Social Primatology (Social Psychology and Social Relevance, Chapter 2)

The howling monkeys of Barro Colorado Island are especially attractive to social psychologists because their natural social life is protected from the destruction that hunters have delivered to their mainland brothers and sisters. Since being made wards of the Smithsonian Institution early in the 20th Century, the BCI howlers have been protected from the whims of fortune and the attacks of human Panamanians. This chapter is partly a report of my observations of howlers in the summer of 1959, when I was part of a field research expedition headed by Professor C. Ray Carpenter of Penn State University, and partly a broad review of the research literature on the social life of howlers and other nonhuman primates up to the early 1970s.

Alan C. Elms

dumbbull monkeys: the howlers of Barro Colorado

These Creatures keep together 20 or 30 in a Company, and ramble over the Woods; leaping from Tree to Tree . . . chattering and making a terrible Noise; and a great many grim Faces, and shewing antick Gestures.

— Captain William Dampier Voyages and Descriptions (1700)

One night when he was twenty-five, my father went out into the hills of north central Arkansas with a giant dumbbull. A dumbbull may be made from an ordinary tin can, but he had stretched a dried groundhog skin over one end of an empty nail keg, A rosined string, with a knot at one end to hold it in place, was threaded through the center of the skin and out the open end of the keg. Sliding his fingers tightly along the string, he could make the dumbbull roar like a wildcat or a bull elephant; in clear weather, the noise would carry for a mile.

That first night, my father moved out along a ridge several miles from town, pulling a scream from the dumbbull every few minutes. Horses and mules began to snort all over the countryside. A family living at the foot of the ridge ran half a mile to a neighbor’s house, afraid the monster was breaking into the stable to eat their cow. The next night my father traveled in another direction. A farmer tried to sic his coon-hounds on the monster, but at one fearsome roar from the dumbbull, the dogs turned tail and ran under the house, while the farmer ran inside and locked the doors. After that, people began to talk about getting up a posse to hunt the crittur down, and word came from other counties about seeing strange beasts. When people got so scared they wouldn’t go out at night to help sick neighbors, my father admitted that he and his dumbbull were the “monster.” He stayed out of public view for a while, but most people took the hoax good-naturedly. They were happy it had been only a dumbbull.

The first time I heard a howler monkey clan roaring at the dawn on Barro Colorado Island, I remembered my father’s dumbbull. The howler’s full-throated vocalization is almost mythological in quality, the kind of echoing cry you’d expect from a banshee or a maddened minotaur. It makes the Panamanian rain forest sound like a real jungle. The howl has been recorded on small tape recorders, but it demands the sound system of a modern hard-rock group. If you’ve never heard a nail-keg dumbbull, try to imagine the noise of a pack of increasingly agitated beagles cornering a great barrel-chested lion, with an electronically synthesized Tarzan yodeling to the rescue.

The howler monkey is not barrel-chested but barrel-throated. An adult male weighs twenty pounds or less, but he has a built-in resonator, an unusual development of the hyoid bones of the throat. As the early evolutionist A. R. Wallace wrote (1878), “It is remarkable that this one group of monkeys should possess an organ not found in any other monkey or even in any other mammal, apparently for no other purpose than to be able to make a louder noise than the rest.” It’s more remarkable still, in that howlers have no obvious reason to make a louder noise than the rest; they usually live in an environment very like a Monkey Eden. This is particularly true of the howlers who have been most studied — the black howlers of Barro Colorado Island, in the Panama Canal Zone.

Barro Colorado has too much rain for paradise; when the rain comes, the howlers hunch their backs and bend their heads and roar at it, as if it were an animate evil. But when the rain stops, little else troubles them. Leaf and fruit eaters, they find it hard to starve in the midst of Panama’s heavy vegetation. The high trees in which they live protect them from most potential enemies. The cecropia leaves that are a favorite food would serve better than fig leaves, were the howlers blessed with modesty; they are not. The adult male’s scrotum shows a clear unmistakable white against his dark fur and the green foliage; the female, although she lacks the vivid sex skin of certain old-world primates, manages to make her needs known unashamedly by waggling her tongue in and out of her mouth, in a sort of remote French kiss. The usual ratio of adults (more than two females to each male), plus their unjealous promiscuity, guarantees a male for every female when she’s in heat, and a female for every male when the urge moves him from lethargy.

As the waters of the Panama Canal rose around the mountain ridge that is now Barro Colorado Island, trapping the monkeys there for good, most large predators escaped, and almost none remain. The most serious threat to howlers, the human hunter, is banned from the island, which is a United States Government preserve. Botfly larvae swell the necks of infants, and older monkeys often wear botfly scars; yellow fever dropped hundreds of howlers during the late 1940’s. But since then, the Barro Colorado howler population has exploded dramatically.

Curiosity, a serious threat to any paradise, seems minimal in howlers. They are imperturbable in the presence of sloths and antcaters, and usually look merely anxious, not inquisitive, about human observers. The effectance motivation described by Robert W. White (1963), which moves humans and perhaps certain apes to explore and manipulate their environment, hardly if ever moves the howler; when he is not eating, he mostly rests or sleeps. To his physical near-Eden, he has added (via evolution) an important element — not better means of need satisfaction, but the restriction of his own wants. Eden is not Eden when Adam desires what he cannot have.

But it is not their nearness to the angels which has made the howlers so attractive to social psychologists. Rather, it is the fact that they are social beings. Other creatures are social as well, from ants to springboks. But howlers and other primates seem to humans particularly social because — as they loll in the trees, snack frequently, grumble at each other occasionally, discipline the children when necessary — they look so human.


The howlers of Barro Colorado Island are especially attractive to the scientist because their natural social life is protected from the destruction that hunters have delivered to their mainland brothers. Since being made wards of the Smithsonian Institution earlier in this century, the Barro Colorado howlers have been protected, like the United States residents of the Canal Zone, from the whims of fortune and the attacks of the Panamanians. Barro Colorado Island is, in this way, a somewhat unnatural environment, but perhaps less so than the howlers’ mainland habitat, with its recent (in the evolutionary time scale) addition of guns and traps and poisons.

This protected status docs not exactly make Barro Colorado a paradise for human researchers. A boat dock, laboratory buildings, and sleeping quarters have been built on one little cove, and several trails have been cleared; but the rest of the 4,000 acres remains undisturbed tropical rain forest. In some areas the trees overhead are so thickly interlaced that little underbrush can grow, and travel away from the trails is easy. In other areas, the brush or the meaty serrated spears of pita are so thick that passage even with a machete is nearly impossible. The howlers have the better idea of how to travel: at treetop level, they miss the occasional coral snake or boa constrictor, the scorpion or tarantula, and all the little bright-colored spike-backed spiders who hang their webs between the tree trunks.

Fortunately, the howlers move through the trees with much caution, so it’s usually easy for an earthbound human observer, even as he dodges spiderwebs, to keep up with them. Their slowness, their noisiness, and their togetherness make it possible to study howler behavior under natural conditions more extensively than can be done with many other primates.

The first important effort to unite the study of comparative psychology with social psychology began on Barro Colorado Island in 1931. Animals have been much used for research by psychologists, but usually only in studying elementary responses — reflexes, simple learning, basic drive-reduction behavior. Until 1931, information about the social behavior of man’s closest surviving relatives, the nonhuman primates, was either grossly unreliable, overly simplistic, or descriptive only of abnormal behavior in zoos. Some writers leapt to make generalizations about man and society from these atrociously inadequate data. Others showed their wisdom by keeping quiet.

Late that year a young psychologist, C. Ray Carpenter, went to Barro Colorado to study howlers. Rather than looking around haphazardly for what seemed momentarily interesting or supportive of a current theory (as in previous “scientific” primate observation), Carpenter spent hours a day observing a single group at a time. For several months he systematically surveyed the island’s entire howler population. An indication of his thoroughness is that although an expert naturalist had estimated the Barro Colorado howler population as seventy animals in 1929, Carpenter was able to identify about four hundred howlers by the end of his survey in 1932, and added a hundred more during further study in 1933. Carpenter not only counted the animals (a harder task than it might seem, when you’re classifying adult males and females, plus six categories of juveniles), but also made careful behavioral observations, largely through field glasses, and studied the howler’s environment (Carpenter, 1934). He returned again briefly in 1935, then turned to studying other primate species. The howlers were left to themselves until the 1950s, when they were twice studied by other scientists; then Carpenter returned once more in the summer of 1959, with several associates and student assistants, including myself (Carpenter, 1965). Others have come since.

So the howler’s social life has been studied scientifically over a longer period than any other primate’s except man’s. Observations of its behavior have raised complex questions that remain unsettled among behavioral scientists, but we can start with a few simple ones. How social is the howler? How does it show its socialness? And is the howler distinctive among primates for anything besides, possibly, its noisiness?


First, howlers are definitely clannish. An adult male can occasionally be seen sitting alone on a high branch in the midst of the jungle, distant from any group, but such isolated animals are rare. Even their isolation may be temporary; one may sometimes be seen tagging after a group, as if he were trying to get in, or to get back in. Most howlers are born, live, and die within the same small group — provincials who are much disturbed by encounters with strangers or loners of their species. (At the most recent full census, Barro Colorado Island held forty-four groups, totaling over 800 howlers; but only a half dozen or so isolated animals were counted.)

The groups are stable, and seldom split. Splitting must sometimes occur, since the number of groups on the island has increased along with overall population growth; but the fission process has never been observed by humans. Groups do not appear to exchange members, unless those rare isolated males are in the slow process of transfer. In general, group composition changes only through birth and death. Varying group sizes have been observed, from 2 to 45; the mean is roughly 18 monkeys. Carpenter has calculated the average composition of a Barro Colorado howler group to be 3 adult males, 8 adult females, 3 dependent infants, and 4 juveniles. These proportions have remained roughly the same for over thirty years, longer than the probable life span of any howler, though mean group sizes have varied as a result of disease and environmental change (Chivcrs, 1969).

Talking about a howler “group” (or “clan,” as Carpenter prefers) may imply both more and less than it should. A howler group isn’t just a random collection of animals in a particular area. The group is usually relatively compact, every animal within a few yards of every other animal in the same tree or adjacent trees. When the group is moving, individuals may sometimes be strung out over several dozen yards; but like a progressing amoeba or a ball of silly putty, the group will contract again into a relatively tight knot of monkeys at feeding or rest time. If most of the group begins to move before one monkey finishes eating, the one will usually follow the group rather than eat.

On the other hand, the group is not overly organized. More often than not, adult males lead and bring up the end of a group progression through the jungle; males consistently initiate moves from one tree to another. But the adult males within a group seem to share this “responsibility”; firm patterns of dominance and submission, or exclusive sexual relationships, or indeed any stable relationships (except between a mother and her immature offspring) have been hard to find. Further, the group is not in constant close social interaction, even of an unorganized nature. This is partly because howlers don’t engage in veiy frequent action of any kind. Carpenter kindly refers to “low energy expenditures,” but the plain fact is that in relation to many other species, howlers are lazy. In general they toil not, neither do they spin — except sometimes slowly, while hanging from their prehensile tails. The action in which they do engage is mostly self-oriented: they feed themselves, they scratch themselves, they sun themselves and take long naps, with little overt reference to other members of the group.

A group’s typical day would involve a dawn awakening; a few howls at the sunrise, or at the howls of distant clans; scattered feeding and then perhaps a slow move to a new food tree; concentrated but leisurely feeding during mid-morning; a long midday siesta, during which the adults sleep or rest and the juveniles may play sporadically; possibly another move, probably another two or three hours of feeding in mid-afternoon, unless heavy rain interferes; then a period of settling down in a lodge tree for the night, with virtually complete inactivity from 7 or 8 p.m. until the next morning. The most remarkable aspect of all this behavior, for an animal who is distinctly a social being, is how little of it can really be called social. Much of it might best be described as conjunctive: the behavior of maintaining a fairly close proximity with other animals of the group, without directly orienting one’s behavior toward them or acting in response to their behavior. Conjunctive behavior could be compared to a quiet human evening at home — the husband reading his newspaper in his easy chair, the wife sewing or knitting on the couch, and both getting a certain satisfaction from being in the same room, though they may hardly speak to each other for several hours. The howlers, however, don’t even go off to bed with each other at evening’s end. They may sleep in the same tree, but they sleep on separate branches.


Occasionally, during all this passive and rather disinterested existence, the individual howler monkey may initiate or be on the receiving end of a specific social act. If it’s in any way positive, the action will most likely come from his mother. If it’s somehow negative, it’s likely to come from another howler group, rather than from within his own clan.

I use the vague word “positive” because so little howler behavior can be called truly “cooperative.” The howler mother gives her infant sufficient attention to keep him alive, and that’s about it. She lets him hang to the hair on her belly or back; as he grows older and begins to explore the tree-world, she’ll retrieve him; when she’s ready to move on, she lets him know with a small noise, so he can climb on. If he gets left behind, she’ll come back for him; when he gets too big to ride on her, but not big enough to make every tree crossing safely, she’ll even stretch herself between two branches to make a monkey bridge for him. The small infant is nursed, of course; but when he’s old enough to eat solid food, his mother will only rarely (and perhaps even then only accidentally) pull a branch or leaf down to let him feed.

Mothers also protect their infants from the attentions of other females, who engage in what has informally been called “nursemaid” or “auntie” behavior. The newborn howler is of great interest to the group’s infant-less females. They may simply remain close to the mother for hours at a time, watching; they may nose or touch the infant; they may try to pick it up. Usually all this attention is quite unhelpful and even annoying to the mother, who never asked for a nursemaid in the first place. But once I did sec an infantless female place her arm around and her body partly over a mother and infant as a summer rain grew harder.

Juvenile howlers occasionally play with each other, in a simple-minded way — chases, wrestling, tail-pulling. Adults are involved in play onlv when they’re the butt of a juvenile joke. Most nonhuman primates, particularly those of the Old World, socialize extensively by grooming each other (picking bits of dirt, dead skin, and so on out of each other’s fur with mouth and fingers). Howlers seldom groom; even mothers groom their infants very little. Since the grooming serves in other primates to remove parasites, its absence in howlers may explain why the howlers have such a hard time with botfly larvae infestations. Extensive social interaction may be unnecessary for howlers in most regards; but in this case, it appears they could use more than they have. (Melvin K. Neville [personal communication] observed an apparently higher incidence of grooming among red howlers in Venezuela, and also observed no botfly problem. There may, however, simply be fewer botflies in the Venezuelan habitat.)


Hostile (or, more technically, agonistic) behavior occurs within the clan even less often than positive behavior — sporadically at most. The low incidence of hostility may result partly from the way it’s handled at first appearance, in juvenile play-fighting. Juveniles usually stop nipping and slapping when they begin to get hurt; but if their fighting becomes serious, an adult male will usually emit a sharp grunt, at which point the juveniles cease all action. It isn’t clear whether their response to the male’s grunt is entirely learned or has an innate basis; but frequent repetitions of the grunt-and-stop sequence among juveniles may be sufficient to instill in howlers learned limits to adult aggression against their own clansmen.

Among the group’s adults, the most frequent of infrequent hostile responses occurs when an animal accidentally leaps onto another’s back, or onto an already occupied limb that can’t support two. Howlers seem properly sensitive about their balance and stability of support, since the ground may be 100 or 150 feet down; an upset howler will often snap at the intruder. A mother may also snap at or cuff a too-attentive “auntie”; I’ve even seen an adult male do the cuffing. In contrast to this unusually “fatherly” behavior, I’ve also seen an adult male squeeze a young juvenile until it screamed, with no apparent provocation. Collias and Southwick (1952) once saw a male bite an infant’s tail in half and hurl the baby to the ground.

These latter incidents are unusual behaviors. Some students of primate psychology object to the citation of unusual behavior, on the grounds that it misrepresents the usual behavior of the species. Homicide and suicide are comparative rarities in human society, too, but they’re considered significant in framing personality theories, particularly as they bear on the question of innate aggressiveness in humans. Just so, examples of howler aggression within the group are mentioned to indicate that these animals aren’t entirely peaceful, but have the potential for aggression under certain conditions.

The isolated males mentioned earlier may represent an even stronger case of within-group aggression. No one has ever observed a howler being driven from a group by his fellows, but there he is, out in the jungle, all by his lone. This occurrence of isolated males is doubly surprising, because howlers don’t show the sharp competitiveness and rigid dominance structure displayed by certain other monkeys. No one has witnessed any fights to the death, or even any spats leading to serious injury among individual adults; no struggles for mates, no squabbling for food. Such behavior seldom occurs in primates having clear dominance hierarchies, either; but the howlers attain this relatively peaceful state without even an apparent rigidifying of social structure. Chivers (1969) suggests that howlers may be getting more combative toward their fellow clansmen as the island population increases; he reports a 15 per cent incidence of “scarred, or freshly torn, lips” among adult males as evidence. But he doesn’t report having seen any battles in which these split lips presumably originated.

Altercations between howler clans, however, are not hard to find. Although a physical clash between two clans, with hand-to-hand combat, has never been reliably reported, howlers do react violently when another group approaches anywhere within several hundred feet. The reaction and counterreaction consist mainly of what howlers do best: they howl.

Robert Ardrey (1961) makes the encounter fit for a Disney film:

When two groups sight each other, each on the fringe of its territory, all break into total rage. Males, females, juveniles and infants become ants on a hot plate, leaping through the branches, scudding through the tree-tops, screeching, barking, chattering in frenzy. The forest cathedral becomes a green asylum for its insane inhabitants, and the howls of apparent melancholia become the shrieks of the truly demented.

This is a pretty example of Ardrey’s literary talent, but it doesn’t describe usual howler reactions. When one group’s nearby presence does become known to another (by smell, by incidental noises, or by the first group’s howls at another provocation), the second group is likely to stop any activity, rather than leap about furiously. Then one or more adult males will begin the bark leading to the roar and howl of the full-throated vocalization, joined in by the other males. Adult females and occasionally older juveniles may add their barks and high-pitched chatter. The males of the second group will usually respond, and the cries of both groups may alternate or continue simultaneously from several minutes to an hour or more. During this time, monkeys in each group may move about cautiously, sometimes closer to and sometimes farther away from the other group; but their vocalization is likely to be their only clear sign of disturbance. If two clans accidentally find themselves unusually close together, some monkeys may shake branches or move about in apparent agitation; but they usually manage to stay so far apart that howling suffices to express their distress. Ultimately, both groups will subside into grumbling and feeding, or one group will turn in its path of progression and make a slow retreat. The remaining group may vocalize for several minutes more without reply from the retreating group; then it will either move into the trees just vacated, remain where it is, or itself retreat in the direction from which it came. In no case will one group pursue another.

These vocal exchanges between groups leave no casualties, but they do help to keep strange howlers out of the territory inhabited by each group. Much has been made of the territoriality concept recently, because it seems to offer a biological basis for certain cultural phenomena among humans; but considerable confusion exists about the howler’s version of territoriality. Howlers don’t act as though they’ve built a fence or drawn a border, and then go tearing around the perimeter looking for invaders. Rather, they’re born into a group that is familiar with a certain part of the jungle; in following the group, they presumably learn the location of food trees, lodge trees, and arboreal pathways in that piece of jungle, and turn back from unfamiliar territory. These pieces of familiar territory overlap somewhat, so that two or more howler groups occasionally come into contact. When they do, their reactions give no observable sign of territorial possessiveness but rather a display of irritability like that shown by a human when someone invades his “personal space.” As Carpenter (1965) has said of howlers, “they defend the place where they are” — an idea quite distinct from defending a big hunk of real estate.

Much has also been made of the observation that howlers “defend” their territory only against other howlers, and don’t react to intrusions by other species, such as white-faced cebus or capuchin monkeys. This isn’t quite true. Howlers do usually save their best howls for other howlers; but in several instances I’ve seen howlers move away from capuchins in the same tree, stop a group progression from one tree to another when capuchins came near, and sometimes even vocalize at capuchins. This isn’t really surprising. Even if capuchins don’t compete much with howlers for food, they are distinctly unpleasant animals in their natural habitat. They chatter altogether too noisily; they shake tree limbs, and sometimes break off and drop branches with an annoying crash; they do not respect howlers’ branch-rights, and will sit down on the same limb even if it’s in danger of snapping. If you lived in a tree, would you want a band of white-faces to move in next door?

Ivan Sanderson (1957) has suggested of the howler that “If the countries of South America ever decided to federate or wished to choose a common emblem, this might well be the best animal to select as typifying the continent, roaring his individualistic defiance at the world, emblazoned in red-gold on a green jungle treetop against an azure tropical sky.” Sanderson is either unaware of, or too embarrassed to mention, the howlers’ alternate means of defense, though it might make the animals even more appropriate for certain national emblems. Very simply, when a threatening figure such as a human being is on the ground, the howler will move directly over him and then release a load of urine and watery dung. The first time an observer is the object of this treatment, he feels insulted and betrayed — he feels the howlers’ meanness has no limit. After a while he becomes accustomed enough to the practice to move out of the way, and begins to wonder how this easily avoidable maneuver can be of any value to the howler. He wonders until the day he discovers a howler’s hidden presence overhead by the gentle rain of excrement all around him.

Several writers have argued that this foul behavior is not intentional, that the howlers let go only because thev’re nervous or excited. Anyone who believes this should be placed unprotected beneath a hostile howler group several times before he makes his final judgment. Both male and female adults have been observed repeatedly to move with care from distant to nearby limbs and discharge their loads of excrement in close proximity to the observer, before returning to their original positions. Carpenter once built a blind under a group’s likely line of march, to observe them unseen. The howlers paused over the blind, each in turn, defecated, then continued on their way.

more monkeys, also apes: primate diversity

Early in 1933 a small band of monkeys was imported from Singapore and liberated in this natural jungle more or less as an experiment. Not only did this little band of immigrants from far off Asia decide the jungle was satisfactory but they set up their own government, ruled by a chief and took over the jungle so completely that it was soon found necessary to construct a cage to protect visitors from the jealous monkey inhabitants.

                                        — Leaflet from Monkey Jungle, 22 miles south of Miami, Florida

“Carpenter has a good thing going with his research on howlers,” our friendly neighborhood Congressman might tell the National Science Foundation, “and the Canal Zone is American soil. But don’t go handing out grants to study foreign species, understand? When you’ve seen one monkey, you’ve seen ’em all. If anybody feels like gallivanting around to look at something different, let him go to the zoo.”

Most behavioral scientists seemed to feel likewise for two decades after Carpenter’s pioneering work with howlers. (Their attitude was helped along by the interference of a world war and the practical difficulty of mounting field expeditions, as compared with doing laboratory research.) No major effort was made to study other primate species in the wild, with the exception of Carpenter’s own observations of gibbons; and the howler data, along with Sir Solly Zuckerman’s baboon studies in the London Zoo, strongly influenced numerous scientific and popular treatises on What Primates Are Really Like. The territorial howler, the tyrant baboon with his harem — why, we can see the same behavior in Florida’s Monkey Jungle! In human society! Primates all, and all brothers beneath their multicolored skins!

The gibbons and Sir Sollv’s baboons will shortly be given their due. First, let’s tell our Congressman a little about zoo studies. The zoo primate is in a strange environment, tremendously simplified in comparison with his own wild homeland, even when zookeepers do their best to supply him with “natural” surroundings. His physical wants are met easily; his area of travel is grossly limited; and he is likely to have been raised not by his own parents but by the zookeeper’s daughter. So he will probably be crazy, to begin with; even if he isn’t, his behavior will be abnormally restricted to little more than playing, copulating, or fighting. A zoo may be quite instructive about the physical form of various species. But if you’re interested in primate behavior, a visit to the zoo is rather like going to a lunatic asylum to laugh at the inmates.

Nor can the howler be accepted as representative of wild, “natural” primate behavior. Primates include about two hundred species, hardly brothers under the skin and mostly distant cousins at best. Not only the physical form but the behavior of each species is distinctive. This significant fact has been made clear by a major resurgence of primate field studies during the past few years. In reviewing the recent findings (made by psychologists, zoologists, anthropologists, and representatives of several other disciplines), we can use howler behavior as a benchmark, in order to note important similarities and differences; but the howler is no more an ideal model of a social monkey than any other species. So many different kinds of monkeys and apes have by now been studied in detail, or are being studied presently, that we’ll look at only a suggestive range of primate behaviors.


Let’s start again with social cohesiveness. Most primates do live in stable groups, and in many species the size of the group is not much different from what you’d typically find in howlers. But an Asian ape, the gibbon, and two New World monkeys, the callicebus and the night monkey, regularly live in nuclear family units only — mother, father, and one or two young. At the other end of the scale, the hamadryas baboons of Ethiopia may gather in sleeping troops of several hundred animals at night, though the troop breaks into smaller units during daylight hours. The proportion of adults to juveniles and infants in a group varies in different species according to birth rates, maturational rates, and survival rates; but there is often a disproportionate ratio of adult females to males, perhaps 2 or 3 to 1. This recurrent oversupply of females in many primate groups has intrigued primatologists ever since Carpenter first gathered reliable data on it in howlers (he called it the “socionomic sex ratio”). K. R. L. Hall and Irven DeVore (1965) explain it rather simply. It’s not the result of male duels to the death, as some writers have speculated; males don’t duel to the death under natural conditions, among howlers or baboons or any other primates studied except man. Nor can it be accounted for as the result of young males being driven out of the group by the old tyrant, or of old males being driven out by the upstarts, though an occasional lone male is seen. Instead, the disproportion is mainly a matter of maturational rates. Females mature from half again to twice as quickly as males among most primates; and since primates have relatively short life spans, adult females may outnumber adult males considerably, even if the total number of males — adult, juvenile, and infant —in a group is about equal to the total number of females. Add to this that males seem more likely to become isolates, that males are more likely to be killed defending the group when predators abound, and that some younger mature males are misidentified as juveniles because their sexual maturity isn’t as obvious as in a pregnant, estrous, or child-carrying female, and there you have the socionomic sex ratio. (Hans Kummer [1968a] reports only about 10 per cent more adult females than males among hamadryas baboons. But in that species, younger animals of both sexes take on a distinctive sex role rather early, so he counts every animal from three and one-half years on. Because he starts his count of sexually typed animals at the same age for both sexes, the numbers come out almost equal.)

Among many primate species, groups are virtually closed units, as with howlers. Additions are made mainly by births; strangers are not welcome; departures are likely to be permanent. Among gorillas and even more so among chimpanzees, groups are rather fluid. George Schaller (1963) observed several instances among gorillas in which adults of both sexes, as well as an occasional juvenile, switched groups with little or no trouble. Vernon and Frances Reynolds (1965) describe chimpanzee groups in the Budongo Forest of Uganda as “constantly changing membership, splitting apart, meeting others and joining them, congregating or dispersing.” Some of these behaviors seem random; others involve group transfers as an animal’s life condition changes. A female in heat becomes an attraction for wandering males, thus forming what Adriaan Kortlandt (1962) calls a “sexual group.” When she gives birth and is no longer attractive, she associates with others in her condition; a “nursery group” results. When no estrous female is present, males find little in mothers or children to attract them, and so they may travel in all-male bands. Larger groups, containing males, females in heat, mothers and young, typically are attracted together by a particularly bounteous food tree. Jane Goodall (1965) reports that mother-and-young families are the only completely stable “groups” she has seen; any other relationship may break off without notice, and an animal in any other category may travel by himself if he has no strong reason to travel with other chimps.


Within primate groups, substantial differences in social structure can be observed. Few species seem to have a much looser structure than the rather free-form howler clan; but many primates show a tighter group organization, founded upon a dominance hierarchy. Dominance hierarchies have been observed in various birds and mammals, starting with barnyard chickens (from which came the alternate term of “pecking order”). The most dominant animal in a group may get first crack at food or sexual mates, or may determine group movements. He may also be able to display hostility toward other group members without being aggressed against in turn. The less dominant animals may make a show of subordinance, or redirect their annoyance toward group members still lower in the hierarchy. When Walt Whitman said he could turn and live with the animals because “Not one kneels to another,” he knew not whereof he spoke.

Dominance in primates was first studied extensively by Sir Solly Zuckerman (1932), under unfortunate circumstances. Sir Solly, a distinguished British zoologist, managed to observe baboons in the wild for a few days in the 1920’s, but he did most of his research on a huge hamadryas baboon colony in the London Zoo. To organize a zoo colony properly, he would have had to know much more about baboon social life in natural surroundings than he was able to learn in a few days. One bit of helpful knowledge would be the slightly greater number of adult females than adult males in the wild hamadryas troop. Another would be the fact that peaceful relationships in a hamadryas troop are structured on adult males’ recognition of each other’s exclusive rights to certain females, these rights being built up gradually from before the time a female is sexually mature. And another would be the baboon troop’s insularity: baboons usually react negatively to strange baboons.

What Zuckerman began with at the London Zoo was about lour times as many adult males as females; males captured from several different troops, and thus unfamiliar with each other; males thrown together with no time for the peaceful establishment of rights to particular females. What Zuckerman got were vicious fights between males, fights over females and even against females and young — baboons mutilating and killing each other until hardly any baboon colony remained. In the process, Zuckerman also saw males repeatedly presenting “sexually” to other males, and this looked to Zuckerman like sexual perversity, though it was most likely a stereotyped communication pattern referring to geographical movement rather than sex. In his famous book, The Social Life of Monkeys and Apes, Zuckerman presented a picture of the dominant male baboon as absolute tyrant, monopolizing all available sex partners, destroying opponents, always craving sex, sex both normal and perverted. Until recently, his work has been the definitive picture of baboon social life, and for many the definitive picture of natural primate social life, from which man escaped by supreme acts of will and morality.

Following Zuckerman’s lead, baboons were long thought to be particularly stern practitioners of the rule of dominance. The top male had his harem, which he defended against all comers; subordinate males virtually starved, if they weren’t killed or driven into the lonely wilds. But that myth has been at least partly shattered: neither hamadryas nor savanna baboons are anywhere near as hierarchy-oriented in their natural state as their zoo behavior had suggested. Indeed, even when a distinguishable dominance hierarchy has been found to exist, researchers have often had to use sophisticated statistical techniques to be sure which animals are dominant over whom, and under what circumstances (Bernstein, 1970).

The savanna baboons of South and East Africa display clearer dominance patterns than most primates, but even their behavior is remote from the heights of despotism. Normally, competition for food is nonexistent; males’ infrequent disputes for other reasons are brief and usually noninjurious. Hall and DeVorc (1965) report that the most dominant males in a group do tend to monopolize sexually receptive females during the peak of each female’s estrus period. But Thclma Rowell (personal communication) detected no such monopoly in the groups she observed; even in Hall and DeVore’s groups, dominant males shared estrus females with subordinate adult males and older juveniles when the females were coming into and going out of heat.

Nor does dominance in the savanna baboon always follow the traditionally assumed single-line pattern, from the top male to the second and so on down, in such other prerogatives as grooming or aggressive displays. Rowell (1969) found adult males of forest-dwelling groups to move from one group to another so often that no stable hierarchies could be established. Among groups where hierarchies exist, the dominant position is often held jointly by two or three adult males, acting in concert, though singly each of these males may be subordinate to other animals in the group. Members of such an alliance come to each other’s defense when attacked by another male; they don’t compete even for scarce food tidbits provided bv human observers. They obviously can’t share the same female simultaneously, but their alliances appear to withstand occasional sexual competition. Such dominance alliances probably develop over extended time periods in most groups, but Hall and DeVore observed one situation in which a very subordinate male of one group moved to another group, quickly allied himself with one of the more dominant males there, and thus began a presumably happy life as joint top dogface.

If a savanna baboon does attain dominant status, he acquires certain responsibilities along with his ambiguous privileges. His principal obligation is to protect women and children. When the group is on the move, dominant males typically remain at the center, surrounded by mothers and infants, with subordinate males on the periphery. At a sign of danger, the dominant males move immediately to the front of the group to face the threat — not only with their bark, but with their daggerlike canine teeth. As with howler males, both dominant and subordinate savanna baboons also protect the group from internal aggression: if an older juvenile is roughing up a younger one, or if two females have become angry at each other, an adult male will usually break up the dispute before it escalates, either by growling or by chasing the offending party away (Hall and DeVore, 1965).

Group structure and dominance relationships are arranged rather differently among hamadryas baboons, the species involved in Zuckerman’s zoo studies. In the wild, hamadryas groups exist on three levels (Kummcr, 1968a). The troop contains as many as several hundred animals, who usually sleep in the same area and leave together in the morning. They then split into several bands of thirty to ninety animals, who remain loosely associated during daytime activities. Presumably both the troop and the band serve useful defensive purposes; the troop may also develop partly because a shortage of suitable sleeping space forces large numbers of animals to sleep in the same area (Kummer, 1968b). Each band is further organized into several one-male units, typically composed of one adult male and one to four females with young. The band’s subadult and young males who have no females of their own may follow a one-male unit at a distance, or they may simply remain in the general area foraged by the entire band. Around 20 per cent of a band’s adult males at any one time are mateless.

Zuckerman identified the one-male units as “harems,” and the name still seems somewhat appropriate, though some harems contain only one female. The male leader keeps his eye on his entourage and permits little association, sexual or otherwise, with other males. A venturesome female is likely to be rewarded with a bite on the neck from her leader. But the harem is not built up through brutal combat, as Zuckerman thought; nor is it quite as closed as it appeared to him in the zoo. The one-male group is likely to begin with a young male’s adoption of a one- or two-year-old female, still sexually immature and still several years from childbearing age. He mothers her, he keeps her close to him and brings her back if she tries to run away, until a firm social bond is established. Only later will sexual activity begin. The male may then acquire other mates in the same way, or take over a mature female who has been “let go” by an aging male. Hans Kummer saw little indication that harems are broken up through fighting, but the older males seem to tolerate their females’ moving gradually farther from them, until the females may affiliate with younger males. The old males, as they give up certain sexual prerogatives, gradually take over direction of the entire troop’s travel — expending their energies as executives rather than as breeders.


Territoriality, in the sense of staying in a home range that overlaps little with other groups’ home ranges, and of somehow defending one’s immediate location from infringement by other groups, is not nearly as widespread among primates as Carpenter thought when he found it in the first two species he studied at length, the howler and — halfway round the world — the gibbon. Perhaps a more frequent pattern is the one observed by Phyllis Jay (1965) among langurs, the sacred Hanuman monkeys of India. In the population Jay studied, each group of langurs stayed in a home range that overlapped with the home ranges of other groups, and largely within a core area that didn’t overlap with others. Groups seldom met, partly because adult males produced a “deep, resonant ‘whoop'” when ready to move to another area (somewhat as howlers roar before they move in the morning). But when groups did meet, conflict didn’t occur; the smaller group gave way to the larger. (Under different environmental conditions, Suzanne Ripley [1967] observed troops of langurs apparently seeking each other out for battle, but territorial issues did not seem to be involved.)

In savanna baboons too, a group’s home range overlaps considerably with areas sometimes traversed by other groups, while core areas are fairly exclusive. When two groups do meet, for instance at a water hole during the dry season, they may act nervous toward their newfound neighbors, but they usually don’t roar, bare their teeth, or otherwise display hostile intent in the howler manner. One baboon group may gradually creep away, and some entire groups appear dominant over other entire groups in this way, at least partly on the basis of group size; but the whole interaction is polite, with no chasing allowed (Washburn and DeVore, 1961). Waterhole sharing under such circumstances may well be necessary for species survival. Howler hostility toward other groups is perhaps not a requirement of primate life but a luxury.

A weird modification of territoriality is practiced by the callicebus or titi, studied in Colombia by William Mason (1966, 1968). Callicebus monkeys live in small family groups; the parents display their mutual attraction by sitting together and entwining their long tails at eventide. Each family group usually stays in its home range, with several such families occupying delimited areas of a small woods surrounded by open savanna. Groups coming into contact vocalize strenuously, and may charge each other with a great display of hostility. So far, they appear howlerishly territorial; indeed, titis show even less overlap of group territories. But during a female’s estrus, something extraordinary happens. When two groups meet for an aggressive interlude at the edges of their respective territories, this usually monogamous and homebodv female may ignore her own mate’s sexual advances and wait for the rival male to come satisfy her. or she may even move out toward him. Her own mate naturally displays signs of annoyance, and may attempt to interfere, but often she’s able to enjoy a brief copulation with the other family’s male, after which she returns to her husband with virtually no marital repercussions. Perhaps her mate is by this time already contriving his own liaison with another titi’s woman. The promiscuous but group-centered howlers would surely deliver a moralistic roar at the very thought of such behavior, were they able to think upon morality.

Among apes, the gibbon seems unique in its strong territorial defense. Orangutans have been too little studied, and their societies perhaps too disrupted by all the forces that have pushed them toward extinction, to say much about their apparent lack of territorial behavior. Gorilla concern about territory is rather casual (Schaller, 1963). The home ranges of adjacent gorilla groups usually overlap a great deal, sometimes completely; encounters between groups elicit only slight excitement and perhaps a staring contest between males, with a little chest-beating sometimes thrown in. Schaller has seen two groups bed down and sleep for a night side by side, then go their separate ways in the morning. Intergroup antagonism among chimpanzees, including territorial defense, simply has not been observed. Each loose “community” of frequently interacting chimps apparently remains mostly in one several-square-mile home range, but individual animals exhibit no proprietary feelings toward this familiar area. The Reynoldses several times watched groups, which spent nearly all their time in one area, spontaneously travel “fast and noisily” into the heart of another area usually occupied by a different set of chimps. No hostile reaction ever occurred on either side. When large numbers of chimpanzees gather in a small area around a common food source, the animals appear extraordinarily excited at the sight of so many unfamiliar faces, and sometimes stage what have been called “carnivals.” Many animals vocalize simultaneously, and males drum on tree trunks for hours at a time. The Reynoldses (1965) describe “chimpanzees coming and going in all directions, some to and some from the centers of hubbub. . . . Sometimes whole valleys along a stretch as much as a mile would resound and vibrate with the noise.” The carnivals occasionally last into the night (as all good carnivals should), and sexual activity may reach unusual heights. It’s indicative of chimpanzee distinctiveness that although howlers roar at unfamiliar howlers in attempted repulsion, chimps treat similar occasions with excitement, curiosity, and what really appears to be delight. Further, though chimps do call out briefly and drum on trees while traveling, this doesn’t seem to space out or prevent group contact, as with other primates; instead it may attract them together.


The howler gives about as little attention to its young as any true primate can get away with; in several other species the attention is substantially greater. In the langur, “auntie” behavior has been highly developed. Females of all ages are immediately attracted to a newborn, and begin to take it from its mother a few hours after birth — in Phyllis Jay’s words (1965), “gently manipulating, nudging, and smelling the infant.” When the infant indicates discomfort, another female (or perhaps the mother) will take it. Jay has seen as many as eight females handle the infant on its first day of life. Older females seem more competent than younger at this baby-handling, but as Jay points out, “Because langur mothers allow other females to hold their infants, no langur female is completely without experience in infant care.” Furthermore, “auntie” behavior often turns into “babysitter” behavior: one mother may deposit her infant with another mother, and at times a single mother may sit with several infants for a good part of the day.

Among a few primate species, paternal behavior is notable. The night monkey father, for instance, apparently spends most of his waking time carrying his infant. By the age of three weeks, the night monkey infant goes to his mother mostly to be suckled, and spends the rest of the night riding on his father’s back. The callicebus father behaves rather similarly. Martin Moynihan (1964) suggests that this may be a useful division of labor for such monogamous family-grouping primates. But it would cause problems for species such as the howler, where no child knows its own father and where infants may outnumber adult males.

In at least some groups of Japanese macaques, adult males take care of the group’s juveniles. They don’t merely protect the juveniles from attack, as with savanna baboons; they don’t adopt female young in mate relationships, as with hamadryas; and they don’t carry young infants around, as with night monkeys. Instead, a male will take over a year-old juvenile of either sex when the mother has had or is about to have a new baby. According to Junichiro Itani (1959), the male hugs the yearling, “takes it on his loins, or walks with it; when sitting, he will groom it.” This paternal care declines after the female delivery season has ended, but it may be repeated again the following year, particularly for juvenile females or those who are “especially undergrown and weak.” Most juvenile males by this time have formed their own play groups.


The adult males’ relationships with juveniles may deliver benefits in both directions. Associates of the Japan Monkey Center have encouraged groups of monkeys to come out of their usual mountainous forest habitat to more humanly accessible clearings by “provisioning”: they supply the monkeys with sweet potatoes and other interesting foods. Even before provisioning began, considerable differences existed between food preferences of various groups. For instance, some groups dug for edible roots; others did not. Provisioning made possible experimentation with the development of such apparently “cultural” differences in behavior — patterns learned and handed on rather than inherited (Kawamura, 1959). Itani passed out candy to one group; the older animals refused to eat it, but the juveniles were more venturesome, and candy eating soon spread throughout the group’s juvenile population. Mothers still in close contact with juveniles began eating candy too; so did the more paternal males. Infants then learned from their mothers. The less contact an animal had with the innovative juveniles, the slower he was to pick up the candy habit. At the end of a year and a half, 51 per cent of group members were confirmed candy eaters.

In another group (Kawamura, 1959), one juvenile female spontaneously started washing sweet potatoes before eating them. Again, other juveniles and then their mothers picked up the practice. But in this troop, adult males had no close “paternal” relationships with juveniles, so the males went on eating dirty sweet potatoes. In still another instance, M. Yamada began giving wheat to the group; this time an adult male was first to accept the new food. Shortly the group’s dominant males picked up the habit, then the head female; and from there it spread downward to the whole clan. Some Japanese monkeys even invented a way to separate the wheat from the chaff, so to speak (Frisch, 1968). The wheat was usually poured out onto a sandy beach, and the monkeys had a hard time making a meal of wheat without getting a mouthful of sand too. So they began tossing the mixture into the sea. The sand sank, and they scooped the wheat off the water.

The founding of another cultural practice may have occurred in chimpanzees. Goodall (1965) observed chimps making and using tools, by trimming twigs that they then thrust into termite holes, so they could draw out and lick off the clinging termites. Only the chimpanzees in Tanzania’s Gombe Stream Reserve have been seen to use this method of catching termites. Young chimps there appear to learn the practice from their elders, so it may well be that one lucky or brilliant chimp developed the practice, which then spread to other members of the chimpanzee community by imitation. In other areas, chimps have been reported to thrust sticks into beehives and then lick the honey off; whether the two practices are culturally related isn’t known. The Reynoldses, in an area where both beehives and termite nests are present, never saw chimps bother either one.

Chimpanzees in particular show a variety of other social behaviors: complex greeting patterns, social facilitation (hyperactivity of various kinds in the presence of several other chimps), sexual and aggressive displays. Indeed, the chimpanzee may prove to be, despite its extremely loose social organization, the most “advanced” social animal besides man. The Reynoldses (1965) have suggested that the chimps’ apparently loose social structure really reflects their relative advancement in social behavior: because they can communicate so efficiently from a distance, and can remember intermittently occurring social relationships (as indicated by variations in greetings to more and less familiar animals), they needn’t maintain tightly organized groups with constant visual contact. Just so, a loving and confident husband might be observed to remain farther from his wife at a party than an uncertain and jealous husband from his. But perhaps we need more research on both husbands and chimpanzees before we accept either hypothesis.

As with howlers, the social behaviors described in this section consume only a small portion of primates’ time and energy — even among chimpanzees. Social animals they may be, but the nonhuman primates are still far less wrapped in a social world than humans are. However, the behaviors I’ve mentioned aren’t just trivial curiosities, a sort of Believe-It-or-Not of the primate world. They are all functional aspects of living behavior patterns that have enabled many different primate species to maintain themselves for several hundred thousand years or more. Socialness is important for the primate, even though his small brain is not constantly concentrated upon it.

how the primate got his gregarity: origins of social motivation

The main factor that determines social grouping in subhuman primates is sexual attraction. Females attract males and males attract females.

                     — Sir Solly Zuckcrman The Social Life of Monkeys and Apes

My daughter Heather, long before she could talk, showed a fierce fascination with other children her size — shouting to them across rooms crowded with adults, grabbing at them as they passed in supermarket shopping carts. Her reaction was so enthusiastic, so free of attempt to win parental favor, that it appeared almost instinctive. Then again, it might have come from her equally vigorous appreciation of herself in a mirror, projected outward onto similar objects (but why did she like her reflection, other than because it seemed to her another child?); or from the pleasure she got out of child-sized tovs; or even from her mother pointing out other infants with the cry, “Bay-bce! Bay-bce!” Who knows? No one can really know, since human babies are complicated creatures reared in a complex world, and since child development researchers are ethically constrained from reducing either complication or complexity experimentally. So the nature-versus-nurture controversy, at least as applied to humans, is hardly a fit topic any more even for junior high school debating societies. Separating the factors in any meaningful way is simply impossible. But wait — who should be available but the relatively simple nonhuman primates, in their relatively simple environments, and with little protection of their rights against drastic experimentation? They may not be of much direct help in explaining my daughter, but perhaps field and laboratory studies of their behavior can help us deal with broader questions raised by her enthusiasm for other children: Why are so many primates attracted to each other? Why do they group together? What are the bases for primate sociability?


Zuckerman’s early hypothesis (1932) is still influential among nonspecialists: society is founded upon sexual copulation and sexual competition. “Reproductive physiology is the fundamental mechanism of society,” he declared — particularly of primate society, since he assumed that primates lack a breeding season and are thus continually interested in sex. Sir Solly devised this position not from reading Freud but from observing his zoo monkeys, who had nothing much besides sex to be social about.

Carpenter had made more reliable observations, from which he developed a more subtle explanation of sex’s role in primate society (1942). His idea was not that sexual desire dominates primate life, but that the social bonds between males and females in a group are repeatedly reinforced through the year by the reward of frequent sexual activity. This would be particularly effective in holding groups together if they practiced the kinds of “rotating mateships” found in howler and rhesus monkeys, because in such situations a female rhesus, for instance, “may be possessed by all the more dominant males in her group during estrus, and with each one social affinitive relations may be strengthened by the coincident positive conditioning (or learning) occurring during copulations.” In other words, all the adult females in the group keep all the adult males happy (or at least the more important males), and all the males keep all the females happy, so they all live together in one big happy family.

The view of sex as the center of primate society remained attractive enough for a distinguished anthropologist, Marshall Sahlins, to write in 1960:

The powerful social magnet of sex was the major impetus to subhuman primate sociability. . . . Subhuman primates are prepared to mate at all seasons, and although females show heightened receptivity midway through the menstrual cycle, they are often capable of sexual activity at other times. Most significantlv for the assessment of its historic role, year-round sex in higher primates is associated with year-round heterosexual social life.

Sahlins qualified his remarks with the acknowledgment that “certain Old World monkeys … do have seasonal declines in breeding without cessation of horde life,” but he insisted that in general, sexuality pervades nonhuman social life throughout the year.

Carpenter many years earlier had noted in howler females a distinct estrus period, not just a period of “heightened receptivity.” He was therefore much more careful to qualify his hypothesis about sex and sociability (Carpenter, 1942):

It must be concluded that strong attachments between individual monkeys or apes may be formed and persist without overt and primary sexual activity, unless we make identical the social and the sexual behavior. Through reciprocal play, through mutual grooming or through communal feeding, social relations and statuses are acquired and these learning processes are strongly motivated and reinforced by reciprocal interactions of drives and their incentives oilier than sexual.

As masses of data have recently become available showing either the distinct periodicity of a primate species’ sexual behavior, or the year-round low level of sexual activity in many primate groups, the assumption of sex’s primary role in sociability has largely been abandoned; and these other rewarding aspects of social life have been given more and more weight. It’s clearly pleasant for the infant primate to have the attentions — the nutriment, warmth, protection — of his mother; pleasant for the young primate to play with his fellows; physically pleasant for most primates of any age to be groomed (as with human back-scratching). It may also be rewarding to see other members of the group defend oneself, perhaps oneself to join in defense, against intruders or predators. When sexual behavior occasionally happens, that should be rewarding as well; but it is not the ultimate and unique gratification of social life. As Carpenter says elsewhere (1940), “Almost every phase of behavior of which a primate is capable enters to some degree into the determination of its ‘gregariousness’ and the qualities of its complex social behavior.”


Whether social behaviors are pleasurable to the individual or not, they may promote survival of the species. Recent writers on primate sociability have chosen to emphasize this feature: the functional basis of group living. If a social behavior is more functional in preserving species existence than a nonsocial one, evolution is more likely to “select” the social; and there are good reasons to believe that social life is and has been highly functional for most primates. This doesn’t necessarily mean that evolution has implanted a social instinct into every monkey and ape, though we’ll explore that possibility later. It may only mean that the primate’s physical structure predisposes him to social living, makes socialness either an easier or a happier way of life than isolation — for instance, because an animal has a back that feels good when it’s scratched, or is so physically disorganized at birth that he must cling to his mother for months, being strongly rewarded by social contact all the while. Or it may mean developing the capacity to transmit traditional patterns of functional behavior: through social communication, through imitation, or through parental reward or punishment of certain acts.

The usefulness of primate social grouping is many-faceted. Childraising may be more effective, because the infant may not only get attention from several adult group members at various times, but also has the opportunity to learn coordination and sex-roles safely in play with other infants. Further, in case of the mother’s death, adoption by another group female may be possible. Growing adolescents have the opportunity to observe various adult roles that may help them mature behaviorally with fewer mistakes. Adults have more opportunity to find sex partners when a partner is desired, and indeed more opportunity for any kind of social stimulation when and if they need it. The group may not only protect the individual from immediate danger, but may help him to cope with potential dangers by teaching him, at least through example, about his environment: what foods to eat, what organisms and natural phenomena (eagles, nettles, bogs) to avoid. Trial-and-error learning may be useful for the laboratory rat, but it could be deadly for the isolated primate.

Because different primate species live in different environments, the same forms of behavior are not universally functional. Development of varying functional responses to varied environments may account in large part for behavior differences between species, as well as for variations of group behaviors within species. The Reynoldses (1965) have suggested that chimpanzees’ relatively large size and their dependence on fruit are at least partially responsible for their loose social organization. Bands of twenty to thirty chimps may form in areas where ripe fruit is heavily concentrated, but at other times and places such a large band could hardly avoid starvation. Mainly ground-dwelling monkeys show, on the whole, distinctive general behavior patterns not found in mainly tree-living monkeys — for instance, the clearer dominance hierarchies of the savanna baboon, which some writers have suggested would be the most useful social organization for defense against ground-dwelling predators. The same baboon species may encounter fewer predators in the forest, and so shows the much looser social structure observed by Rowell (1969). A related species, the hamadryas, may encounter such severe survival problems in the harsh Ethiopian environment that its social structure in some ways is even stricter than that of savanna-dwelling baboons, though more fragmented as well. (On the other hand, as Washburn and Hamburg [1965] note, different adaptations are possible even within similar environments. For instance, patas monkeys, “unlike baboons, have adapted to ground life by speed rather than by size and social organization.”)

The origin of primates’ ancestors in trees rather than on the ground was itself probably a significant factor in the development of strong social motivation, simply because the infant had to cling so closely to its mother for so long a time before it could safely venture out into the tricky arboreal environment. At the same time, Hall and DeVore (1965) have suggested, development of the abilities necessary to respond to a complex social life may in turn be the source of much primate adaptability to the physical environment: if a primate can change his behavior to suit his neighbors, he should be able to change even more easily to suit a new but relatively stable climate or geographic area.


With all this talk of functionality and evolution, the word “instinct” may keep bobbing around in the back of your head. Has evolution been so impressed by the functionality of group living that it has implanted a seed of gregariousness in us all, monkeys, apes, and men? How do we answer that question? Well, we can’t go depriving human children of social contacts, or stripping their environment to the bare essentials, to see if gregarity remains. But we can do — and have done — such things to monkey children.

Much laboratory work in this area was initiated by Harry Harlow and his colleagues and students at the University of Wisconsin. Harlow is well known for raising infant rhesuses on imitation mothers, shaped from heavy wire screen that sometimes remained bare and sometimes was upholstered with terry cloth. The infants (now we can say, “Of course!”) vastly preferred the cloth-covered mother even if she was milkless while the bare-wire mother had a milk-dispensing “unibreast” in mid-thorax. Perhaps less well known is a later Harlow study in which the infants were given a choice of two cloth-covered mothers, one with unibreast and one without. Some infants had a brown nursing mother and some a green nursing mother; the non-nursing mother got whichever color was left. All good humanitarians and Martians should be glad to know that color of mother made no difference to the infants. More significantly for our purposes, preference for the nursing mother lasted less than four months. After that, though the young monkeys were still nursing, they spent roughly equal amounts of time with the nursing and non-nursing mothers. Harlow (1962) concludes, “Certainly nursing and activities associated with the breast are not variables underlying the persistent, relatively inextinguishable affectional bonds of the infant for the mother.” All good behavioral psychologists, who’ve attributed human social attraction mainly to maternal reduction of such “basic” drives as hunger and thirst, should be properly dismayed.

Harlow’s research has settled no arguments about the existence of monkey social instincts. The clinging reflex that predisposes the infant rhesus to “love” the cloth mother is not complex or long-lasting enough to account for social behavior by itself. But Harlow has made clear that a social instinct, if it exists, is not enough for social life. His infant rhesuses, raised for six months or more with an artificial mother, developed serious disturbances, particularly in their social behaviors: they withdrew into themselves when faced with more sociable monkeys, or fiercely attacked other monkeys who had undergone even greater social deprivation. Sexual behavior, the most obviously biological and “natural” of social behaviors, was just as severely blocked as other social responses. Harlow delights in telling how carefully his socially deprived females had to be coaxed and coached by experienced males before they ever conceived. Once they became parents, these females were often miserable mothers with their firstborn, completely ignoring the infants or grinding their faces into the wire mesh of the cage. Infants raised without mothers but with other infants seemed to develop somewhat more normally; and the miserable mothers, having contradicted all assumptions about maternal instincts with their firstborn, generally became capable mothers with their second. In each case, social “instincts” had to be supplemented by social contact with a real live responding primate before something approaching normal social behavior could emerge. As Harlow says about the sexual development of his socially deprived males, both “yearning and learning” are essential — both social motivation and social experience.

William Mason, a former colleague of Harlow’s, has approached the development of primate sociability from another direction. Mason (1965) hypothesizes the existence of an optimal stimulation level in any primate. When the degree of stimulation rises above the optimum, the primate will try to reduce it; when stimulation falls below the optimum, the primate will try to increase it. This basic idea, shared with several other psychologists, is distinctly different from the more popular drive-reduction theories, in which the organism is assumed to strive perpetually for a minimal level of stimulation.

Mason demonstrated how this hypothetical process of maintaining optimal stimulation operates, through a series of simple experiments. He dressed two experimenters in different costumes, each with a sort of abstract primate mask over his face. One costume was always associated with the experimenter’s quiet holding of a baby chimpanzee to his chest; the other costume was always associated with the experimenter’s playing with (tickling, bouncing, gently pushing) the baby chimp. Given a choice, the chimps generally preferred the play-person over the holding-person by a considerable margin. But in an unfamiliar room, which presumably raised the chimps’ overall arousal level, they usually didn’t want to be aroused further by play; they preferred the person in the holding-costume. (This preference decreased, however, as they became more familiar with the unfamiliar room.) The same was true in other disturbing situations, such as when a young chimp was removed abruptly from its cage-mates.

Mason uses the data from such experiments to explain at least partly the primate’s preference for social companionship. A world of inanimate objects could become very boring very quickly; other living organisms provide the most complex stimuli. Other organisms with which one is familiar — the members of one’s own group — could provide the most pleasing alternative to too much stimulation or too little. This is particularly true of the young primate, who can turn to his familiar mother in time of excessive stimulation, and to his playmates when more stimulation is needed. Even if adult life in the group becomes rather tame, the primate has received so many rewards — so many hours of optimal-level existence — from the group as a child, that he is unlikely to leave it without hesitation as an adult. Besides, there’s nowhere else to go. A solitary life would be even more boring, and an existence of flitting from one strange group to another too upsetting.

But what of primate species where adult group life is not so tame — where an aggressively dominant animal may at least occasionally threaten or chase or bite his subordinates? Wouldn’t his offensive behavior severely strain the group’s social ties? Michael Chance and Clifford Jolly (1970), using Mason’s hypothesis, argue that the reverse may be true. Initially, a subordinate animal may scurry away from his attacking leader, who has raised his stimulation level too high. But as time passes and the attack is not pursued, the subordinate animal’s stimulation level will drop to the preferred optimum, and he should therefore be attracted more strongly than before to the source of this recurrent pleasure, his exciting leader. In the words of many a harried spouse or parent, “Well, at least life is never dull with him [or her] around.”


Other research indicates that a primate needn’t play with others, or be groomed by them, or have sex with them, to find them rewarding; all he may need is a vision of another primate like himself. Robert A. Butler (1965) has conducted numerous studies in which a monkey housed in an opaque box is taught to perform tasks for the reward of getting a quick look at laboratory goings-on, or at a picture, or perhaps at a cagemate in an adjacent pen. Although the monkeys will work for a sight of just about anything (perhaps to raise their stimulation level to optimum), they appear to prefer looking at other monkeys, particularly at familiar monkeys, and more particularly at sexual mates — though males may prefer to look at a strange female rather than at their own mates, “especially if the stranger is in full sexual coloration”!

The reward value of social sneak-looks has been explained by some writers in terms of secondary reinforcement: the monkey has been satisfied in the past by sexual mates, or has been played with by his cage-mates, so he gains an associated pleasure merely from seeing them. Butler rejects this idea, because monkeys will persist at tasks for long periods with no primary rewards to strengthen the “secondary” visual one, and because very young monkeys — one to two weeks old — seem to find such visions rewarding too. One of the most convincing demonstrations of Butler’s point, and indeed one of the most convincing indicators of innate social responsiveness in primates, is a set of studies done by Gene Sackett.

Sackett (1966) raised infant rhesus monkeys from birth in individual opaque cages. Each monkey never saw another monkey and saw humans only during its first five to nine days of life, when it had to be hand-fed. One wall of each cage was a projection screen, and beginning at two weeks of age, the monkey was exposed to various projected color slides on this screen. Sometimes the experimenter himself turned the slides on and off; sometimes the monkey, after a brief initial exposure to the slide, could himself turn it on repeatedly for as many as five minutes, by touching a lever in the cage. The monkeys’ preference for particular kinds of pictures could be gauged either by how often they pressed the lever when they were able to control the length of the slide’s exposure, or by what behaviors they showed when the experimenter controlled the exposure of a particular slide. Month-old monkeys showed no differences in preference. But older infants showed a clear preference for pictures of other monkeys, particularly infant monkeys, rather than for control pictures such as sunsets, trees, or “a pretty adult female human.” Between ages two and four months, negative responses (“fear, withdrawal, rocking, and huddling”) were given especially often to pictures of a threatening monkey; but “this apparently innate fear response to threat stimuli” started declining midway during that period, presumably because the threats were never carried out. Pictures of monkey infants elicited more playing than any other stimuli, though the threat pictures were accompanied by almost as much play. The threat and infant pictures each generated more overall activity than any other types of monkey pictures (mother-infant, sex, play, etc.); these “other-monkey” pictures in turn generated more activity than the non-monkey control pictures.

Maybe seeing humans during the first week of life, or seeing the fur on its own body, could predispose the infant monkey to prefer other anthropoid forms on the screen (though why not the pretty female human?). But such experience would not account for its preferring to see an infant monkey, since it had never seen its own face — a distinctive cue for identifying infancy, in the pictures — nor for its showing fear responses at pictures of threatening monkeys. Furthermore, when Sackett allowed other infant rhesus monkeys to look at one of several live adult monkeys of various species (in a more complicated apparatus known as the Sackett Self-Selection Circus), the infants generally preferred monkeys of their own species, and particularly females. Since the infants had been removed from their own mothers shortly after birth, had seen no other monkeys since, and did not resemble adults of their own species either in color or in facial appearance, their preferences are hard to explain as the result of learning (Sackett, 1970). Sackett seems to have found, as he suggests, clear evidence for certain rather general types of innate social response tendencies, both positive and negative, in at least one primate species. Whether similar innate tendencies are present in other primates will be answered by further research. Whether such innate response tendencies exist in humans may never be definitely known.

Is primate social behavior determined primarily bv heredity or by environment? However dull or uncontroversial the answer sounds, it must be: by the interaction of both. As with various other organisms studied by ethologists, primate behaviors seem to have evolved in such a way as to interlock with the environment. Given certain environmental and social cues, certain heredity-based behaviors will emerge and will undergo subsequent modifications by the environment. Given a drastically different environment, such as a zoo or laboratory, these behaviors will emerge only in stunted, disorganized form, will emerge as perversions of “natural” behavior, or will be inhibited from development in any recognizable form. Primates differ from other organisms in that these emergent behavior patterns show more modifiability, more possibility of “cultural” variation as a function of variations in the social and non-social environment. This relatively greater modifiability itself suggests a major lesson about the basis of primate sociability: such sociability is not tied to one single factor or type of factor, but expresses a complex interaction of factors. Harlow’s research has destroyed the idea that positive reinforcement from breast or bottle is the basis for human sociability; DeVore, Washburn, and other field researchers have finally overturned the idea that sex is the single key to primate sociability. These single-factor hypotheses may be good for generating aphorisms, or for starting a field of research on its way from infancy, or for giving the lazy writer convenient cubbyholes in which to fit his accumulation of disorganized facts. But we must now accept the necessity for far greater intricacy in our explanations of social life, among monkeys and apes as well as among ourselves.

the human primate: generalizations to homo sapiens

How like to us is that filthy beast the ape.

— Cicero,  De Nature Deorum

What does all this information about nonhuman primates add up to? Some scientists would argue that it’s simply interesting in its own right, a description of yet another chunk of the universe; that science admits no scale of values which renders one topic more important than another; that we needn’t worry about any further implications of our findings, as long as we understand monkeys or apes better than before; and that therefore you might as well study monkeys as, say, humans. I wouldn’t dismiss such a position altogether, because I do find monkeys and apes interesting and entertaining in themselves — at least as interesting and entertaining as some people I’ve tried to study. But I’d still have to say that this is not enough: that at least as a social psychologist, and at most as a human being, I must search the nonhuman primate data for whatever information will help in understanding the world’s most social primate.

I do not, however, fall into that small but intense crowd who feels that primateness is all — that mankind is ruled by his primitive primate passions, and that with knowledge of his evolutionary kinfolks’ behavior we can plumb him to his depths. Nor am I, on the other hand, among those who, like the anthropologist Leslie White (1949), feel man to be so qualitatively superior to other primates that we can ignore them altogether. Though this may sound like mass-media mugwumpishness, I must conclude that neither extreme embodies the most tenable position. which of course I am now ready to offer, in four parts.


Undoubtedly, the nonhuman primates resemble us more in physical form and physiological function than do any other organisms. So if we’re to gain any useful knowledge about humans from animal studies, we might best turn to these primates. Experimental psychologists have long argued the validity of using rats as analogs for humans, because both organisms learn — both show persisting behavioral modifications as a result of experience. But beyond this similarity, without which any fairly complex organism in a very complex world couldn’t exist, there’s probably little important resemblance. Between the rat’s slow modification of behavior after repeated trials and rewards, and a human’s one-trial learning of a multitude of memories, there may be an unbridgeable gap. Kohler’s classic studies of “insightful” ape problem-solving (1927) suggest at least somewhat greater similarity in learning potential between us and other primates than between us and rodents. Greater similarities also seem to exist in such things as curiosity level (for most primate species) and amount of unstructured play — a worthwhile sort of behavior for incidental learning.

That we do need to turn to other species for certain information has already been emphasized. We don’t permit ourselves to do certain things to humans, even in the interests of science, that we can do to captive monkeys. We can’t find natural human social groups anywhere near so simple in structure as nonhuman primate social groups, and we can’t create really primitive human groups artificially. But information about the effects of biological endowment on such simple social groups may be useful to us in tracing our own psychological history. Other primates have had to face some of the same problems as we, and have had some of the same opportunities, based on similar physical structure and developmental patterns. Both human and nonhuman primate infants, for instance, are born unable to care for themselves, and must have parental attention over a period of years to survive. This means there’s a longer time in which learned behaviors can be passed from one generation to another, and in which opportunities for large amounts of social reinforcement abound. Also, the lack of efficient biological defense measures in both human and nonhuman primate adults means a greater necessity for mutual defense. Also — there are several dozen alsos, based on such biological similarities as these. Some of the other primates’ simple responses to such similar problems and opportunities may be the very ones with which we began our initially slow but gradually accelerating march to civilization.


Certain nonhuman primates live in environments that appear similar to those in which man’s immediate ancestors and earlv man lived, and in which man’s basic social patterns may have developed. Man’s direct ancestors may not have closely resembled today’s primates, but some were arboreal, as most primate species are today. Early Homo sapiens very likely existed rather like present-day savanna baboons in many areas, traveling on the ground and hunting for food during the daytime, sleeping in trees at night. (Adolph Schultz [1961] has pointed out that many areas where early man lived had too few caves to maintain an adequate population of “cavemen.”) Early man was not an individualist living in his own cozy hole or camping with his family under the stars. Most ground-dwelling primates live in groups; this is one of their major defenses against possible predators. If early man had a cave, he probably shared it; if he lived in the open, plenty of other men were around — at least if we can judge from what works for the baboon or rhesus. Man then gradually made various discoveries that enabled him to take advantage of other types of surroundings. But as with his biological structure, his early environment must have played a major role in forming the social patterns that gradually became elaborated into those we use now.


Paradoxically, certain differences we observe between our social relationships and those of other primates may reveal more than the similarities about our own social psychology. Given somewhat similar biology and similar early surroundings, why are the nonhuman primates as different from us in behavior as they are? In trying to discover why these differences are so great, we may be able to develop questions about human social behavior that we’d never have imagined if we’d limited ourselves to studying only humans. Maybe later we can develop answers to these questions.

Perhaps the major questions, from which many answers may flow, are: Why can we talk when other primates can’t? What is it about us in particular that has both allowed and compelled all human groups, even the most “primitive,” to develop complex languages, while other primates have failed to go beyond a few dozen strongly stimulus-tied sounds and gestures? What differences did this crucial distinction make in the early as well as the later social life of man? The first of these questions has already received the beginnings of some convincing answers (Bastian, 1965; Lancaster, 1968). Other primates not only lack vocal apparatuses capable of complex speech, but also lack certain elaborations of brain structure characteristic of humans. Evolution apparently has not endowed them with such brains because in their environments and with their patterns of living, speech-enabling brains would be of little advantage. As with the howler, most primates can get enough to eat and drink, and a decent place to sleep, with hardly a “word” to or from their fellows. Even species that have the capacity to acquire a simple symbolic language, such as the chimpanzee (Gardner and Gardner, 1969; Premack, 1971), apparently have not utilized this capacity in the wild; and until they find such capabilities useful, evolutionary processes will not select animals whose linguistic capacity is unusually high. These answers still need a great amount of detailed research support; but already, by getting into the vital human topic of language, we can see how stimulating the differences between us and the other primates may be.


Perhaps even more paradoxically, the behavioral differences we find not between humans and other primates, but among the various species of nonhuman primates themselves, may help us understand our own behavior better by making us more reluctant to apply any findings about animal behavior directly to human behavior. I’ve already mentioned the frequent assumption that a human behaves like a slightly more complex rat. The assumption has also been made occasionally that any significant behavior found in wild primates will appear in some form in tame humans. These assumptions don’t seem to include the recognition that because the many species of nonhuman primates show such wide ranges of behavior among themselves, no one behavior can be attributed to man until that behavior has itself been studied in man and established as part of his repertoire. Those who apply primate findings directly to man still fail to appreciate what primatologists have known for many years: that behaviors are species-specific, that species variability is not only physical but also behavioral and psychological. Some primates appear to defend a “territory”; some howl at others on sight; some don’t care. Some monopolize mates, or monogamize them; others live in pleasant promiscuity. Some show distinct dominance structures; others display apparent democracy (though not necessarily with female suffrage). Some pass their babies around; others protect or overprotect. Out of all this, where is man? Who can say? Who dare try? Some do try, but none should. Primate studies generate many questions about humans, and perform very usefully in that function. The answers should be obtained from humans themselves.

Even if the nonhuman primates were sufficiently cooperative to resemble each other more, we should not forget how much our generalizations from their behavior must be limited by their overwhelming differences from man. Not only their speech centers but their entire forebrains are vastly more limited than those of humans, not only in size (which may not be significant in itself) but in structure and function. The biological relationship is not as close as the fundamentalists fear; even the anthropoid apes cannot locate a common ancestor with us closer than one or two million years back. Nonhuman primate females do not naturally show continuous sexual receptivity; human females do, at least in a manner of speaking. Nonhuman primate babies never display that interesting evolutionary invention, the social smile, which certain stimuli will elicit automatically in human babies at ten to twelve weeks of age. When other primates do defend their “territory,” they stop at its edge; they do not engage in crusades into enemy territory, defensively killing and raping the opposition. The search for novelty in other primates is not absent, but it is muted in comparison with man. Man’s dissatisfaction apparently did not come from losing the easy availability of food and drink; that came after he was dissatisfied, and because he was dissatisfied, as the Good Book says.

We so often insist that every bit of knowledge discovered by science must mean something for humans, that it may be hard to accept the lack of such immediate meaning in research on primates or other animals. But again, the lack of direct applicability — the fact that I cannot list for you all the positive contributions that primate research makes to humanity — should not conceal its great heuristic value, its capacity for revealing to us important questions we must ask about the foundations of human social psychology. To know whether a particular motivation observed in gorillas also occurs in humans, how strong it is in humans, whether it may be strengthened or weakened or deflected and in what manner, we must turn to the empirical study of man himself, however much more complicated that may prove to be. But monkeys and apes, by their very unlikeness to us, will compel us to make this complicated effort.

[From Social Psychology and Social Relevance, by Alan C. Elms, pp. 12-45. Originally published and copyrighted by Little, Brown and Company, Boston, MA, 1972. Copyright reassigned to Alan C. Elms.]

0 comments on “Social Primatology (Social Psychology and Social Relevance, Chapter 2)Add yours →

Leave a Reply

Your email address will not be published. Required fields are marked *