcement.]

Hereditary factors also help to determine the development of the relationship. If one bird is a male and the other a female, the male almost invariably wins. Among females, large hens are likely to win over small ones, and aggressive breeds over peaceful ones (Guhl 1953).

Thus, the development of a social relationship may be considered as a process of differentiation of behavior, affected by both heredity and learning. As will be seen below, the development of social relationships in young animals is a slower and more complex process.

Social relationships can be classified in two different ways. In Carpenter’s system, classification is based on the different kinds of individuals present in the system (1934). In howling monkeys and most other mammals there are three major kinds of individuals (males, females, and young) combined in six kinds of social relationships (see Table 1). The same scheme applies to all social vertebrates, but it becomes more complex in some of the social invertebrates. In bees there are three kinds of adults—males (drones), females (queens), and sterile females (workers)—as well as the young. In addition, the workers exhibit different behavior at different stages of development. In termites there may be as many as three different kinds of individuals capable of reproduction and as many more different kinds of workers.

Table 1 — Social relationships classified according toage and sex
Source: After Carpenter 1934.

Social relationships may also be classified according to the behavioral system or systems involved. On this basis a large number of social relationships are theoretically possible, but actually observed, important social relationships are relatively few in number. There are four which involve one behavior system only. Dominancesubordination relationships are developed in any animal system where agonistic behavior is found. Sexual relationships depend of course on sex behavior. Leader-follower relationships are based on allelomimetic behavior and result from an unequal degree of mutual imitation. Finally, mutual-care relationships, often seen in the mutual grooming of primates, are formed when both animals exhibit epimeletic behavior.

There are two important relationships composed of combined systems: trophallaxis and care-dependency. Trophallaxis involves mutual feeding and mutual care and is very important in the social insects. The workers feed the young and in turn lick off secretions from them. In mammals, birds, and other vertebrates which care for their young, epimeletic and et-epimeletic behavior are combined to produce the relationship of care-dependency.

Both of these schemes of classification may be combined for systematic study. For example, the behavior exhibited in a male-male relationship may be examined in detail and classified as dominance-subordination, leader-follower, or whatever it may be in the particular animal society.

The number of possible social relationships which an animal develops depends on the size of the group. Carpenter has presented a simple formula for the total number: n(n — l)/2, which is of course the number of possible combinations between any two individuals in a group of n individuals. When examined in more detail, this formula has certain interesting properties which have implications for the social organization of groups. If we consider groups of different sizes, starting with one and going up to ten, we see that an individual added to a group adds a number of relationships equal to the number of individuals already present, i.e., an individual joining a group of three adds only three new relationships, but adds ten to a group of ten. There is a very rapid increase in the number of relationships as the group gets larger, resulting in extremely complex organization. Experiments indicate that hens in a flock of one hundred can recognize every other individual, making a theoretical total of 4,950 relationships in the flock. Social activity tends to increase in proportion to the number of social relationships.

Primary socialization

In highly developed societies the individuals must become related to particular members of the species. An ant colony would become totally disorganized if its members reacted toward members of other colonies in the same way as to their own. Ants would wander from one nest to another, overcrowding one and leaving young unprotected in others. Therefore, at some period in their development, all individuals must learn to differentiate their own family or colony from others. As Fielde (1903) showed, there is a critical period in early development in which the immature ant can be transferred from one nest to another or even to that of another species and subsequently act as if it belonged with the new individuals. This transference of social relationships occurs regularly in the slave-raiding ants, a species of red ant commonly found in North America. The raiders remove the pupae from nests of black ants and take them to their own, where the black individuals subsequently rear and care for the offspring of the raiders.

A similar phenomenon takes place in all highly social animals which have been studied so far. Birds hatched in incubators or removed from their nests in early stages of development and reared by hand become attached to the human species and often fail to react to their own. In precocial birds, such as chickens, ducks, geese, and turkeys, the formation of the first social relationship takes place within the first 24 hours after birth. In the more slowly developing altricial birds, such as song sparrows, jackdaws, or doves, the process takes place at a later period in development, and the degree of attachment to the human handler varies inversely with the age at which the young are taken from the nest [seeimprinting].

Among mammals, also, the process varies with speed of development. In sheep and other herd animals the young lamb becomes strongly attached to its mother within the first few days, whereas in the more slowly developing dog the period of socialization begins at about three weeks of age, reaches a peak at about six or seven weeks, and slowly declines thereafter. The process of socialization is of course a reciprocal one, the parents becoming attached to their offspring as well. In sheep or goats the mother becomes attached to her offspring within the first two hours or so and thereafter will repulse any strange infant. Primary socialization has been little studied in primates other than man, but it takes place within the first six months in rhesus monkeys, as also appears to be the case in human infants [seeinfancy, article onthe effects of early experience].

The time at which primary socialization takes place is obviously a critical period in development, determining which individuals shall be the close relatives of the young animal. Later he may exhibit appropriate patterns of sexual behavior and other types of social interaction toward similar individuals [seeAffection; Socialization, article Onpsychological aspects].

What is the nature of the process involved in primary socialization? In the young animal there are positive behavioral mechanisms such as the“following” reaction of a young chick or lamb, which brings it into contact with members of its own species. There are also negative mechanisms, particularly the behavior patterns of escape, which develop somewhat later and prevent socialization with other individuals by inhibiting contact. As to the nature of the process itself, in precocial birds it obviously takes place in the absence of food rewards, and, under experimental conditions, without any obvious external rewards or reinforcement of any kind. While food rewards do affect the behavior of a young mammal, experiments have shown that the process of socialization takes place without them. All evidence indicates that the process is an internal one and is closely connected with emotional responses. Experimental studies indicate that any kind of strong emotion, whether it be that of loneliness, fear, pain, or hunger, will speed up the process of forming a social bond. While the capacity to form such a bond is strongly developed in the infant animal, the same process can take place at any time in later life at a somewhat slower rate, except when it is completely suppressed by the development of interfering behavior patterns (Scott 1962).

Animal and human societies

The kind of social organization developed by a species depends upon the nature of its behavior patterns and the ways in which they are organized into behavioral systems. Thus we may state a general law: Social behavior is a determinant of social organization.


The simplest kind of animal society is the temporary aggregation. Such groups occur widely in the animal kingdom and were extensively studied by W. C. Allee, who was able to show in a large number of cases that living in groups promotes the survival of individuals, whether they are protozoans or vertebrates (Allee et al. 1949). This kind of temporary aggregation is based on shelter– or comfort-seeking behavior.

A second kind of temporary aggregation is based on sexual behavior. This may involve only two individuals, as it does in paramecia, or vast numbers, as in the mating swarms of the palolo worm of the Pacific. The great majority of invertebrate societies belong to one or the other of these temporary aggregations.

In contrast, the higher invertebrates, and many vertebrates, form long-lasting groups with highly complex internal organization. Insect groups first received scientific attention, but it is now realized that the social behavior of insects is somewhat specialized, being limited by certain anatomical peculiarities such as the external skeleton and the relatively inefficient compound eyes. Because of these limitations, insects are small, relatively shortlived, and unable to respond to complex visual stimuli at a distance.

Insect societies

The most highly developed insect societies belong to two orders, the Hymenop-tera, including ants, bees, and wasps, and the Isoptera, or termites. In both orders the societies consist of permanent groups associated with nests that are constructed by the combined efforts of the members, and the most prominent social behavior is care giving. In an ant colony two workers meet, feel each other with their antennae, and if one has recently fed, it will regurgitate a drop of honey dew to be eaten by the other. This, together with care of young, constitutes most of their social behavior and is what Wheeler termed trophal-laxis (1923).

In any given species of insect, social organization tends to be rigid and stereotyped, repeating itself generation after generation. There is relatively little modification by experience, and the combined efforts of the individuals produce a stable social environment. Alfred Emerson points out that insect societies have many of the characteristics of a single organism, and he has called them “supra-organisms” (see in Allee et al. 1949).

The peak of development of the insect society is reached by the common honeybee, which maintains a permanent social group. In contrast to ants, in which each new colony is formed by a mated pair, large numbers of old bees accompany a queen moving to a new hive. Mating takes place at this time, and the fertile female, or queen, continues to reproduce for a long period, sometimes two or three years. The vast majority of the colony members are short-lived sterile females or workers which go through a regular series of activities as they pass through a few weeks of active life, first working around the hive and finally going out to forage for food. As von Frisch discovered, foraging bees are able to communicate to other workers the direction and distance at which food is found by means of body movements, or “dances,” when they return to the hive (1950). Present evidence indicates that the nature of the signals and the ability to understand them are determined by heredity rather than learning.

Vertebrate societies

The societies of vertebrate animals are also related to basic anatomy and physical capacities. Vertebrates have an internal skeleton and hence are capable of continuous growth. They can become immensely large, and their life spans may be enormously extended incomparison with insects. The capacity for learning is consequently much more valuable and plays a more important role than it does among invertebrate societies, so that there are many examples of cultural as well as biological inheritance. Furthermore, the vertebrate eye is vastly more efficient than the compound eye of arthropods, making mutual imitation or allelomimetic behavior possible, particularly for animals active in the daytime.

There is an enormous variety among vertebrate societies, ranging from temporary aggregations, like those of the lower invertebrates, to societies far exceeding those of the insects in complexity. The most highly developed societies are found among animals which occupy dominant ecological positions.

In the water, fish are still the dominant form of life, and their societies have taken two forms. One of these, the school, which is based on allelomimetic behavior, is very widely found. In herring or mackerel the fish move in groups throughout their lives, even spawning in a mass and paying no attention to the eggs or young. In contrast, there are many kinds of fish, like the sunfish and stickleback, which emphasize nest building and the care of the young. In these cases the male guards the nest and sometimes the young for a short period after hatching. The period of dependency remains short, however, as in most cases the young must feed themselves. The males guard definite boundaries around the nest and thus exhibit territoriality (Tinbergen 1953).

Amphibians, which usually occupy a relatively inferior ecological niche and are under constant pressure for survival, show almost no social organization except temporary mating aggregations and a few examples of care of the young. The same can be said for many reptiles, although some lizards develop dominance and territoriality and alligators build nests and briefly protect their young.

Birds, which are the dominant form of life in the air, show a great variety of highly developed societies. One of the characteristics of bird societies is the phenomenon of seasonal change. For example, during their brief mating season, song sparrows develop one sort of social organization involving sexual behavior, territoriality, and care of the young and then spend the rest of the year in flocks which are much like schools of fish. Since the young are produced as eggs which have to be incubated, a high degree of coordination of care-giving behavior and physiological processes is necessary, and in many birds this is exhibited by both sexes, which combine to build the nest and care for the young. Many species feed their young, and pigeons and other members of the dove family nourish their offspring with crop milk. A “nuclear family” of male, female, and young is characteristic of these and many other bird societies, but in chickens and other members of the same family, mating is polygamous, and the brood of young is raised by the female alone.

In mammals there is an enormous variety of societies differing in structure and degree of complexity. Among rodents, woodchucks are relatively solitary animals, reacting to each other with agonistic behavior except during the brief mating and reproductive seasons (Bronson 1964). At the opposite extreme are the prairie dogs, which run to groups numbering thousands of individuals. Each prairie dog colony is subdivided into territories, usually occupied by a sexually mature male, a group of females, and their young. Each year the adults in the territory move out to new locations on the edge of the colony, leaving the burrows behind for the next generation. Thus the oldest and most experienced members of the colony are found on its edges, where adaptation is more difficult (King 1955).

Other examples of mammalian societies are the great herds of hoofed mammals. In a typical society, such as that of mountain sheep, the two sexes mingle during the mating season. Males compete in individual combat, and the more successful ones do most of the mating. During the rest of the year there are separate male and female herds, the latter being the group in which the young are raised. Among these herds a definite leadership organization develops, the older animals of either sex tending to lead in their respective groups.

Some carnivores are quite solitary in their habits, but others, like wolves and dogs among the Canidae and lions in the cat family, live in permanent groups. These are, however, much smaller than those of the herd animals on which they prey.

Thus, we see that there is an enormous number of different kinds of animal societies. There is great variation among human societies, but the variety possible in other animals far exceeds it. It is therefore of considerable interest to see what kinds of social organization are present in man’s closest biological relatives, the other members of the Primate order.

Primate societies

Like other important orders of mammals, the primate group includes a wide variety of species, ranging from the small “bushbabies,” or galagos, of Africa to the large manlike apes and man himself. Primates can be at least as different from one another as cats are from bears in the order Carnivora. Their chief limitation of variation is an ecological one. Except for man, they live entirely in tropical and subtropical regions, and while many primates like to play with water, none of them is truly aquatic.

That primates other than man are highly social was first established by Carpenter’s classical study of the howling monkey (1934). These are tree-dwelling primates of Central America. A band of males, females, and young wanders from tree to tree, eating fruits and leaves and rarely descending to the ground. Their most obvious kind of social behavior is allelomimetic, since they are constantly following each other. The more active males generally take the lead, but there is no single established leader in their wanderings. Agonistic behavior is greatly reduced and expressed almost entirely as the vocalization that gives the species its name. Females in estrus pass from male to male, and there is no indication of sexual jealousy or rivalry over females. The most important social relationship within the group is that between mother and offspring, the mother constantly carrying and caring for her baby for the first two years of its life. Care-giving behavior is also shown by the males if a young animal falls to the ground or is threatened in some way. Different bands have roughly marked territories, and groups keep each other apart by means of threats.

Like man, baboons are a plains-living primate, and for this reason their social life has particular interest. Biologically they have evolved toward a doglike structure, often running on all fours and developing enormous teeth. However, they are largely vegetarian rather than hunting animals, often eating grass which they pick and eat a blade at a time. They readily adapt to living off farm crops, particularly melons and vegetables, and can be enormously destructive (Washburn & DeVore 1961; DeVore 1965).

As with howlers, a large part of their social behavior is allelomimetic, and they move in a group of males, females, and young which may include thirty to one hundred individuals. Agonistic behavior is more prominent than in howlers, and the males develop a definite dominance order. As the group moves, the most dominant male stays in the center, while the more subordinate animals keep their distance from him and each other. Living on the plains, the group is constantly threatened by large predators such as lions and cheetahs. Females and their young tend to group themselves toward the center of the group, near the dominant male, this being the safest spot. When a member of the group is attacked or threatened, all the males combine in its defense. Consequently, baboons are seldom actually harmed by predators. When in estrus, a female may mate with the most dominant male first but then move on to others as he becomes satiated. As with howlers, there is no evidence of sexual jealousy or permanent bonds between particular males and females. The most important social relationship is that between mother and offspring; in fact, all adults are extremely responsive to young animals and will approach and care for them if the mother permits. As the infant becomes more independent, it joins a play group of other young animals within the group. In general, the baboon society gives the impression of a closely knit and highly cooperative group.

Macaques are related to baboons, and their behavior is somewhat similar. The Indian macaque, or rhesus monkey, commonly lives near human habitations and is often fed by people. Carpenter studied an artificial colony of rhesus monkeys on Santiago Island off Puerto Rico and demonstrated that a dominance order between the males was an important part of their social life (1942). As with baboons, females in estrus pass from male to male. Overt aggressive behavior is perhaps more common than in baboons, possibly because of the disorganizing effect of human contact. Southwick found that where rhesus troops went into cities and competed for food, there were many cases of serious injury (1963).

The behavior of the Japanese macaque is quite similar. Imanishi reports that the animals spend their lives within a troop composed of males, females, and young (1960). The most dominant male usually occupies the central position, surrounded by females, while the younger and less dominant males keep to the outside of the troop and also precede it when it is moving. As with baboons, there is no subdivision of the troop into families or any permanent consortship between individual males and females.

The Indian langur, or Hanuman monkey, is a tree-living species and exhibits much less aggressive behavior than the macaques. However, Jay has described a definite dominance order among males (1965a; 1965b). AH females are subordinate to all males, except when an infant is threatened, and their own dominance order is not a clearly marked one. Sexual behavior is relatively unimportant, as the females have only a five-day estrus period per month and this is absent during pregnancy and lactation. Females have some tendency to consortwith the most dominant male, but there is no absolute correlation, and in one group studied by Jay the most favored male was a low-ranking one. Male-female consortships never last more than a few hours. In contrast to baboons, males pay little attention to the young, but infants, and particularly the newborn ones, are the object of much attention from all females in the troop. Males tend to take the lead when the group moves.

The manlike apes have been very difficult to study in the wild. In his original study Nissen (1931) was able to see very little besides the retreating chimpanzees disappearing into the tops of the tropical forest. More recently, Kortlandt located a favorable area on a Congo plantation where chimpanzees were protected and were in the habit of coming into the open (1962). Reynolds and Reynolds were able to keep track of a group of chimpanzees in the Budongo forest of Uganda with the aid of a large group of spotters scattered through the dense vegetation (1965). Unlike baboons, chimpanzees do not live in compact groups, but forty to fifty individuals may occupy an area of six to eight square miles. Except for females and their offspring, there are no permanent associations between individuals, and temporary groups of males and females, females and young, males, and groups containing all kinds of individuals are found in approximately equal numbers. Agonistic behavior is uncommon, and there are few indications of dominance and subordination. During estrus the same female may mate with several males in succession, and no fighting is observed between them. Chimpanzees appear to keep in touch with each other by vocalization, and on many occasions groups will join in vocalization over a large area. There is little evidence of leadership under most conditions and no reports of defense of territorial boundaries. Thus, chimpanzees form a very loosely knit but mutually tolerant society, with loose temporary associations between individuals, except for the mother-offspring relationship.

Goodall has studied chimpanzees in the comparatively open forests of the Gombe Stream Reserve in Tanganyika and has been able to follow the behavior of identifiable individuals over a period of several years (1965). Her more detailed accounts of behavior confirm the general picture of chimpanzee social organization. In addition, she has observed occasional predatory behavior and meat eating. These chimpanzees also exhibit use of tools, in particular, poking twigs or rolled-up leaves down the holes of termites in order to collect them for food.

The behavior of gorillas under natural conditions is also surprising. With their immense size these animals have little to fear from predators and within their own social groups are placid and peaceable. They are strict vegetarians, and the two sexes are so much alike that they are almost impossible to distinguish at any distance. Their sex organs are very small, and sex behavior plays a very minor part in their social life. Unlike chimpanzees, gorillas live in small compact groups containing both sexes and all ages of individuals. According to Schaller, there is a strong system of leadership, all members of the group responding to one of the older males (1963). Dominance is expressed almost entirely in the form of allowing precedence, and it is strongly related to age.

The only one of the manlike apes that shows anything like the human nuclear family of male, female, and their offspring is the gibbon. These long-armed tree-dwelling animals of southeast Asia are highly aggressive in both sexes, and the largest social group is a male, a female, and their immature offspring. Other groups are kept away by threats and fighting. Rather than supposing this to be the beginning of evolution toward a nuclear family, we can advance the hypothesis that the gibbon group is formed by a reduction of the troop to the smallest unit which is still biologically functional.

Primate societies do not present an evolutionary history of human social organization. Rather, each species has evolved its own social organization, producing an astonishing variety against which it is possible to compare and contrast human social organization. The one thing which is constant in all primate societies so far studied is the emphasis on the care of the young, particularly by the mother but also by unrelated males and females in the highly social forms. Agonistic behavior varies a great deal, as does sexual behavior. Most primates show regular periods of estrus, although there is some indication of a tendency toward a more extended receptivity in female chimpanzees. Was the social organization of our primitive human ancestors more like troops of rhesus and baboons, or was there a tendency to form temporary separate groups based on sex and age, as in the chimpanzee? The almost universal tendency toward division of labor in human societies suggests the latter. The animal data also suggest that the human nuclear family is perhaps not the beginning but the end point of human social organization, resulting from situations which demand great fluidity; that is, the nuclear family represents the ultimate reduction of family organization to the simplestunit which is still capable of all the functions of biological reproduction [seeFamily].

The evolution of social organization

Social organization is based on behavior, and behavior leaves no fossils. Therefore, any reconstruction of the evolution of animal societies must always be hypothetical and based on what we know about living animals. Assuming that the more lowly organized forms of living animals are similar to those which existed millions of years ago, we can suppose that there are two general bases for the initiation of social life. One of these is shelter-seeking behavior, in which animals stay together because the bodies of their fellows form a favorable environment. If this sort of behavior is extended, it should result in the animals’ following each other around, i.e., allelomimetic behavior. Animals that do the latter must have both motor equipment for rapid coordination and the necessary sense organs to keep in touch with each other. This kind of social group approaches its highest development in schools of fish, flocks of birds, and herds of mammals.

The other kind of social behavior which we may assume to be primitive is sexual behavior, which results in more efficient means of reproduction than scattering the sex cells broadcast. Retaining the eggs after they are fertilized and giving them care after they are laid still further increases the efficiency of reproduction, and the height of this kind of behavior is seen in two widely different parts of the animal kingdom—societies of insects and those of many birds and mammals.

These two basic kinds of social organization may evolve independently of each other, or both may occur together in the same species. In many cases social organization is quite independent of physical form, and closely related species may be either highly social or relatively unsocial. For example, the deer family range from the highly social elk through the more solitary Virginia deer to the moose, which is almost completely solitary except during the brief mating season and the association of a calf with its mother. Among rodents there is the prairie dog, a ground squirrel that lives in colonies of thousands of individuals, and at the opposite extreme, woodchucks, which lead highly solitary lives except for the brief necessities of mating and rearing the young. Even in the social insects there are close relatives of honeybees which build solitary nests. Thus there is no consistent relationship between biological form and social behavior, and the evolution of societies has become largely, though not entirely, independent of the anatomy of the individual. Consequently, in searching for the origin of human sociality, it is not possible to find anywhere the recapitulation of human social prehistory. The study of animal societies can at most suggest ideas and provide a basis of comparison.

The problem of basic human nature

In addition to biological evolution, human societies have added a new dimension: cultural evolution. The beginnings of this can be seen among the higher animals where the offspring learn fears from the previous generation. However, human language has enormously increased the potentiality of passing information along from one generation to another, and written language permits its accumulation over almost infinite times and in enormous amounts.

Compared to biological evolution, the study of cultural evolution is still in its infancy. Very little progress was made in the theory of biological evolution until the mechanism of heredity was discovered. Reasoning by analogy, we may suggest that the development of an adequate theory of cultural evolution will in part depend on our knowledge of the mechanism of cultural heredity, namely learning, and its application to cultural phenomena.

Unlike biological evolution, which moves in terms of generations, cultural change can occur with extreme rapidity and can be measured in terms of years. This raises the possibility that cultural change may move so fast that it exceeds the biological capacities of man in respect to social behavior. [SeeCulture, article onculture change].

Here again we are handicapped in our scientific information. There is no way of restoring precul-tural man to life and discovering what sort of social existence he led. Judging from existing animal societies, however, we may assume that his social behavior was well adapted to a particular kind of stable social organization. Beyond this, all we can do is to observe present-day human beings and study their social behavior under varying conditions. We know that sexual behavior plays an important role. Instead of definitely limited periods of sexual behavior, human females have very nearly reached the situation of constant receptivity, if not constant estrus, and there is no seasonal limitation on either sex. We know also that, because of the relatively slow developmental rate of human infants, care-giving behavior is highly important, as is the care-soliciting behavior of the young. It is likewise obvious that care-giving behavior is often extended to adults as well as to children and that it exists in both sexes. Young children are there-fore one of the focal points of human societies, and as with all primate societies, the relationship between old and young is an important one. As with any highly social species, we may assume that human beings have evolved behavioral capacities for the control of agonistic behavior, partly through forming habits of not fighting while young and partly through the formation of dominance-subordination relationships. These facts are almost self-evident. It is not as well recognized that human beings are also highly allelomimetic, with strong tendencies to do what others around them are doing. Because of differences in age and sex, allelomimetic behavior may be difficult in mixed groups, and we can observe a tendency for human groups to sort themselves out on the basis of age and sex. [SeeAggression; Imitation; Leadership].

Besides these major types of advanced social behavior, human beings continue to show all of the more primitive sorts of social behavior, for example, ingestive, eliminative, shelter-seeking, and investigative. In regard to the last, another characteristic of human beings is that they are highly curious. We can conclude that human beings show all of the basic kinds of social behavior in strongly developed form, and we can advance the hypothesis that any successful form of cultural organization of society must provide for reasonably satisfactory development and expression of these types of behavior. Conversely, any form of organization which attempts to completely suppress or distort a fundamental kind of behavior will result either in the dissolution of the society or maladaptive behavior on the part of the individual within it.

J. P. Scott

[Directly related are the entries Ethology; Imprinting; Instinct; Psychology, article on COMPARATIVE PSYCHOLOGY. Other relevant material may be found in Affection; Aggression, article on Psychological Aspects; Collective Behavior; Communication, Animal; Culture; Evolution; Genetics; Groups; Imitation; Kinesics; Sexual behavior, article on Animal sexual behavior; Social psychology.]


Allee, W. C. et al. 1949 Principles of Animal Ecology. Philadelphia: Saunders.

Bronson, F. H. 1964 Agonistic Behaviour in Wood-chucks. Animal Behaviour 12, no. 4:470–478.

Carpenter, Clarence R. 1934 A Field Study of the Behavior and Social Relations of Howling Monkeys (Alouatta palliatd). Comparative Psychology Monographs 10, no. 2; serial no. 48.

Carpenter, Clarence R. 1942 Sexual Behavior of Free Ranging Rhesus Monkeys (Macaco mulata). 2 parts. Journal of Comparative Psychology 33:113–162. → I: Specimens, Procedures and Behavioral Characteristics of Estrus. II: Periodicity of Estrus, Homosexual, Autoerotic and Non-conformist Behavior.

Clark, Wilfred E. Legros 1960 The Antecedents of Man: An Introduction to the Evolution of the Primates. Chicago: Quadrangle.

Devore, Irven (editor) 1965 Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Fielde, Adele M. 1903 Artificial Mixed Nests of Ants. Biological Bulletin 5:320–325.

Frisch, Karl von 1950 Bees: Their Vision, Chemical Senses, and Language. Ithaca, N.Y.: Cornell Univ. Press.

Goodall, Jane 1965 Chimpanzees of the Gombe Stream Reserve. Pages 425–473 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Guhl, A. M. 1953 Social Behavior of the Domestic Fowl. Technical Bulletin No. 73. Manhattan: Kansas State College, Agricultural Experiment Station.

Imanishi, Kinji 1960 Social Organization of Subhuman Primates in Their Natural Habitat. Current Anthropology 1:393–407.

Jay, Phyllis 1965a The Common Langur of North India. Pages 197–249 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Jay, Phyllis 1965b Field Studies. Volume 2, pages 525–587 in Allan M. Schrier, Harry F. Harlow, and Fred Stollnitz (editors), Behavior of Nonhuman Primates. New York: Academic Press.

King, John A. 1955 Social Behavior, Social Organization, and Population Dynamics in a Black-tailed Prairiedog Town in the Black Hills of South Dakota. Contributions from the Laboratory of Vertebrate Biology, No. 67. Ann Arbor: Univ. of Michigan.

Kortlandt, Adriaan 1962 Chimpanzees in the Wild. Scientific American 206, no. 5:128–138.

Nissen, Henry W. 1931 A Field Study of the Chimpanzee. Comparative Psychology Monographs 8, no. 1; serial no. 36.

Reynolds, Vernon; and Reynolds, Frances 1965 Chimpanzees of the Budongo Forest. Pages 368–424 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Schaller, George B. 1963 The Mountain Gorilla: Ecology and Behavior. Univ. of Chicago Press.

Scott, J. P. 1958 Animal Behavior. Univ. of Chicago Press.

Scott, J. P. 1962 Critical Periods in Behavioral Development. Science 138:949–958.

Southwick, Charles H. (editor) 1963 Primate Social Behavior; an Enduring Problem: Selected Readings. Princeton, N.J.: Van Nostrand.

Tinbergen, Nikolaas 1951 The Study of Instinct. Oxford: Clarendon.

Tinbergen, Nikolaas 1953 Social Behaviour in Animals, With Special Reference to Vertebrates. London: Methuen.

Washburn, Sherwood L.; and DeVore, Irven 1961 Social Behavior of Baboons and Early Man. Pages 91–105 in Sherwood L. Washburn (editor), Social Life of Early Man. Viking Publication in Anthropology, No. 31. Chicago: Aldine.

Wheeler, William M. 1923 Social Life Among the Insects. New York: Harcourt.


Man is a large, bipedal, diurnal primate closely related to the living great apes. Human behavior is based on a rich social heritage made possible by a tool-dependent culture and the unique properties of language. However, in most of the fundamental features of his social life, such as prolonged care of immature offspring and lifelong association between related adults, man is a typical Old World primate. Because man is most closely related to the Old World monkeys and apes, the following discussion is confined primarily to these forms and directed to those issues of most interest to the social sciences.

Living primates. There are about six hundred varieties of living nonhuman primates, divided into some fifty genera and two hundred species. Because of this extraordinary diversity, ranging from such forms as the small, insectivorelike tree shrews and “mouse lemurs” to the great apes and man, it is sometimes assumed that the behavior of living primates can be arranged along an evolutionary scale of increasing behavioral similarity to man. However, no living primate is the ancestor of any other, and many varieties have had separate evolutionary histories for tens of millions of years. Every living species has survived by specialized adaptations, and while it is possible to make some broad generalizations about the physical characteristics of prosimians, monkeys, and apes (Clark 1960), behavioral comparisons are more difficult. Washburn and Hamburg (1965) have recently discussed the classification of the primate order from a behavioral point of view.

The many varieties of monkeys alone range in size from creatures weighing less than a pound to others weighing more than one hundred pounds. They have exploited a wide variety of jungle and open woodland habitats, frequently achieving population densities of one hundred or more individuals per square mile. The Old World monkeys (Cercopithecidae) are divided into two subfamilies, the Cercopithecinae and the Colobinae; the latter group is distinguished by a specialized stomach capable of digesting large quantities of mature leaves. Brief observations have been made of several species of Colobinae, but the only form studied in detail is the common langur of India and Ceylon, Presbytis spp. (Jay 1965). Among the Cercopithecinae, numerous studies have been made of the two closely related ground-adapted forms, the African baboon (Papto) and the Asian macaques (Macaca). Of the other two basic groups of Cercopithecinae, the mangabeys (Cercocebus) and the guenons, or vervets (Cercopithecus), a long-term field study has been made of only one, a ground-adapted species, Cercopithecus aethiops (Struhsaker 1965).

The four apes, the Pongidae, are the gibbon (including the siamang), the orangutan, the gorilla, and the chimpanzee. All of the apes differ from the monkeys in that the trunk is “short, wide and shallow ... the lumbar region is short. . . and the shoulder and arm muscles are especially adapted to abduction, flexion, and rotation” (Washburn & Hamburg 1965, p. 9). Because of this adaptation to climbing and the ability to hang or swing from branches, to brachiate, apes have a “vertical” posture, thus differing from the horizontal, quadrupedal monkeys, and it is no surprise that in their structure and bodily movements the apes closely resemble man. Physically, and to some extent be-haviorally, man is most like the African great apes —the gorilla and the chimpanzee. But apes are confined to heavily wooded areas, and some of the behavioral patterns which have enabled man to travel long distances in open country are more closely paralleled in the behavior of terrestrial monkeys, such as baboons and macaques. In fundamental senses such as hearing, eyesight, and smell and in the organization of the brain, the viscera, and the reproductive organs man, the apes, and the Old World monkeys share a basic biological pattern which distinguishes them from all other mammals, including New World monkeys and other primates.

Field studies of primates. Although attempts were made to study the African apes soon after their discovery in the nineteenth century, most reports of monkey and ape behavior in the wild remained anecdotal until Carpenter observed the howling monkeys of Panama in 1931. In succeeding years, other field studies were attempted, but these were usually brief; then, in the 1950s, investigators in Japan, the United States, and Europe independently initiated a variety of field studies emphasizing long-term, systematic observation of primates living in natural habitats.

Most field studies have been made on monkey species that spend large amounts of time on the ground, where observation conditions are easiest. Because of their similarity to man, all the living apes except the orangutan have been the subject of at least one major field study. Long-term studies of free-ranging groups in monkey colonies have also contributed valuable data. The Japan Monkey Center has kept records of individual monkeys in its colonies since the early 1950s. Beginning withAltmann in 1958 (see Buettner-Janusch et al. 1962), various investigators have restudied the rhesus monkey colony established on Cayo Santiago by Carpenter in 1938; in fact, since 1958 almost all the members of this free-ranging colony have been tattooed for identification. Because the life cycle of primates is so long, field workers cannot ordinarily ascertain sibling or mother-offspring bonds between adults, and yet these relationships are proving to be very important in some species. (For a more complete history of field studies see Southwick 1963, pp. 1–6; and Wash-burn et al. 1965.)

Laboratory studies of primates, especially of the common Indian rhesus monkey, have continued without interruption since the 1930s. Contemporary views of such topics as discrimination, operant conditioning, perception, and learning have been summarized by Schrier, Harlow, and Stollnitz (1965). Recent studies of hemoglobins, blood serums, chromosomes, the nervous system, and the skin have been reported by Buettner-Janusch (1963–1964). There are two journals devoted entirely to primate studies, Primates and Folia prima-tologica, and there are many films of primate behavior.

The formerly facile generalizations about monkey and ape behavior are no longer tenable. Although only about a dozen of the several hundred varieties of living primates have been studied to date, it is already clear that primates exploit a wide variety of ecological niches, that they live in different kinds of social groups, and that different species display quite different temperaments. Some species are distributed over large geographical areas, and group structure and behavior have been found to vary significantly between different populations of the same species (Jay 1968). The unique life history of every individual produces distinctive patterns of temperament and social behavior among the members of a single group. All of these variables make it inevitable that there will be important exceptions to many of the following generalizations.

Primate group structure

Most vertebrate social groups undergo a variety of changes in group composition during the year. Changes are correlated with such phenomena as the abundance of food, the mating season, the birth season, and the maturation of offspring. During the mating season adult males may drive off young males and form harems of receptive females; mothers may leave the group altogether when they give birth, living apart until their offspring are able to support themselves, or females with immature offspring may form separate bands of their own. Some or all of these groups may participate in seasonal migrations. The result is a “population” with changing aggregations of individuals during the yearly cycle.

The social group of the Old World monkeys and the apes is very different. Membership in the social group is usually continuous, and group composition may remain stable over many years. Typically, an individual is born, matures, leads its adult life, and dies in the same group. While there are interesting exceptions to these generalizations, they are so characteristic of a wide variety of monkey groups that it is convenient to speak of a distinctive group organization, the “troop.” A troop so defined is a group of several adults of both sexes, together with juveniles and infants, which maintains a social identity and spatial unity that persists through the annual cycle and transcends the life-span of individual members. Such a troop is easily recognizable to the field observer because its boundaries are typically defended against outsiders. This social organization describes the basic social group of such Old World monkeys as lan-gurs, baboons, macaques, vervets (Cercopithecus aethiops), and colobus; of the New World howling monkey (Alouatta palliata), and perhaps spider monkeys (Ateles geoffroyi) (see Table 1). Less complete studies suggest that this type of group organization may also be characteristic of at least some species of mangabeys and some other species of Cercopithecus. The troop is also characteristic of Lemur macaco and probably of some other species of lemur. The description of the troop would apply to the social organization of the mountain gorilla, except that members may move in and out of some mountain gorilla groups rather freely.

Old World monkey and gorilla troops average about 25 individuals, but normal groups may be as small as 10 or number more than 200. The fact that primates normally remain in one social group throughout the year means that group members with needs and motivations as different as a small, helpless infant and a large, potentially dangerous adult male must accommodate each other. Among most vertebrates, when antagonisms arise between adults or between adults and mature offspring, usually during mating or birth periods, the result is a change in group structure. In some primate;

Table 1 — Types of social groups among primates
Type of groupProsimiansNew World monkeysOld World monkeysApesPreagricultural man
a. Inadequate field data.
b. Changes during annual cycle.
c. Changes during daily cycle.
d. Geographical variation.
Solitary individuals (usually nocturnal or crepuscular)Tree shrew  Orangutana 
 Galago b    
Mated pair with offspringHapelemurAofes Gibbon“Family”
 Galago b    
One-male group  Patas  
   Theropithecus geladac  
   Theropithecus geladac  
Troop, but oriented to one male Spider monkeyaCercopithecus ascaniusGorilla 
Troop (multiple adults of both sexes)Lemur macacoHowlerMacaquesChimpanzeed“Band”
 Lemur cattaSquirrel monkeyaSavannah baboon  
 Lemur fulvusCapuchinaVervet  
 Propithecus verreauxiSpider monkeyaMangabeysa  
Unstructured aggregationGalagobTheropithecus galadacChimpanzeed“Dialect group” 
   Paplo hamadryasc  

species young adult males may change groups or live for a time in an all-male group or as solitary animals. But most remain in their natal group, where mutual tolerance and group cohesion are accomplished by such different behavior patterns as the dominance hierarchy, the persistence of mother-offspring bonds into adult life, an elaborate repertoire of affective communication, and the conservative tradition of the group. In such groups each individual has a status with regard to every other individual, and through the long period of infancy and adolescence, the individual learns to anticipate the responses of other group members and behave accordingly. It is in the rich behavioral context of these Old World primate social groups that many of the behaviors which we have come to consider “human” were first developed.

In patas monkeys (Erythrocebus patas), hama-dryas baboons (Papio hamadryas), and geladas (Theropithecus gelada) the basic social unit comprises several adult females and their offspring but only one adult male. Still another primate grouping pattern is the mated pair with offspring, reported to be found in the Indridae, in some small South American monkeys, and in the gibbon. This form of social organization has been compared to the human family, but it is actually the result of very different behavior patterns. An adult gibbon, for example, is antagonistic toward all other gibbons except its mate and its immature offspring. As a result, each gibbon pair lives apart from all other pairs, and even their own offspring are apparently driven away at maturity. In this respect the gibbon group is more comparable to a mated pair of birds during the breeding season than it is to the human family. (Unlike most birds, the gibbon group is stable and unaffected by seasonal variation.) The human family cannot occur as an isolated pair of adults but is always a subgroup within a larger social unit.

Most of the prosimians are small, arboreal, nocturnal creatures which, like many other such mammals, forage as individuals, coming together only rarely. Yet some, such as Propithecus verreauxi and some species of lemurs (such as Lemur macaco), live in small troops. The small galago or bushbaby (Galago senegalensis) forages singly at night but may sleep in small groups during the day and form mated pairs during part of the year. (The most complete discussion of the classification, social groups, and activity patterns of Madagascar lemurs is in Fetter 1962 and is summarized in Buettner-Janusch et al. 1962 and in DeVore 1965a.)

Still another form of social group is described for the chimpanzees by Goodall (1965) and byReynolds and Reynolds (1965). Small subgroups of a population, such as females with young, adult males, and mixed groups of males and females, roam freely in a local area of forest. These clusters are in frequent contact and may gather together, for example, at food trees but do not appear to be organized into large persistent social units. On the other hand, recent observations of chimpanzees by the Japan Monkey Center personnel suggest a group size and organization much like a monkey troop: when crossing open country between forest patches, the chimpanzee group consistently numbered 43 and contained a nucleus of adult males, females in estrus, and females with infants. On the basis of brief observations some monkey species, especially in the New World, may also live in loose aggregations. The aggregations of hamadryas baboons are different in that the one-male units remain intact even though more than seven hundred individuals may gather at the same sleeping place (Kummer & Kurt 1963).

As a result of field studies in progress, it will soon be possible to refine the descriptive categories shown in Table 1 and to add many new species. While the table reveals certain trends, such as the solitary life of many nocturnal prosimians and the frequent occurrence of the “troop organization” in monkeys, it also illustrates the difficulty of abstracting a model of “monkey” or “ape” social organization. Further, gross categories such as those in Table 1 should not obscure major behavioral differences: the frequency, intensity, and complexity of social contacts among members of a lemur troop are far fewer than comparable contacts among members of a monkey troop. This list of primate species could be tabulated with respect to characteristics such as habitat, diet, and group size (Crook & Gartlan 1966). For example, with the possible exception of some prosimians, males of all these species are dominant over females, and this is broadly reflected in the degree of sexual dimorphism they display. There is great variation in this characteristic, ranging from the gibbon, in which the sexes are scarcely distinguishable in size, to the baboon and the gorilla, where the male may be two or three times as large as the female. Sexual dimorphism with respect to body size, size of canine teeth, and general robustness of the skeleton and musculature all tend to correlate with the extent to which a species is adapted to the ground. Indeed, the degree to which an Old World primate species is adapted to life in the trees or to life on the ground is one of the best ways of predicting the size of the group, the degree of sexual dimorphism, the intensity of dominance relations among adults, the use of loud vocalizations in intergroup spacing, and the size of the area over which the social group ranges (DeVore 1963).

Dominance and aggression

“Dominance” has been denned in many ways by different workers, and further study will undoubtedly show that quite different behaviors have been subsumed under this rubric. Common to all concepts of dominance, however, is the notion that the dominant individual can assert priority over a subordinate in order to gain some desired objective. Although expressions of dominance vary widely from one species to another, between groups within the same species, and even in the same individual at different times, it is clear that dominance interactions are fundamental to primate behavior and that they are especially prominent in the Old World monkeys, apes, and man. It is in the primarily ground-dwelling species such as baboons and macaques that dominance behavior is most apparent, and it is in the social groups of these species that dominance hierarchies exert the most influence in social organization. In these monkeys an adult male is about twice the size of an adult female, and there is typically a pronounced gap between the bottom of the male dominance hierarchy and the top of the female hierarchy. In more arboreal species, females are usually more nearly the size of males, and the dominance hierarchies of the sexes may overlap to some extent. In such species expression of dominance is often less frequent and more subtle, and mere avoidance, for example, may be as important as a challenge or attack in the demonstration of dominance status. In one-male groups the male is larger and more dominant than the females and does not tolerate any other adult males. In the loose aggregation of a species like the chimpanzee, there are frequent dominance interactions between adults, but it has not been reported whether these result in a hierarchy such as gorilla groups show.

The rank of any individual in a dominance hierarchy was formerly considered to be the result of the success of that individual in dominance interactions with each of the other individuals in the group. However, from recent studies of Old World monkey and ape groups, it appears that dominance status is seldom, if ever, simply the outcome of an interaction between two individuals. Any of a series of circumstances may intervene in the free expression of dominance between two individuals. The mother comes to the aid of her offspring during childhood and, at least among rhesus monkeys, into adult life as well (Sade 1965). Other individ-uals frequently join the combatants when a fight starts, and adults who tolerate each other will often support each other when one of them is challenged or attacked (DeVore 1965b). Thus, an individual’s dominance status is the result of a variety of factors, including past success in dominance encounters, the status of its mother, and the number and status of the individuals who will support it when challenged.

Threat behavior

Baboons and macaques, among whom thresholds of aggression are lowest and agonistic encounters most frequent, exhibit the most elaborate repertoire of gestures and vocalizations conveying anger, threat, and fear. One form of threat behavior which is common in all the Old World primates (including man) is the “redirection of aggression,” in which one animal threatens another and the threatened animal redirects the threat to a third individual or even to an inanimate object such as a stone or branch. For example, the loser of a fight between two adult males may chase a more subordinate male or a female or he may shake a tree limb vigorously. If he has chased a female, she will often chase or threaten a more subordinate female or a juvenile. This chain reaction of aggressive acts probably serves to periodically reinforce the various dominance statuses of all the participating group members.

Male dominance

Dominance relationships among adult males in baboon and macaque groups are often stable over many months or even years. A male’s status is ultimately based on his strength and fighting ability, and unless the group is very large, the males tend to be arranged in a linear hierarchy, or “peck order,” in which each male is dominant over all individuals beneath him. This stable hierarchy of recognized status serves to reduce the number of agonistic interactions among males. In some savannah baboon, rhesus monkey, and Japanese macaque troops, several males in their late prime may be so tolerant of each other that dominance between them is rarely expressed, and in such troops these males may constitute a nucleus in which the members support each other in dominance interactions. As a result, these older males, some of whom would lose a fight to a younger male in individual combat, are able to hold the young males in check by combined action (DeVore 1965a, pp. 54–71; Imanishi 1960). It is this nucleus of older supporting males that takes the initiative in protecting the group from outside danger, in repelling individuals who try to join the group, and in breaking up fights among group members. On the contrary, among gorillas and in small monkey groups this initiative is taken by a single dominant, “alpha” male. Dominance behavior is difficult to understand in captivity, where it seems to be disruptive and antisocial. But in natural surroundings, dominant males protect weaker group members, and females and juveniles seek out the dominant males to sit .beside or to groom. So, although the dominance hierarchy depends ultimately on force, it leads to relative peace and stability within the group.

Female dominance

The dominance hierarchy of females tends to be unstable and to depend to some extent on the’female reproductive cycle. Females in estrus are more active and aggressive and are frequently under the protection of a consort male. Among baboons and macaques, adult males are extremely protective of young infants, and this protection also extends to the mother as long as she is with a young infant. Thus, although it is clear that some adult females are more dominant than others in almost all circumstances, the interference of reproductive cycles in female-female dominance interactions seems to preclude the establishment of stable, linear dominance hierarchies.

Sexual behavior

As in humans, most female monkeys and apes have a menstrual cycle of about thirty days. Because humans do, and monkeys in captivity may, breed at all times of the year, the belief has been widely held that Old World primates show no seasonal variation in mating activity. Since the publication of Zuckerman’s The Social Life of Monkeys and Apes (1932), it has often been asserted that the basis of monkey and ape societies is sexual attraction; however, we now know that the non-human primate group performs many functions besides providing its members with sexual outlets.

There are many differences in patterns of sexual behavior between man and the nonhuman primates. In the first place, in at least some species of macaques, there are sharply defined mating and birth seasons. The data are best for the Japanese macaque and for the rhesus on Cayo Santiago, where records have been kept longest. Lancaster and Lee (1965) show that Old World monkeys tend to have at least a birth peak, if not a sharply defined birth season. Most monkeys give birth every year, and as in many birds and mammals, their reproductive cycles are presumably affected by such phenomena as day length and temperature. The reproductive cycle of some species, such as the baboons and the apes, is longer than a year, and if marked seasonality of sexual reproductive behaviorexists in these species, it has not yet been demonstrated.

Another reason that opportunity for mating is an inadequate explanation of social cohesion in nonhuman primate groups is the fact that the female is sexually receptive for a relatively small portion of her life. Females will accept a male only on those days of their sexual cycle when they are in estrus, and sexual cycles are inhibited during pregnancy and lactation. Females and their young constitute the core of monkey groups, and yet the amount of sexual activity in the life of any female, or within any group during the year, is very small. In a species with a mating season, disruptions are most common during this season. At this time ex-tragroup males attempt more actively to join the group, and the existing competition among males already in the group may become more intense. Ko-ford (1966) reports that on Cayo Santiago 66 per cent of all rhesus males who leave a group do so during the mating season. One of the Cayo Santiago groups subdivided into five groups between 1958 and 1964, each subdivision occurring during the mating season. On the other hand, social bonds within groups are probably strongest during the birth season, when mothers are clustered together with their new infants and adult males are actively caring for the infants of the previous year.

Mating patterns. In such species as the gibbon and the patas monkey (and in small groups of any species), only one adult male does the mating, but in most Old World primate groups the female is potentially accessible to every male in the group during some part of her estrous cycle. In such dominance-oriented species as baboons and macaques, the tendency is for the most dominant male to have priority of access to an estrous female, but his mating activity is usually confined to her period of maximum receptivity. For example, a female baboon is receptive for about 12 days in a thirty-day sexual cycle; the degree of receptivity can be gauged by the amount of tumescence in the sexual skin of the perineal area. During the early period of swelling she is mounted by juvenile and subordinate adult males, and it is only when tumescence is at a maximum that she is sought out by the dominant males. If there are a number of adult males of approximately equal dominance and only one receptive female, the males may threaten and fight for her possession. Although copulations early in her cycle are casual and transitory, dominant adult males usually form a consort pair with a receptive adult female. The pair may remain together for several days, or if there is competition among the males, the pairing may be interrupted after only a few hours. A baboon male’s ability to maintain possession of an estrous female when he is being threatened will depend in part on the support he can expect from other males (DeVore 1965fo). The female must cooperate with the male during copulation, and she is thereby able to exercise some personal preference and to take some initiative in making sexual contacts. Thus, mating patterns in even a sexually dimorphic species like the baboon are more than a simple reflection of the male dominance hierarchy. There are opportunities for subordinate males to copulate before the female becomes attractive to the dominant males; a male’s ability to mate may depend on his relationship to other, supporting males; and the female herself is able to exert some initiative.

Prolonged immaturity

By comparison to other mammals of similar size, the period of infant dependency is remarkably long in primates, and the necessity for providing care and protection of immature offspring is a dominant feature of monkey and ape social life. Ordinarily an infant in danger is retrieved by its mother, but adult males of many varieties of monkeys will threaten, distract, or attack a potential predator. Aggression toward predators is most prominent in ground-dwelling species such as baboons and macaques; for example, on an average day a baboon group may range as far as two or three miles from the safe refuge of trees.

The birth of an infant not only changes the quantity and quality of its mother’s social interactions, but every member of the group is affected in some measure. Adults and juveniles are highly attracted to young infants and repeatedly approach them to inspect, fondle, groom, and play with them. The newborn infant primate clings to its mother at all times, imposing substantial behavioral constraints not only on her but on the entire group. This prolongation of preadult life is so biologically expensive for the species that there must be major compensatory advantages. In a recent discussion, Washburn and Hamburg (1965) have suggested that this prolonged period of immaturity makes it possible for the species to adapt to a wide variety of environmental contexts because of the opportunity it provides for complex learning in the young. A large number of laboratory and field studies support the fact that much of the normal adult behavior of the species must be learned. In nature immature monkeys are in frequent contact with adults of both sexes and with their peers. Themajor activity of the young primate is “play,” and it is during these long hours of daily play that the juvenile learns both the features of its physical environment and the appropriate responses to the other animals in its social environment. The development of normal sexual behavior, the expression of effective patterns of affection and dominance, and the ability to rear an infant competently are all social skills which must be learned, and monkeys raised in isolation show major deficits in all of these behaviors (Harlow & Harlow 1965; Mason 1965). Social deprivation experiments also indicate how even less drastic events in the early life of a young primate reared under normal conditions may lead to the striking individual differences in adult temperament and social skills that are so frequently reported by research workers.

Mother-offspring relations

It has long been obvious to observers that the relationship between a mother and her offspring is the most intense and prolonged in primate life. During the usual field study it is impossible for the observer to know what genealogical relationships may exist between adults or between any group members who are beyond the weanling stage. Rejection by the mother when she weans her infant seems so firm, and the effect on the infant so traumatic, that the tendency has been to assume that their early close bonds have been severed permanently. (Moreover, the mother’s attentions are soon redirected to the care of another newborn infant.) But these conclusions are clearly incorrect, at least for some species, as prolonged observation of identified individuals has shown. Studies reported by Imanishi (1960) and Koford (1963) indicate that the rank of a young male in the dominance hierarchy is frequently very closely related to the rank of his mother; high-ranking mothers tend to rear high-ranking male offspring. The evidence is less clear for females, but it seems likely that high-ranking females may also be related as siblings or as mother and daughter.

Sade (1965) has offered the most detailed accounts of the importance of social bonds between uterine kin in rhesus monkeys. The most frequent subgroups are clusters of mature and immature offspring around an old female. These animals not only sit together and groom each other more often than they do other group members, but they frequently support each other in dominance interactions as well. Data indicating the importance of genealogically related subgroups are most convincing for the Japanese macaque and the rhesus on Cayo Santiago, where individual records have been kept for many years, but Jane Goodall (1965) has also described similar clusters of offspring around old female chimpanzees. Troops of Japanese macaques and rhesus monkeys which contain no adult males (except during the mating season) have been reported, but they are rare.

The achievement of high rank by the offspring of dominant females appears to be caused both by childhood experience and by the mother’s support of her mature offspring. Infants of dominant females stay with their mothers and other dominant individuals near the center of the group, and in the Japanese macaque they may never go through the period of social peripheralization that is characteristic of most young males. They are born in the middle of the group, grow up there, and remain there as adults, assuming at a young age the roles peculiar to the dominant males at the heart of the group. Also, in day-to–day agonistic interactions, a young male who is supported in aggressive encounters by a dominant mother will simply win more of these encounters than one whose mother is subordinate. Presumably the long period of conditioning during infancy and adolescence strongly reinforces a male’s expectation of success or failure in the social dominance interactions of adult life. The question of the effect of genetic inheritance on the achievement of social dominance should be investigated; this could easily be done by cross-fostering experiments, in which the offspring of a subordinate female and a dominant female would be exchanged at birth.

Sade reports that rhesus mothers on Cayo Santiago continue to support their adult sons in dominance interactions with other group members. Furthermore, a mother ordinarily remains dominant over her son all of her life. Apparently correlated with this behavior is the fact that copulation between mother and son is very rare. In one of the few known cases of mother—son incest, the two had engaged in a vicious fight, and the son had subsequently become dominant over his mother before the mating occurred. Although they are not yet well understood, it appears that the complicated patterns of nurturance and dominance behavior between mother and son may inhibit copulation between them and that the aversion to incest may be present among members of a number of Old World primate species.

Adult male-offspring relations

Paternity in nonhuman primate groups is rarely known or knowable either to the investigator or to the monkeys. Adult males seek to protect any infant from injury either from group members or from predators. However, in some species an adult male mayform a special protective relationship with a particular infant. In the Japanese macaque, for example, males frequently establish such a relationship with yearlings when these are rejected by the mother and replaced by a new infant (Itani 1959). Analogous behavior has been reported in other macaques and baboons (Washburn et al. 1965). Among some New World marmosets a male may carry an infant at all times except when the mother is nursing it. The one-male unit of the hamadryas baboon develops because young adult males forcefully adopt immature females, “mothering” the females until they reach sexual maturity (Kummer 1968).

There are no indications from field studies that the adult male in arboreal Old World primates establishes strong social bonds with infants or juveniles beyond protection of infants when the group is threatened by a predator. Even in the ground-adapted patas monkey and in the vervet monkey, the role of the male seems limited to protection or to distracting predators, and the present evidence indicates that more elaborate male-offspring relations arise only in terrestrial primates living in large social groups.

Human and primate social organization

Only a few years ago nonhuman primate behavior was assumed to be simple, stereotyped, and easy to observe; with notable exceptions (Carpenter 1964), field reports were largely anecdotal accounts by casual observers. Observations of zoo colonies suggested that primate groups were a disorganized rabble, tyrannized by aggressive males and held together only by sexual attraction. From the perspective of modern field studies, the conclusions based only on the behavior of primates in captivity bore as little relation to the social organization of free-ranging groups as would a monograph on middle-class society based solely on a study of inmates in a maximum-security prison.

Social organization is not identical for any two primate species. Human social organization, with its varied cultural manifestations, is unique; but studies of monkeys and apes are helping to distinguish those behavior patterns that are truly unique from those which man shares with his nearest relatives. Before the urban revolution human social organization was centered on small, face-to–face groups of adults involved in the quest for food and the protection of immature offspring, and it is not surprising that patterns of dominance and aggression and of nurturance and affection seem so similar in men, monkeys, and apes. The division of labor by sex, with females specializing in child care and males as protectors, and the performance of different social roles according to status are familiar patterns to the sociologist. The evolutionary basis of man’s unique biological and social system is treated elsewhere in this encyclopedia, but it seems appropriate to outline here some of the features of human organization which seem to distinguish men most clearly from the other primates [seeEvolution, article onHuman evolution].

Much of the contrast between human and primate social organization is based on human language. Communication in monkeys and apes is a rich complex of vocal, gestural, and tactile signals, but the majority of these signals are reflections of changing emotional states among the communicants and are comparable to the communication systems of other animals, not to language [see COMMUNICATION, ANIMAL; see also Marler 1965; Bastian 1965; Altmann 1967]. Language is so fundamental to human life that it is impossible to dissociate it from other distinctly human behaviors, such as toolmaking and the willingness to share food. Nonhuman primates do not share food, even with their own infants, but neither do they have tools to gather food or an improved site or camp from which they can disperse during the day in complementary gathering and hunting activities and where they can convene again to share the results of the day’s quest.

The most distinctive feature of human social organization, the family, is adaptive only in the context of an economy based on tool use and sharing; presumably it also requires other uniquely human traits such as the ability of the female to mate at all times. On the other hand, many behavior patterns that have been viewed as ramifications of the human family are actually present in the Old World primates—the maternal affectional system and its extension to uterine kin, the close association of siblings, the protection and nurturance by adult males of the young—and it is no longer so difficult to understand how the human social systems could have evolved. With language, an incest aversion between mother and son became an incest taboo; with language, the already close ties among uterine kin were explicitly recognized and socially extended.

Field and laboratory studies of primates were not begun on a large scale until the 1960s, but they are now attracting the attention of persons in biology, ethology, and all branches of the behavioral sciences. Techniques of observation and analysis are constantly being refined, and behavioral studies are being combined with experiments in pharma-cology, hormonal studies, and neurophysiology (Washburn & Hamburg 1968). For example, miniature telemetering devices implanted in the animal have been used to record and direct the behavior of individuals in a free-ranging group. Ethological principles and techniques of observation are being applied to humans and the results compared to studies of nonhuman primates [seeEthology]. There is every indication that in the next few years this renewed interest in the non-human primates will yield results of increasing relevance to the social sciences.

Irven Devore


Altmann, Stuakt A. (editor) 1967 Social Communication Among Primates. Univ. of Chicago Press.

Bastian, Jarvis R. 1965 Primate Signalling Systems and Human Languages. Pages 585–606 in Irven De-Vore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Buettner–Janusch, John (editor) 1963–1964 Evolutionary and Genetic Biology of Primates. 2 vols. New York: Academic Press.

Buettner–Janusch, John et al. (editors) 1962 The Relatives of Man: Modern Studies of the Relation of the Evolution of Nonhuman Primates to Human Evolution. New York Academy of Sciences, Annals 102: 181–514.

Carpenter, Clarence R. 1964 Naturalistic Behavior of Nonhuman Primates. University Park: Pennsylvania State Univ. Press.

Clark, W. E. Leghos 1960 The Antecedents of Man: An Introduction to the Evolution of the Primates. Chicago: Quadrangle.

Conference on Sex and Behavior, Berkeley, California 1965 Sex and Behavior. Edited by Frank A. Beach. New York: Wiley.

Crook, J. H.; and Gartlan, J. S. 1966 Evolution of Primate Societies. Nature 210:1200–1203.

Devore, Irven 1963 A Comparison of the Ecology and Behavior of Monkeys and Apes. Pages 301–319 in S. L. Washburn (editor), Classification and Human Evolution. Chicago: Aldine.

Devore, Irven (editor) 1965a Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Devohe, Irven 1965fc Male Dominance and Mating Behavior in Baboons. Pages 266–289 in Conference on Sex and Behavior, Berkeley, Calif., Sex and Behavior. Edited by Frank A. Beach. New York: Wiley.

Devore, Irven; and Lee, Richard 1963 Recent and Current Field Studies of Primates. Foiia primatologica 1, no. 1:66–72.

Eimerl, Sarel; and Devore, Irven 1965 The Primates. New York: Time Inc.

Goodall, Jane 1965 Chimpanzees of the Gombe Stream Reserve. Pages 425–473 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Harlow, Harry F.; and Harlow, Margaret K. 1965 The Affectional Systems. Volume 2, pages 287–334 in Allan M. Schrier, Harry F. Harlow, and Fred Stoll-nitz (editors), Behavior of Nonhuman Primates: Modern Research Trends. New York: Academic Press.

Imanishi, Kinji 1960 Social Organization of Subhuman Primates in Their Natural Habitat. Current Anthropology 1:393–407.

Itani, J. 1959 Paternal Care in the Wild Japanese Monkey, Macaco fuscata fuscata. Primates 2, no. 1: 61–93.

Jay, Phyllis 1965 The Common Langur of North India. Pages 197–249 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Jay, Phyllis (editor) 1968 Primates: Studies in Adaptation and Variability. New York: Holt.

Jolly, Alison 1967 Lemur Behavior: A Madagascar Field Study. Univ. of Chicago Press.

Koford, Carl B. 1963 Rank of Mothers and Sons in Bands of Rhesus Monkeys. Science New Series 141: 356–357.

Koford, Carl B. 1966 Population Changes in Rhesus Monkeys: Cayo Santiago 1960–1964. Tulane Studies in Zoology 13, no. 1:1–7.

Kummer, Hans 1968 Two Basic Variations in the Social Organization of the Genus Papio. In Phyllis Jay (editor), Primates. New York: Holt.

Kummer, Hans; and KURT, F. 1963 Social Units of a Free-living Population of Hamadryas Baboons. Foiia primatologica 1:4–19.

Lancaster, Jane B.; and Lee, Richard B. 1965 The Annual Reproduction Cycle in Monkeys and Apes. Pages 486–513 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Marler, Peter 1965 Communication in Monkeys and Apes. Pages 544–584 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Mason, William A. 1965 The Social Development of Monkeys and Apes. Pages 514–543 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Fetter, J. J. 1962 Recherches sur I’ecologie et 1’etho-logie des lemurs malgaches. Musee National d’His-toire Naturelle, Memoires Serie A 27, no. 1:1–146.

Reynolds, Vernon; and Reynolds, Frances 1965 Chimpanzees of the Budongo Forest. Pages 368–424 in Irven DeVore (editor), Primate Behavior: Field Studies of Monkeys and Apes. New York: Holt.

Sade, Donald S. 1965 Some Aspects of Parent-Offspring and Sibling Relations in a Group of Rhesus Monkeys, With a Discussion of Grooming. American Journal of Physical Anthropology New Series 23:1–17.

Schaller, George B. 1963 The Mountain Gorilla: Ecology and Behavior. Univ. of Chicago Press.

Schrier, Allan M.; Harlow, Harry F.; and Stollnitz, Fred (editors) 1965 Behavior of Nonhuman Primates: Modern Research Trends. 2 vols. New York: Academic Press.

Southwick, Charles H. (editor) 1963 Primate Social Behavior; an Enduring Problem: Selected Readings. Princeton, N.J.: Van Nostrand.

Struhsaker, T. 1965 Behavior of the Vervet Monkey, Cercopithecus aethiops. Ph.D. dissertation, Univ. of California.

Symposium on The Primates, London, 3962 1963 The Primates. Edited by John Napier and N. A. Barnicot. The Proceedings of the Symposium of the Zoological Society of London, No. 10. London: The Society.WASHBURN, S. L. (editor) 1963 Classification and Human Evolution. Chicago: Aldine.

Washburn, S. L.; and Hamburg, David A. 1965 The Study of Primate Behavior. Pages 1–13 in Irven DeVore (editor), Primate Befiauior: Field Studies of Monkeys and Apes. New York: Holt.

Washburn, S. L.; and Hamburg, David A. 1968 Aggressive Behavior in Old World Monkeys and Apes. In Phyllis Jay (editor), Primates. New York: Holt.

Washbuhn, S. L.; Jay, Phyllis C.; and Lancaster, JANE B. 1965 Field Studies of Old World Monkeys and Apes. Science New Series 150:1541–1547.

Zuckehman, Solly 1932 The Social Life of Monkeys and Apes. New York: Harcourt.


Ecologists have been studying the dynamics of animal populations for roughly forty years, trying to discover how their numbers are regulated. The subject is of much practical importance in a world in which man, either knowingly or unknowingly, interferes drastically in this process and in which our own exploding population has become a central threat to human survival. Scientific knowledge in this field has reached a stage at which generalizations are possible, although final agreement has not been reached.

Populations of animals obviously cannot all have the same life-cycle characteristics. Many of the insects run through several generations in a single summer, after which the population remains dormant during the winter and regenerates in the spring. Larger forms of animals, on the other hand, tend to live longer, and most of them have populations that keep going all the year round. Some populations fluctuate in numbers from season to season; others remain practically constant from year to year, even from decade to decade. Striking examples of the fluctuating kind include the arctic lemmings, many of the northern game birds and fur bearers, and the subtropical locusts. The steady state typically applies to most kinds of big game, many birds, and other animals both large and small, especially in temperate and tropical regions with equable climates. There is a continuous gradation between the extremes, and under natural conditions the populations of most animals in most situations resemble one another in one very important respect, namely, that there is no long-term trend either up or down.

The normal situation is one in which the density of population moves or fluctuates irregularly, either weakly or strongly depending on circumstances, to a varying distance on either side of a constant average value. If and when permanent changes take place in the environment, the average is also likely to shift and will assume a new level once the environmental change is complete. Therefore, we need to find out what determines the average, what causes the fluctuations and prevents them from swinging out of control in one direction or the other until the population either approaches infinite size or falls to zero.

Population income and loss

Animals must in general possess the fecundity necessary to augment their numbers; otherwise they would quickly die out. In the course of time local populations receive an income of new recruits, largely as a result of reproduction, although some may come through immigration. They are also subject to toss, some of it by emigration and the rest through mortality. If the population stays roughly constant in number over the years, it means that the cumulative totals of recruitment and loss have balanced out.

The fact that populations tend to maintain a constant average level and that when fluctuations occur they do not proceed very far in one direction or the other without being checked indicates that there must be some stabilizing process at work. For example, if the recruitment rate were always 20 per cent per annum and the death rate 25 per cent, the population would steadily diminish to the vanishing point; similarly, if the two rates were reversed, it would expand at 5 per cent compound interest. But let these rates vary according to whether the population density is high or low— so that high density produces low recruitment or high mortality or both, and low density the opposite—and then any continuous trend of numbers up or down can always be prevented. The phenomenon typically found in nature, of population density balanced within relatively narrow limits of variation, cannot in fact be accounted for in any other way than by regulators that operate in this manner and are “density dependent” in their action.

There is one important school of thought which believes that density-dependent action arises mainly from the factors causing losses in the population. According to this hypothesis, animals are supposed always to reproduce as fast as they efficiently can, the surplus so produced being removed in a density-dependent manner by the various destructive agencies, namely predators, parasites and disease, starvation, and accidents of the physical world, such as floods and droughts. On the face of it, there seems nothing impossible or improbable in this: disease, for example, would be expected to spread fastest in overcrowded conditions, and starvation would have a similar incidence. Therate of accidental mortality seems perhaps the factor least likely to have any connection with population density.

There are, however, many large animals, of which primitive man was a good example and lions and eagles are others, which effectively have no predators and appear not to be regularly subject to death from disease on a significant scale, except perhaps among old or outcast members of the population. We find ourselves, therefore, forced back on starvation as the only remaining density-dependent agency that might be capable of removing a population surplus. This possibility calls for close examination.

Population density clearly has a very special relationship to food supply. Animals “select” their environments or habitats in accordance with their natural inclinations. The habitat selected has to provide them with the resources they need for sustaining life, including food of adequate quantity and quality, space, shelter, and breeding sites. In the great majority of species, food is the resource that sets the upper limit on population density, and determines the “carrying capacity” of the habitat. Habitats can therefore generally be graded as richer or poorer almost entirely in terms of their food productivity, and the average population density of animals living in them has been shown, in situations where both can be measured, to be closely related to it.

However, the correlation between food supply and population density has no direct connection with starvation. The normal condition is that all individuals living in a particular habitat get enough to eat. If starvation occasionally occurs, it almost always comes as an emergency resulting from an accidental cause, such as exceptional cold or drought, and is not in any way dependent on the density of population that happens to be affected by it.

Food resources harmed if overtaxed

Human experience in exploiting purely natural resources of animal and plant life, whether they happen to be the virgin pastures of the prairies or fish or game, shows that most of them can all too easily be overtaxed. Persistently overgrazed pasture quickly deteriorates and especially in arid climates can be reduced to a worthless desert in a few centuries. If we take too big a crop from our fisheries, whether for commerce or sport, or press too hard on fur bearers, wild fowl, or game, the stocks and the yields decrease. Unfortunately, long after the demand has become excessive, it is almost always still possible to go on making money, and we tend therefore to shut our eyes to the increasingly disastrous cohsequences piling up for the future. This highly characteristic danger inherent in the exploitation of “recurrent resources” has often led in the past to the decimation of the more lucrative and vulnerable prizes of the hunter, such as the passenger pigeon, the northern right whale and the blue whale, the fur seal and marine turtle, or the rhinoceros of Africa and Asia, now threatened because their horns are highly valued by people in the Far East for their supposed aphrodisiac powers.

In exploiting their own sources of natural food, animals face exactly the same dangers as man. For the same reason, therefore, they must exercise restraint while there is plenty, if that plenty is to be conserved and future yields sustained.

This is a deduction of great importance, for it implies that starvation can be ruled out as a practicable means of curbing population growth. Long before starvation occurred, disastrous damage would have been done to the food resource by over-exploitation, and the survival of the entire population would be threatened.

It can be concluded, therefore, that not only are there many kinds of animals which are not subject to predation or disease on an effective scale, but that starvation can also be eliminated as a practical regulating factor, because damage to the habitat is a precondition and, consequently, constitutes a continual threat to the safety of the group as a whole.

Regulation through “conventions”

It is now necessary to explore the alternative hypothesis that animals themselves possess the means of controlling numbers and population density. If they were able to do this, it would obviously be an adaptation of great value, because an ability to control population density would imply that it could be adjusted to some particularly favorable level, and that level could very well be the optimum one, as far as exploiting the food resource is concerned. Ideally, it would then be possible to take the maximum sustainable crop from the food resource and always feed the largest number of mouths that could be safely allowed without ever running the risk of overtaxing or depleting the supply.

It is a common situation for an animal population to depend for a period of weeks or months on a standing crop that is not at the same time being replenished. This applies, for example, to birds feeding on seeds and berries in the fall or hunting for hibernating insects during the winter. When they first broach the food stock, it is atmaximum abundance and for perhaps hours or days could provide food for a prodigious number of exploiters. But it must be made to last much longer than that, and somehow the number or population density of exploiters must be restricted from the outset. The food itself will not restrict them, because it is superabundant; and yet the reason for imposing the restriction is related to food and food alone. What is required is some artificial method of limiting the population density, one that is, if possible, closely correlated with food production, serving as a buffer or shield in preventing overexploitation: something, in fact, closely analogous to the restrictive conventions entered into by states and nations to prevent commercial overfishing in the sea.

It is at once apparent that a great many conventions of this kind exist in the natural world. The best known of them is the territorial system adopted by many kinds of birds during the breeding season. In the case of robins, cardinals, song sparrows, and a host of other birds, the males of a species compete with each other, each claiming and defending a territory on which pairing, nesting, and rearing of a young family subsequently take place. The males will not tolerate the crowding of territories beyond a certain limit, so that once a particular habitat has been taken up in claims, no further males can get in; any that attempt it are soon forced to leave and try somewhere else.

This is a perfect example of a conventional density-limiting mechanism, and one easily understood. What has happened is that direct contest between rival members of the group for actual food has been sidetracked into equally furious competition for a piece of ground—a conventional substitute. Once the ground is “owned,” the owner and his dependents can use the food that the territory contains in perfect freedom, without further dispute, and provided the conventional territory size is large enough, there need be no danger that their demands will overtax the supply.

In a short article one cannot hope to do more than suggest the broad implications of this hypothesis. It must suffice to say that many of the density-limiting conventions that can be observed in nature are less direct or more sophisticated than the simple territorial one. For example, birds that depend on the resources of the ocean cannot stake individual territories for gathering or nesting on the sea itself. Instead, they breed in a colony on shore which in most cases is quite artificially limited in size and in which each successful pan-gains a token territory in the form of a small nest site. Adults that have not succeeded in winning a nest site take on the status of onlookers and are frequently found at seabird colonies. They are evidently inhibited from nesting outside the traditional boundary or from starting a new colony around the corner. This conventional recognition of a colony is presumably related to the food supply, limiting the local population to a size that will be permanently immune to overfishing; however, as compared to the territorial system, the form it takes is more abstract or artificial because the object competed for is reduced to a few square feet or a hole or crevice, rather than the entire food-supplying territory. Abstraction can go even further than this. In some cases it is only the right to membership in a group that is recognized; the alternative is rejection as an outcast. Thus, there may be no contest for actual property at all, but only for individual or personal status.

These conventions are normally concerned with competition between the members of the population of a single species of animal: that is to say, with intraspeciflc competition. This competition provides conventional goals or rewards that are substituted for actual food, and their effect is to limit the number (or density per unit of food) of members in the group and to unload any unwanted surplus to a safe distance.

It is important to note that the competition itself assumes artificial or conventional forms. Birds contesting for territories seldom fight and kill each other, although winning a territory can in fact easily be a matter of life or death as far as the respective consequences to the victor and vanquished are concerned; instead they sing and threaten by voice or display fine crests or splendid or startling patterns of plumage.

Social behavior provides the key

It takes but a short step at this point to reach the most striking and important generalization to emerge from the new hypothesis: Any group of individuals competing for conventional prizes by conventional methods automatically constitutes a society. Behind this there is the deeper implication that here is the essence or root from which all social behavior has stemmed: that the evolution of a primordial social organization occurred as an adaptive mechanism for the control of population density. A society could be defined as an organization capable of providing conventional competition among its members; in these terms, sociability is seen to have a biological basis. One does not need to ponder very deeply to see how closely this cap fits even in human social groupings, where leaders emerge even though individuals need not be engaged inany direct form of competition. A society, whether human or animal, is in essence a brotherhood tempered with rivalry.

Inasmuch as it incorporates the mechanisms of competition, social organization appears to be indispensable in the process of maintaining an optimum population density. This is readily apparent in a territorial system, like that of birds, where neighbors of the same species are in constant social competition and by their behavior directly control the population density in the habitat. The long-term research on the Scottish red grouse being carried on near Aberdeen has shown, for example, that the grouse population on a heather moor consists of individuals all mutually known to their neighbors and greatly different in social standing. Dominant cock birds hold territories almost all the year round, covering the habitat like a mosaic; the most aggressive have the biggest territories. Socially subordinate cocks and unmated hens are often carried as members of the population although there is no room for them to establish any territorial claims for themselves. With the onset of winter, or when the productivity of the food supply declines for any other reason, it is the subordinates at the bottom of the social ladder that get squeezed out, leaving behind a surviving group adjusted in density to suit the lowered food level.

The social hierarchy, or peck-order system, thus works as a safety valve, or overflow mechanism, identifying and unloading those who cannot be supported when there is a drop in food supply. If the population were not thinned down in this way, damage would be done to the habitat by overtaxing the resource. By means of the social hierarchy not only is damage prevented but the remaining population is assured of an adequate diet and a reasonable chance of survival.

The hierarchy also operates in relation to breeding. Only established territory owners and their mates can breed, although unestablished non-breeding adults can still be carried in the population and form a valuable reserve to fill any gaps that appear in the ranks of the establishment. It may be noted that the hierarchy is another well-known biological phenomenon for which no general functional explanation has been given, although it finds an automatic place under the new theory.

Characteristically, red grouse compete under conventional rules, one of which is that, at least in the fall, the competition takes place only between dawn and about 8 A.M. After that the social groups flock amicably, and all feed side by side for the rest of the day. So fierce is the aggression among the cock birds at dawn, however, that it is enough to subject some of the less-spirited competitors to a stress that not only drives them away from the habitat altogether but leaves them defenseless to predators and liable to succumb quickly to disease.

The convention of competing at dawn, and often at dusk, and leaving the rest of the day free for other necessary activities like feeding is exceedingly common and appears to shed a flood of light on the timing of such familiar phenomena as the cockcrow and the dawn chorus of birds, the dusk flighting of ducks, the massed maneuvers of starlings and blackbirds at their roosts, the synchronized choruses of some kinds of tropical bats and cicadas, and the activity of marine life such as the croaker and the snapping shrimp; there are innumerable other analogous phenomena in which an indication of present numbers seems to be provided through synchronized communal activity. Dawn and dusk are of course the two most easily perceived hours of the day, and this no doubt explains their wide use for brief social activities that call for a simultaneous general participation in order to achieve their effect.

Homeostasis and epideictic feedback

Community displays and contests provide the essential index of population density from day to day, producing “information” which can lead to an appropriate response when an adjustment to density is required. The whole process of balancing the population against the changing food productivity of the habitat appears to be a “homeostatic,” or self-stabilizing, one, based on a feedback of information that sets the wheels of the adjusting machinery in motion whenever the balance has swung away from the optimal level. Community displays can not only lead to the expulsion of a population surplus, as in the case of the red grouse, but when the breeding season comes round, they can also serve to indicate how large a quota of recruits will be required to replenish the population or build it up to the particular level that is justified by the current carrying capacity of the habitat.

We need a technical term to describe this kind of activity, whose purpose is to sample population pressure, and “epideictic” has been chosen as having the right derivation and meaning. One could easily predict that epideictic displays would be especially conspicuous before the breeding season, when the population is about to be built up to its annual peak level; indeed, this is the time at which some of the displays already mentioned reach their height.

At that time in many species only the males arerequired to take part in communal displays, each commonly acting as the representative of a mated pair; the females are otherwise occupied, frequently in more domestic duties. In these circumstances the males can be described as the “epideictic sex.” The stimulation they receive from competing with rival males or sharing in exciting communal activities is density-dependent in its intensity: the more males there are, the hotter the pace. Rather frequently they assemble in unisexual swarms or groups, dancing in the air, perhaps, if they are flying insects or taking part in ritual tournaments, gymnastics, or parades, in the case of sage grouse and prairie chickens, tropical hummingbirds, manakins, and birds of paradise. They may have special vocal organs, as in songbirds, cicadas, frogs, drumfish, most crickets and katydids, howling monkeys, and many others. These are primarily used not to woo and win females, as was formerly thought, but to gain the prizes of social competition in the form of real estate and personal status. Exactly the same use applies to adornments, weapons, and scent glands. Much rethinking is therefore required on the hitherto vexed subject of sexual selection.

Epideictic displays are especially prominent among migratory animals as a prelude to their departure and as a means of smoothly handling heavy traffic over the vast tracts of country that often separate summer and winter quarters. This is especially true of birds. Without such displays the food resources would tend to be dangerously overtaxed in congested stopping places. The excited flocks of swallows and swifts, the spectacular massed maneuvers of locusts building up for a great emigration flight, and many of the big gatherings of birds, fruit bats, and insects to nightly roosts are associated with impending changes in population density through migration.


In the light of this theory of self-regulation of animal populations effected through the agency of social conventions, we can turn back once more to the original questions and see how they have been answered. The average population level would appear to be set for most species by the size of sustained crop that the food resources can yield, and this level is actually achieved through a code of conventional behavior that prevents the growth of population from going above the optimum density. Some fluctuations in population level are produced by accidental events in the environment, including predators and disease, which may cause extensive mortality; but in most cases they appear to arise from the operation of the homeostatic machinery, which automatically allows the population density to build up when resource yields are good and thin down when yields fall below average. The homeostatic machinery responds directly to the pressure of social competition. The latter is part and parcel of the struggle for existence and depends in turn on food availability and the number of mouths to be fed. It mirrors (inversely) the standard of living prevailing at the time. In this, therefore, lies the answer to the question, What establishes the balance of numbers and prevents the progressive long-term increase or decline of populations? Fluctuations tend in practice to be checked by homeostatic processes, which can control the recruitment rate, the pressure to emigrate, and, sometimes on an immense scale, the amount of mortality resulting from stress. Only when there are long-term alterations in the environment itself will the average population levels undergo permanent change.

This sketch can give only an imperfect picture of the wide ramifications of the hypothesis. There is still one most important point to be mentioned, which is that man himself is conspicuously out of line with all the higher animals in being exposed to progressive population growth. The history of this seems perfectly clear. In many animals, including some of the mammals, fertility itself is under density-dependent control; overcrowding automatically leads to lowered, and finally suspended, recruitment. In primitive man, on the contrary, fertility was limited not through the internal effects of hormones on the rate of ovula-tion or on the resorption of embryos in the uterus (as in rabbits, foxes, deer, etc.) but through cultural conventions, customs, and taboos which imposed restrictions on sexual intercourse for mothers who already had a child at the breast or prescribed compulsory abortion, infanticide (especially of female infants), human sacrifices, or head-hunting expeditions. These had evolved in prehistoric man as part of a socially integrated control mechanism which kept populations nicely balanced against the carrying capacity of the habitat, exactly as in the higher animals. Then, eight or ten thousand years ago, the agricultural revolution began. Food productivity increased by leaps and bounds, and power and wealth accrued to those groups which allowed their populations to multiply, settle villages, and build towns. The old checks on population growth were insensibly for-gotten and discarded, and none have since taken their place. We should not therefore look to nature to halt the human population explosion: it must depend on our own efforts.

V. C. Wynne–Edwards

[Directly related are the entries Ecology; Population. Other relevant material may be found in Evolution; Homeostasis; War, article on PRIMITIVE WARFARE.]


Jenkins, David; Watson, Adam; and Millek, G. R. 1963 Population Studies on Red Grouse, Lagopus lagopus scoticus (Lath.), in North-east Scotland. Journal of Animal Ecology 32:317–376.

Wynne–Edwards, V. C. 1962 Animal Dispersion in Relation to Social Behaviour. Edinburgh and London: Oliver & Boyd; New York: Hat’ner.

Wynne–Edwards, V. C. 1963 Intergroup Selection in the Evolution of Social Systems. Nature 200:623–626.

About this article

Social Behavior, Animal

Updated About content Print Article