MultiO Studies
Sex as an Alternative to Aggression in the Bonobo
by Franz B. M. de Waal and Takayoshi Kano
(Excerpt. Sexual Nature, Sexual Culture (Ed. by P. Abramson & S. Pinkerton). Chicago: University of Chicago Press, 1995. [pp. 37-56],
Takayoshi Kano. The Last Ape: Pygmy Chimpanzee Behavior and Ecology. Stanford: Stanford University Press, 1992

For years it was thought only humans engaged in face-to-face sex. But it turns out there is at least one other creature that enjoys frontal intercourse — our closest cousin, the bonobo. They are also the only other animal we know of who use tongues whilst kissing. And they do it a lot. It's not all just for fun or procreation: sex plays an important role in resolving disputes and maintaining social cohesion. Put simply, bonobo societies get along brilliantly because they're always getting it on with each other. In any possible combination. Girls carry other girls about on their backs and rub clitorises together. Boys find other boys and rub their rumps together. They happily engage in fellatio and mutual masturbation all over the place. And they're not selfish about it either — during sex the male bonobo will alter the speed and intensity of his thrusting based on the facial expression and vocalisations of the female. Whereas other animals get into fights over food or territory, bonobos just have sex instead. Imagine a world where all your petty little disputes are solved like that. “Oh, I'm sorry, Jennifer Garner, I appear to have eaten your termites. Listen, I'd hate for there to be any bad feeling between us. So...”
   (Defoe)

INTRODUCTION
If, as both Christian doctrine and biology would have it, the purpose of sex is reproduction, why do millions of human couples engage in it on a daily or weekly basis even though the average family-size in industrialized nations has dropped to only one or two children? Perhaps they do so simply because it makes them feel good, almost like an addiction. Yet this possibility raises the question: Why does it make people feel so good? After all, most other animals restrict their mating to a particular season or a couple of days in their ovulatory cycles; they do not seem to feel any need for nonreproductive sex. Are we the only ones with such sexual appetites, or are there other species like us?

There is one, and it can hardly be coincidental that it is, together with the chimpanzee (Pan troglodytes), our nearest relative. Known as the bonobo (Pan paniscus), or pygmy chimpanzee, the creature was discovered only in the 1920s (see Coolidge, 1984, and de Waal, 1989), and has recently become a hot topic of debate among primatologists. Whereas the chimpanzee shows little variation in the sexual act, bonobos behave as if they have read the Kama Sutra, performing every position and variation one can imagine. Yet, their rate of reproduction in the wild is approximately the same as that of the chimpanzee, with single births to a female at intervals of five to six years. So, bonobos share with us both a rich sexuality and a partial divorce of sex from reproduction.

In biology, sexual behavior generally is investigated from the perspective of reproduction. Although the nonreproductive use of the same behavior is common to many species, this use is considered of secondary importance. From an evolutionary perspective, the primary function of sexual behavior, that is, the function most directly relevant for natural selection, is its capacity of producing a zygote. But if fertile and infertile partner combinations engage in sexual behavior with equal intensity and equal frequency, as is the case in the bonobo, it would seem that the reproductive function has decreased in relative importance.

Before a partial divorce between sexual behavior and its fertilization function is accepted, two conditions must be met. First, sexual contact in infertile partner combinations—such as individuals of the same sex or adults and juveniles—should not be a mere substitute for heterosexual contact between adults. In other words, given a choice, the animals should not necessarily give priority to the second type of contact. The frequent isosexual mounts that Yamagiwa (1987) observed within an all-male band of wild gorillas (Gorilla gorilla beringei) probably are examples of redirected sex. The dominant silverback males of this band treated the younger males as a “harem,” competing over ownership in the same way known of bisexual units. Yamagiwa suggested that the formation of this special male band may have been related to an increased male/female ratio in the population, i.e., a relative lack of females.

Second, infertile sexual contact would be considered deviant behavior unless such contact is functionally integrated into the natural social life of the species. To give an extreme example: a sexual motor pattern that is observed exclusively in individuals reared in isolation would not meet the requirement because it is not possible that sexual behavior evolved under this condition. In short, it would be accepted that sexual motor patterns serve functions other than fertilization if their infertile use occurs both by choice and in a naturally adaptive context. Both criteria must be kept in mind, as the current study concerns captive bonobos.

The following descriptions are given from an unusual perspective, because my main research focus is not primate sex per se but rather aggression and aggression-control. Many primate species have evolved special reassurance gestures that maintain peaceful relationships. Calming behavior occurs in response to social tensions (i.e., when there exists a high probability of interindividual conflict) and in the aftermath of fights. Reunions between former opponents following aggression, known as reconciliations, are characterized by kissing and embracing in the chimpanzee (de Waal & van Roosmalen, 1979), whereas these contacts often involve genital stimulation in its congener, the bonobo (de Waal, 1987).

Neither the exchange of sexual signals outside the copulatory context nor nonreproductive mounting are absolutely restricted to the bonobo. Wickler (1967) proposed the label “sociosexual behavior” to cover the wider employment of sexual patterns. For example, de Waal and Ren (1988) reported a significant increase in “hold-bottom” gestures among stumptail macaques (Macaca arctoides) during reconciliations following fights. These monkeys even may show their so-called “orgasm face” during postconflict reunions: a particular facial expression demonstrably associated with physiological signs of sexual climax, such as increased heart rate and uterine contractions (Goldfoot et al., 1980). Normally, these behavior patterns are part of copulations in this species (Nieuwenhuijsen, 1985). The bonobo goes further than the stumptail macaque, and most other primates, however, in that complete sexual sequences are used for nonreproductive purposes rather than mere sexual gestures or perfunctory mounts.

Sociosexual Behavior
Traditionally, face-to-face copulation has been seen as reflecting the dignity and sensibility that separate civilized humans from subhumans. It was believed, for example, that preliterate people would benefit from education about this mode of intercourse, hence the term “missionary position.” The position was elevated to a cultural innovation of great significance, one which, according to Hockett and Archer (1963, 34), fundamentally changed the relationship between men and women: “Our guess is that it changed, for the adult female, the relative roles of the adult male and of the infant, since after the innovation there is a much closer similarity for her between her reception of an infant and of a lover. This may have helped to spread the “tender emotions” of mammalian mother-infant relations to other interpersonal relationships within the band, ultimately with such further consequences as the Oedipus complex.” Wescott (1963, 92) took this idea a step further: “It seems not only that the adult male becomes, in face-to-face copulation, a surrogate suckling to the adult female by virtue of his position; but also that the adult female becomes a surrogate suckling to the adult male by virtue of her behavior, which is that of soliciting and receiving a life-giving liquid from an adult bodily protuberance.”

Not surprisingly, given the entirely cultural interpretation often given to face-to-face copulation, Tratz and Heck (1954) carefully wrapped their findings in Latin when first reporting that chimpanzees mate more canum (like dogs) and bonobos more hominum (like people). Published in the German language, their study as well as other early observations of bonobo behavior (Rempe, 1961; Kirchshofer, 1962) were ignored by the scientific establishment, however. The bonobo’s humanlike sexuality needed to be rediscovered, in the 1970s, before it became widely accepted as a species characteristic. This rediscovery was based on studies of both captive bonobos and fieldwork by a Japanese team at Wamba and a Euro-American team at Lomako Forest, both in Zaire (Hübsch, 1970; Jordan, 1977; Savage and Bakeman, 1978; Savage-Rumbaugh and Wilkerson, 1978; Kano, 1980; Kuroda, 1980, 1984; Thompson-Handier et al, 1984; Dahl, 1985, 1986, 1987; Blount, 1990).

My own study concerned the world’s largest captive collection of bonobos, at the San Diego Zoological Garden. Ten of these rare apes were kept in three separate subgroups. One subgroup included an adult male/female pair and an adolescent male. Another subgroup consisted of a mother-infant pair and an adolescent male. The third subgroup included four juveniles, two of each sex. The first two subgroups were merged in the course of the study, allowing the observation of interactions between adult females and between a fully adult male and adolescent males. I observed the apes for nearly 300 hours, standing in front of their enclosure. Oral accounts of the bonobos’ social behavior were recorded either on an audiorecorder or on the audio channel of a videorecorder. Video was added at moments of great social activity, such as at feeding time or during the introduction of new group members.

From these observations, and the studies mentioned before, the following conclusions can be drawn regarding the bonobo’s sexual peculiarities.

Extended Receptivity
The period of sexual receptivity of bonobo females is dramatically extended compared to the chimpanzee, and most other primates. Whereas the chimpanzee has a menstrual cycle of approximately 35 days, the bonobo’s is closer to 45 days, and the period of genital tumescence encompasses a greater proportion of the cycle than in the chimpanzee (Furuichi, 1987; Dahl et al., 1991). In captivity, low-ranking bonobo females may copulate throughout most of the cycle, but females in the wild or high-ranking captive females (i.e., females with greater control over when and where to engage in sexual intercourse) are most receptive during the phase of maximum swelling (Dahl, 1987; de Waal, 1987; Furuichi, 1992). Also, the rhythm of male thrusting—perhaps a measure of sexual arousal—is higher during matings with tumescent than with detumescent females (de Waal, 1987).


Figure 1. Ventro-ventral copulation between an adolescent male and adult female, while the female’s infant daughter looks on (infants often try to squeeze themselves between adults so as to “participate” in the intercourse). Photo by Franz de Waal.

Variable Mounting Positions
Because the clitoris and vulva are ventrally oriented, the female’s genital anatomy seems adapted for face-to-face copulation, a frequently adopted position (fig. 1). Table 1 provides the distribution of six behavior patterns over 42 dyadic directions among the San Diego bonobos (2 directions per pair of individuals) and the combined frequency of sociosexual interaction per dyadic direction per hour of observation. Dyadic directions have been arranged according to the rate with which sociosexual behavior was initiated (the actor is defined as the individual making the first invitational gesture or approach, even if that individual is not the more sexually active partner).

Table 1. Frequency of six sociosexual interaction patterns
among different partner categories of bonobos at the San Diego Zoo.
Initiator Partner Mounting Oral and Manual Total Per
Hour
Ventro-
ventral		 Ventro
dorsal		 Opposite Genital
massage		 Oral
sex		 Mouth-
kiss
Adol. Male Infant 4965      114 1.64
Adol. Male Female * 63 18      811.17
Infant Adol. Male47 2 3   1 530.76
Male Adol. Male5 7  23  1 360.64
Female Female37  4    142 0.63
Adol. Male Female * 33 1      340.51
Male Infant19 8      27 0.49
InfantMale 15 5      200.36
Female Female10 1 11   1 230.34
Juv. Male Juv. Male  9    314 26 0.32
Adol. MaleFemale * 25 8  1   34 0.31
MaleAdol. Male 11 71 10  2 310.28
Juv. Male Juv. Male 7 1  15 6 200.25
Adol. Male Adol. Male 12 3      150.22
Adol. Male Male 28  1    110.20
Adol. Male Infant 164 1     210.19
Male Female *9        90.16
Male Female *17        170.15
Juv. Female Juv. Male     1 43 8 0.10
FemaleInfant 4 2      60.09
Female Adol. Male * 23      5 0.07
Infant Adol. Male8        80.07
Adol. Male Male 3  2  12 8 0.07
Juv. FemaleJuv. Male 2     3  50.06
Female Male *2 1      3 0.05
Adol. MaleFemale * 5   1    60.05
Female Adol. Male * 3       3 0.04
Juv. FemaleJuv. Male 3       3 0.04
Juv. Male Juv. Female        33 0.04
Juv. Female Juv. Male        33 0.04
Juv. Male Juv. Female 1     11 3 0.04
Juv. MaleJuv. Female 1      2 30.04
Infant Female  1 1    2 0.03
Adol. Male Adol. Male 2       2 0.03
Female Infant3        30.03
Female Adol. Male * 21      3 0.02
Female Adol. Male * 1      12 0.02
Female Male *  1  1   2 0.02
Juv. Male Juv. Female        11 0.01
Juv. Female Juv. Female        11 0.01
Infant Female1        10.01
Juv. Female Juv. Female         0 0.00
Total  420 15623 39 1743 698 

Adults are indicated as Male and Female; adolescents as Adol.;
juveniles as Juv., and the two-year-old female infant as Infant.
Each of 42 directions between individuals is represented separately.
Dyadic directions are ordered according to the rate of sociosexual contacts
initiated per hour of observation (observation time is not the same for all dyads).
Asterisks mark potentially fertile partner combinations.


By far the most common pattern was the ventro-ventral mount, observed in no fewer than 33 different dyadic directions. There is no dyadic category in which this behavior did not occur, except for the one female-female relationship among juveniles. Characteristic are genital rubbing movements between females mounted in a ventro-ventral position, one female carrying the other (fig. 2). This carrying posture—with one female being lifted off the ground while she clings to her partner much like an infant clings to its mother—allows both females to make sideways rubbing movements. The females rub their clitori together with an average of 2.5 lateral moves per second; this is approximately the same rhythm as that of a thrusting male (de Waal, 1987). The pattern is now widely known as GG-rubbing; an abbreviation of genito-genital rubbing first proposed by Kuroda (1980). It has been observed by all students of bonobo behavior, and is unique to the species.


Figure 2. Two adult females during GG-rubbing, both with extensive baring of the teeth at the climax. The bottom female also squeals. Notice the eye-contact. Photo by Franz de Waal.
We cannot say, however, that aggressive interactions never occur between females. Occasionally, such as right after arrival at the feeding site, a female may jump on a neighboring female, drag and bite her, hold her down, and steal her sugarcane. Such attacks by females differ from those by males. Before physical contact, males glare at each other and usually progress through stages of threat in which they approach while glaring. Aggression between females, however, has the special feature of occurring suddenly, without any warning.

There are times when aggression between two females starts silently, but then screams rise as a fight develops and the two combatants begin rolling around on the ground. Other females may join, creating great confusion. In contrast, a fight between males is usually one-sided from start to finish, and is over quickly. Moreover, a female whose food was snatched and who was treated badly may roll on the ground on her back and scream in a ''temper tantrum,” whereas adult males do not express this kind of childish anger.

The pattern of antagonistic behavior between females strongly suggests that dominant-subordinate relationships (or the rank system) are underdeveloped in females compared to males. In general, relations between females are peaceful. Even when feeding close to each other, they are usually relaxed, and they rarely show expressions of submission such as screaming or grimacing.

The most conspicuous and striking social behavior between female peers is genito-genital (GG) rubbing, in which females rub against each other’s external genitalia. The following is a typical example. First, female A approaches female B, who is in the midst of feeding. Female A, who approached in an unconcerned manner, lingers a bit and sits down at a distance, but within easy reach of B. Then, “a demand for invitation” follows, either by A or B. Let us suppose that A makes the invitation. A stands bipedally, extends her hand to B, puts her face close to B’s face, and peers directly at her. If they are in a tree, A grasps a branch of the tree with one or both hands and peers at B while hanging above her. If B does not react, A may grasp B’s knees with her feet, or some other body part, and shake her. Alternatively, A may touch B’s shoulder and peer at B in a request. These are all requests that say, “Please associate with me in genital rubbing.” Then, B rolls over on her back and spreads her thighs. When A starts to mount her, B embraces her ventrally as if she were a baby. A and B put their genitalia together and rub them against each other. If in a tree, one will hang from a limb above and open her thighs. Her partner will face toward the offered genitalia, and will either embrace or hang from the same branch as they begin to GG rub.

The part of the sexual organ that touches is the pointed part at the tip (clitoris). The quick, rhythmic thrusting differs from copulation, in that the hips sway from side to side. A female engaged in GG rubbing juts out her hips, and often has an expression (called “pout face”) that conveys an uncontrollable emotion. She emits the same kind of scream as when copulating.
 
GG rubbing is evidently comparable to copulation between female partners. The pygmy chimpanzee is the only one, out of 200 species of primates, to devise this behavior, and whenever they are delightedly absorbed in this “lesbian” behavior, they seem proud of their splendid invention.

GG rubbing often occurs in the same social context as copulation, rump-rubbing, and mounting, when excitement is elevated because of tension between individuals. The frequency of GG rubbing is highest for a while after arrival at the provisioning site, and we infer that this seemingly erotic behavior has the important social function of reducing tension between females. The effect of the GG rubbing may be that females are able to contact each other openly, even while crowded together. In this way, female pygmy chimpanzees raise the upper limit of their sociality and succeed in creating conditions for binding themselves into intimate social relationships within a unit group.

Aside from the unique nature of this behavioral form, the striking thing about GG rubbing is that it creates a harmonious stability in the relationship between two individuals, not by fixing a superior-inferior relationship as in a rank system, but by creating a condition of equality.

The females who approach and “beg” for an invitation to GG rub are younger and, consequently, many probably have lower status. Influenced by females who are senior in power, perhaps they obtain their own '‘sense of security” by relating through GG rubbing and soliciting “invitations ”

Older, influential females do not perform dominance displays to show off their high position. (I call them “influential” because we do not use the word “high-rank” with respect to females.) These older females solicit various social contacts from their followers such as grooming, GG rubbing, and copulation. They seldom receive threats or attacks from others (including males), and also rarely display aggressive behavior. They are respected out of affection, not because their rank is high.

Curiously, when a female wants GG rubbing, an influential female will lie on her back, which tends to lead to mounting. When the females are close in age, they both present lying on their back, and may even do a compromise half-mount. If we were on the bottom being held down, we would probably feel submissive and inferior, but female pygmy chimpanzees seem not to take it that way. When GG rubbing, the female on the bottom takes the position of the female during ventral copulation. In this position, she looks proud and affectionate.

The only copulatory position in most primate species is the dorsal mount. In Japanese macaques and baboons, the male who takes that dorsal position is the dominant one, except in special cases. The female who is mounted is in a condition of nonresistance, facing away from her partner. The mounter has in his field of view the whole body of his partner, but the female, if she does not look back, cannot even see the facial expression of her mounter. By contrast, in the ventral copulatory position, there may not be an element of dominance-subordinance. Except for the physical differences of being on top or being on the bottom, the positions of both actors are identical and equal in status. Each can read the other’s facial expressions.

The external sexual organs of a female engaged in GG rubbing do not always reach maximal tumescence. The larger her sexual organ and the more distended her swelling, however, the more attractive she is. An attractive female is often invited, and also often invites, GG rubbing. Consequently, the long estrous period of pygmy chimpanzees is not just for engaging in copulation, but also for GG rubbing. That is, it is necessary in creating stability in the relationships between females and in deepening friendships.

The external genitalia that are distinctive of female pygmy chimpanzees have a location and shape that are well-suited to females facing each other to rub their clitorides. I am convinced that this evolved for GG rubbing rather than for ventral copulation.
   (Kano)

Figure A. A female screams during GG rubbing. Photo by Takayoshi Kano.

Another rare mounting pattern occurred between adult females, with the partners facing in opposite directions. While one female lies on her back, the other stands over her, with her back turned, rubbing her genitals against her recumbent partner’s. A similar posture occurs between males, with both males standing quadrupedally back-to-back, rubbing their scrota together.
One important element characterizing the relationship between adult males is aggressive interactions. Adult males have the highest frequency of aggressive interactions, relative to associations between members of other age-sex classes.

The term “aggression” has been defined as behavior in which one or several individuals injure and threaten or try to injure another (McKenna, 1983). Aggressive behavior is inevitably accompanied by a reaction. The reaction may be aggressive, or it may be submissive. We also use the term “agonistic” behavior for behaviors exhibited by both aggressor and victim.

The aggressive behavior pattern of pygmy chimpanzees abounds with variety, from violence, including physical contacts such as biting, hitting, kicking, slapping, grabbing, dragging, brushing aside, pinning down, and shoving aside, to glaring, bluff charging (the appearance of charging), charging, and chasing. A pygmy chimpanzee may also approach another with exaggerated gestures, wave his arms around the other’s head, and leap over the other’s body, threatening body contact. Aggressive behavior is accompanied by a vocalization that sounds like “kat-kat.”

Prostration, grimacing, flight, avoidance, extending one’s hands, and touching the other’s body are classes of submissive behaviors in response to aggression. Three kinds of shrieks are emitted by victims. They are, in order of increasing intensity, “gyaa-gyaa,” “kii-kii,” and “ket-ket.”

Low-level aggressive interactions, such as approach-avoidance or bluff charge-flight, occur very frequently between males and seldom involve other patterns of behavior. Furthermore, the aggressor and victim are constant; that is, the victim and aggressor do not switch roles.

Aggressive interactions mainly occur at feeding time and, most often, when individuals gather in the trees to feed on concentrated foods such as Dialium and batofe. This suggests that proximity between individuals may induce aggressive behavior.

Successive mounting or rump-rubbing often occur during aggressive interactions. Mounting is riding horseback on another, taking the position used in dorsal copulation. In rump-rubbing, two individuals in a “presenting” posture direct their rear ends toward each other and press their buttocks against each other. Sometimes the positions are assumed and held motionless, but usually they are accompanied by slight rhythmical thrusting movements. Frequently, both actors have erect penises, but insertion into the anus during mounting has not yet been verified.


Figure B. Two males rump-rubbing. Photo by Takayoshi Kano.

In an aggressive interaction, the attacker may spring on a male who, cut off from escape, is groveling and screaming, and the attacker will mount or rump-rub the victim. Or, the attacker may confront his victim, suddenly facing the victim’s buttocks and demanding mounting or rump-rubbing. The order of these events is not fixed.


Figure C. A male grimaces while mounting a higher-ranking male. Photo by Takayoshi Kano.

Immediately after arrival at the feeding place, a male may make a gesture similar to a sexual display to another male, and then, either mounting or rump-rubbing may occur. Mounting and rump-rubbing may have the same function because the social context in which they occur is practically the same. These actions put an end to aggressive interactions, preventing anything injurious from happening, and have the effect of '‘appeasement” or “pacification.” The dominant male presents a guarantee of safety to the subordinate.
   (Kano)
In contrast, the posture during so-called mutual penis rubbing resembles that of a heterosexual mating, with one male (usually the younger) passively on his back, the other male thrusting on him. Because both males have an erection and because intromission does not occur, their penises rub together. Attempts to achieve anal intromission were not observed, and ejaculation never resulted. Kano (1989) further describes penis fencing: a rare behavior, thus far observed only in the wild, in which two males hang face to face from a branch while rubbing their erect penises together as if crossing swords.


Figure 3. A 7-year-old adolescent male is mounted ventro-ventrally by a 2-year-old infant. The infant presses her vulva against the male’s erect penis (no intromission occurs) while he performs a series of upward pelvic thrusts. Photo by Franz de Waal.

A frequent posture involving the female infant was one in which she climbed on the belly of an adolescent male and pressed her vulva against his genitals, whereupon the male, either in a sitting or a recumbent position, made a series of pelvic thrusts (fig. 3). On other occasions, the infant presented for a ventro-dorsal mount. Mounts with this infant never resulted in intromission or ejaculation. As can be gathered from table 1, sociosexual contacts of adults and adolescents with the infant were frequently initiated by the infant herself (i.e., 32.9 percent of the time).

Ventro-dorsal mounting positions are typical of most primates. Chimpanzees, for instance, mate in this position almost exclusively (McGinnis, 1973; Savage-Rumbaugh and Wilkerson, 1978; personal observation). The low frequency of ventro-dorsal relative to ventro-ventral mounts in the San Diego colony may not be general for the bonobo; according to most reports on other populations, both captive and wild, the ventro-dorsal mounting position is the more common one, employed between 62 and 74 percent of the time (Jordan, 1977; Kano, 1980; Thompson-Handier et al., 1984). The important point, though, is that thus far all investigators have reported the regular use of both positions, which means that both can be considered species-typical.


Figure 4. Mouth-to-mouth kissing, with tongue interaction, between two juvenile males. Photo by Franz de Waal.

Other Acts of Stimulation
Two oral sociosexual patterns occurred almost exclusively in the group of juveniles (83.3 percent; table 1). The first is mouth-mouth kissing, i.e., one partner places his or her open mouth over that of the other. In about one-quarter of the instances this involved extensive tongue-tongue contact (fig. 4). The second pattern is fellatio, i.e., one partner taking the penis of another in the mouth. These two sociosexual patterns frequently occurred in the context of rough-and-tumble play. A bout of chasing and wrestling would be interrupted by sociosexual games in which all four juveniles might participate, some of them mounting, others engaging in the just-described oral patterns. Play would resume within a few minutes.

The sixth sociosexual behavior pattern is manual massage of another individual’s genitals. The large majority of instances (84.6 percent) was directed by the adult male to one of the adolescent males. The younger male, with back straight and legs apart, would present his erect penis to the adult male, who would loosely close his hand around the shaft, making caressing up-and-down movements (fig. 5).


Figure 5. A 8-year-old adolescent male presents his erect penis to the adult male (left) following an aggressive incident between them. The adult male performs a genital massage. Photo by Franz de Waal.

This pattern is the social equivalent of masturbation, which the bonobos also showed. Neither genital massage nor masturbation was observed to produce ejaculation. Of the 39 observed masturbations, none were performed by the adult male, 58.9 percent were performed by the two adolescent males, 23.1 percent by the two adult females, and 17.9 percent by immatures. Another form of self-stimulation was nipple-caressing, i.e., an individual stimulating one or both nipples with rapid movements of the thumb(s). This behavior may have served a self-reassurance function as it occurred often after frustrating events, such as when begging for food had been unsuccessful or an aggressive rebuff had been received. The behavior was observed in five different individuals of both sexes, but 82.8 percent of the 99 instances concerned a single adolescent male (fig. 6).


Figure 6. At times of stress or insecurity some individuals would self-caress their nipples with rapid movements of the thumbs, here shown by an adolescent male. Photo by Franz de Waal.

Pleasure
To what extent these sociosexual patterns are accompanied by pleasure is of course impossible to determine through mere observation, but if the sounds and facial expressions of this species are any indication, we estimate these contacts to be gratifying. Females frequently bared their teeth during sex (fig. 2), particularly towards the end of the interaction (at its climax), an expression interpreted as an Orgasmusgesicht by Becker (1984) and as a pleasure grin by de Waal (1988). Wide baring of the teeth also occurred during solitary masturbation, or during excitement over new play objects or food. Furthermore, females often uttered characteristic screams or squeals before or during copulation (Savage-Rumbaugh and Wilkerson, 1978; de Waal, 1988).

Partner Choice
Assuming that the two adolescent males were fertile (both were capable of semen production), 12 dyadic directions concern potentially fertile partner combinations (table 1). The mean (± SE) hourly rate of sociosexual initiatives in these directions was 0.22 ± 0.10, compared to 0.23 ± 0.06 in fertile directions. The difference between these two means is nonsignificant (t-test, t = 0.16, df = 40, P > 0.05). If the analysis is limited to mounting behavior or if the data are combined for both directions per dyad, the same conclusion applies, that is, fertile and infertile partner combinations show virtually identical rates of sociosexual behavior.

Another way of summarizing the data is to calculate the rate of sociosexual initiatives dependent on two factors: (a) relative age of the partner and (b) sex combination. Four age classes were distinguished: adult (10 years and older); adolescent (7-9 years); juvenile (3-6 years); and infant (0-3 years). Partners can belong to the same age class as actors or to an older or a younger class. Three intersex directions were distinguished: male toward female; female toward male; and same-sex combinations. This analysis indicated a relatively high level of heterosexual initiatives toward older partners by both males and females. Males more often initiated contact with younger partners of the opposite sex than did females. Heterosexual contacts were relatively uncommon between partners of the same age class. Isosexual initiatives (interactions between partners of the same sex), finally, were mainly directed at partners of the actor’s own or a younger age class. See de Waal (1990) for a more detailed presentation.

Partner choice was not unlimited at the San Diego Zoo, however. The two juvenile females, for example, exhibited relatively low rates of sexual initiative, yet might have been more active in the presence of adults. All that can be concluded is that given the conditions under which these bonobos lived— allowing all individuals a choice of partner sex and most individuals a choice of several partner age classes—no evidence was found that potentially fertile partner combinations engaged in sociosexual behavior more often than infertile combinations.

Moreover, it must be emphasized that heterosexual intercourse between sexually mature individuals is only potentially fertile: first, because not every copulation involves ejaculation, and second because females are fertile only during a few days of their menstrual cycle. As an external sign of receptivity, they develop a conspicuous genital swelling, but this is an unreliable indicator of fertility; the swelling phase far exceeds the period of ovulation, and swellings also may be shown by pregnant or lactating, hence noncycling, females (Dahl, 1986).

Whereas it is virtually impossible to know with certainty which copulations involve an ejaculating male with an ovulating female, it is not difficult to distinguish, on the basis of behavior, infertile mounts. These are mounts without intromission; mounts broken off well before the male partner slowed down for the final, deeper thrusts indicative of ejaculation; and/or mounts involving a female with detumescent genitals. Because the dominant male tended to interfere with heterosexual intercourse of the adolescent males and because the females mated throughout all cycle phases (cf. Savage-Rumbaugh and Wilkerson, 1978; Thompson-Handier et al., 1984; Dahl, 1987), fertilization could be excluded for a considerable number of mounts in partner combinations marked as potentially fertile in table 1.

Lest the rich variety of sexual patterns and the bonobo’s evident bisexuality leaves the impression of a pathologically oversexed species, I must add, based on hundreds of hours of watching bonobos, that their sexual activity is rather casual and relaxed; it appears a completely natural part of their group life. Also, even though the bonobo is a serious contender for the title of sex champion of the primate world, including the prize for the longest penis—a distinction thus far erroneously reserved for Homo sapiens (e.g., Morris, 1967)—its sexiness should also not be exaggerated. With the average copulation lasting 13 seconds at the San Diego Zoo (de Waal, 1987) and 15 seconds in the wild (Kano, 1992), sexual contact among bonobos is rather quick by human standards. Thus, instead of an endless orgy, we see a social life peppered by brief moments of sexual activity.

An average copulatory bout (from the male’s mount until his dismount) lasted 15.3 seconds at the feeding site. During that time, the male maintains a rapid, rhythmical thrusting motion. From observations of 8-mm film, the average number of thrusts per second and per copulation were 2.7 times and 43.8 times, respectively.

Whether or not ejaculation occurs during copulation is difficult to determine, and the female’s vaginal plug, formed by the white ejaculate, is rarely seen. For two reasons, however, I think that ejaculation during copulation is not common. The first reason is that after copulation the male’s penis is frequently still erect. The second is that the number of copulations per day per male is great. It is not unusual for the same male to copulate several times in one day. At the feeding site alone, the same individual was seen copulating 11 times, the maximum number of copulations per day recorded for one male. At the feeding station, the total number of copulations by males who copulated twice or more in one day was 252, approximately half the number of all observed copulations. Because some copulations must occur outside the view of the observer, males probably engage in many copulations every day.

During dorso-ventral copulation, females frequently extend a hand or foot between the male’s legs to touch the bottom of the scrotum. The variations on this theme are great, from pressing against the scrotum to intensify the pressure of the thrusting movement to not really touching the scrotum at all in a merely perfunctory manner. A female in the same posture will occasionally push the lower part of her genitalia upward as if to increase the friction of a full thrust of the male’s penis.

Many copulations happen silently, but also commonly, the female screams when a couple approaches the end. This scream may relate to ejaculation and female orgasm, because often when the same couple copulates repeatedly in a short interval, the scream is more likely to be heard during the last copulation. All of the copulatory screams of female partners, however, are not related to ejaculation or female orgasm. For example, I observed a case in which a female gave a scream while copulating with a male who had no external sexual organs except a stunted penis.

After copulation, one or both partners remain sitting in the same place or separate from each other by several meters. Very rarely, after a male dismounts a young female, she gives a scream and runs several meters away. Usually, however, both partners are calm after copulation.

Pygmy chimpanzees are promiscuous. All kinds of pairings among group members are possible, except between a mother and her mature son. Consequently, members of the same sex within a group can all be sexual competitors. Adult pygmy chimpanzees apparently are generally tolerant of each other's mating activities. We cannot assert this unconditionally, however, because a male may intentionally solicit and copulate with a female who is separated from other members precisely to avoid interference from other individuals.
   (Kano)
Recently, I observed a colony of chimpanzees at the Yerkes Primate Center in a study modeled after the bonobo study. The rate of sociosexual behavior by adult and adolescent chimpanzees was significantly lower than that by the same age category in bonobos (fig. 7; t-test, t = 5.23, df = 16, P = 0.0001). This difference existed despite a greater partner choice for the chimpanzees, which lived in a group of 19 individuals. Rates of sociosexual behavior of immatures were similar in the two species.


Figure 7. Mean (+ SEM) number of sociosexual acts initiated per hour per individual for the San Diego bonobos and an outdoor colony of chimpanzees at the Yerkes Regional Primate Research Center’s Field Station, separately for (a) adults and adolescents (black) and (b) younger individuals (grey).

Context of Sociosexual Behavior
Seven conditions were distinguished in my continuous records of bonobo behavior:
    Pre-feeding   15 min prior to feeding time
      Feeding-1     first 15 min of the feeding session
      Feeding-2     second 15 min block following feeding
      Pre-conflict     15 min prior to an aggressive conflict unrelated to food
      Post-conflict-1     first 15 min following aggressive conflict
      Post-conflict-2     second 15 min block following aggressive conflict
      Baseline     all remaining observation time

Hourly behavioral rates under these conditions were compared for each pair of individuals. The following changes occurred in a significant majority of dyads (de Waal, 1987):
       1.    An increase in aggression following food provisioning.
       2.    A decrease in social grooming following food provisioning.
       3.    An increase in nongrooming contact, including sociosexual behavior, following both food provisioning and aggressive incidents unrelated to food (fig. 8).


Figure 8. Mean (+ SEM) rate per partner combination of sociosexual and affiliative interactions, excluding grooming, in the San Diego Zoo bonobo colony. The rate is the percentage of 5-minute blocks in which the behavior occurred. Grey: three 15-min periods around feeding time (before the introduction of food and immediately following food provisioning). Black: three 15-min periods relating to spontaneous aggressive incidents in the absence of food (before aggression occurred, and immediately following aggression). The baseline concerns all remaining observation time. From de Waal (1987).

Hence, food provisioning stimulated the bonobos’ affiliative and sociosexual behavior (with the exception of grooming behavior, which was suppressed). In theory, this effect could have been caused either directly, by the presence of food (excitement over food is transformed into sexual arousal), or indirectly, in response to the competitive atmosphere inevitably created by an attractive resource (as reflected in the observed aggression increase). The second hypothesis is better supported, because affiliative and sociosexual behavior also increased following aggression that was not food-related. In other words, the presence of food was no prerequisite for sociosexual behavior, and the causal factor most elegantly explaining both measured increases was tension among individuals.

Typically, the bonobos would become very active upon introduction of food, engaging in aggressive competition but also inviting one another for sociosexual contact. These contacts appeared to reduce the tension and to allow for food sharing. Thus, it could be demonstrated that subordinate group members more often removed food from the hands of dominant possessors following a sociosexual contact than without such prior contact (de Waal, 1987). The interaction could even take the form of an exchange, e.g., a female presents to a male who is holding a large bundle of branches with leaves and takes the entire bundle out of his hands following sexual intercourse. On other occasions, sociosexual behavior was used as a reconciliation. The majority of instances of genital massage, for instance, followed aggressive incidents in which the adult male had chased one of the adolescent males. After a couple of minutes, the younger male would return to the aggressor to present his genitals.

In short, the bonobo’s sociosexual behavior serves important tension-regulating functions. These functions explain why the behavior occurs with high frequency in all possible partner combinations, because tensions are obviously not limited to male-female combinations. It is instructive to compare these observations with the behavior of chimpanzees. This species regulates social tension by means of nongenital contact forms; in various studies of captive chimpanzees, mounting and mating never even ranked among the ten most common modes of reconciliation (de Waal and van Roosmalen, 1979; de Waal, 1992).
Immature and Adolescent Social Relationships

Because an offspring is carried by its mother almost all the time, during infancy there are few opportunities for others to contact the infant. About six months after birth, the infant extends its delicate hands to other individuals that approach the mother. These are the infant’s first attempts to make active social contact. The majority of attempts of this kind are ignored, but occasionally, the infant receives a finger from somebody. When the mother is grooming someone else, the infant often crawls over the bodies of both adults, but they do not get annoyed or push it aside. The infant gradually increases its activity and broadens its behavioral repertoire. Elders are tolerant of behavior of any kind until the infant is about two years old.

Male infants are more precocious than females and begin to show sexual behavior in less than one year. When a mother finishes GG rubbing, her male infant clings to her partner and inserts his erect penis into her. Throughout the juvenile period, sexual behavior gradually becomes a regular activity. When the juvenile encounters his mother or other adults engaging in copulation and GG rubbing, he immediately runs and clings to either one’s stomach or back, and screams. Then, when the adults conclude their activity, they embrace the juvenile and practice similar behaviors with him. There are juveniles that cannot wait to thrust; they cling to the male’s hips during copulation and insert their penis in the female in the midst of GG rubbing.

Adult males in the midst of copulation sometimes get annoyed and push away clinging juveniles with their elbows, but adults are generally cooperative in awakening the erotic impulses of these juveniles. A female that has finished copulating and GG rubbing will stop and wait when a male juvenile clings to her behind; the female will grasp the juvenile’s hips and insert his penis. Occasionally, a female will lower her hips to accommodate and make insertion easier. If so, the young male begins to thrust awkwardly but enthusiastically. At these times, other juveniles come and touch the female’s behind, and wait for their turn. The female does her best to accommodate the partners, one by one, without much ado. Once in a while, the female will sit down and ignore them, but when this happens, the juveniles start to scream. The female may then lose patience and wearily lift up her hips as if to say, “he’s just a hopeless child.”

Adult males are also enthusiastic about '‘sex education.” After copulation, they mount and thrust at juveniles, either male or female, that approach and present. Without inserting their penis, the adult males rub against the top of the hips and the thigh. Often, instead of thrusts, an adult male raises one ankle in the air and lightly scratches the side of the juvenile’s chest.

Adult males are frequently the ones who invite the juveniles. A male may take a juvenile from its mother and thrust while it is clinging to his belly, or he may approach a solitary juvenile and mount. Carried away by enthusiasm, an adult may repeat this kind of behavior several times with the same individual. As a rule, one continuous bout lasts longer than true copulation, and examples of thrusting have lasted close to two minutes.

The frequency of sexual behavior in juvenile females is low relative to juvenile males (31 cases for females vs. 227 cases for males, Table A), because unlike juvenile males, who vigorously engage in copulation with females, juvenile females hardly ever engage in GG rubbing with adult females. In contrast to males, who can have a penile erection from about six months after birth, juvenile females have extremely small external genitalia throughout the juvenile period. Only after they are about six years old and close to adolescence can females join in true GG rubbing.

Table A. Pseudocopulatory and copulatory behavior
among immature individuals (juveniles, infants)
(based on 330 observation-hours at the feeding site, 1978-79).

Position and kind
of sexual behavior

 Partner
Adult
male		 Adoles-
cent
male		 Adult
female		 Adoles-
cent
female		 Imma-
ture
male		 Imma-
ture
female		 Total
Immature males
Was mounted dorso-ventrally 32 101 0 70 50
Mounted dorso-ventrally 7* 139 27 —1 75
Ventro-ventral copulation 11 739 20 112 90
Copulation with position change 0 01 1 00 2
Copulation in undetermined position 2 05 0 00 7
Genito-genital rub 0 1** 3 00 0 4
Rump-rump contact 1 00 0 00 1
Immature females
Was mounted dorso-ventrally 7 50 0 — 0 12
Ventro-ventral copulation 6 20 0 —0 8
Copulation in undetermined position 1 00 0 —0 1
Genito-genital rub 0 03 4 —0 7
Undetermined sex
Ventro-ventral copulation 1 00 0 00 1
Total 68 2691 52 183 258

* In three cases, an adult male was mounted and thrusted against when he was in the midst
of dorso-ventral copulation (multiple-individual copulation).
** A behavior of rubbing with the penis (see Table 24).

Relative to adult sexual interactions, those of juvenile pygmy chimpanzees are rare (Table A), which differs from other hominoid species. Juvenile pygmy chimpanzees do not need to get involved in immature sexual games with juvenile peers, because they have the good fortune of receiving direct coaching from experts on sex, the adults.

“Play” begins one year after birth, and with the exception of slow-moving play with the mother (“doll play”), it is solitary play. “Locomotor play,” in which youngsters hang and clamber in the branches, is representative. After one year of age, they begin to play with other infants, but basically as an extension of locomotor play in which peers are together.

During the juvenile period, individuals enter into ordinary “social play,” consisting of chasing each other from branch to branch and grappling and wrestling. They do this tirelessly, generally during the adult rest period. In one game, they jump into the nest, wrestle for a while, and leave. Then, after running around, they jump into the nest again. Longterm play occurs in the trees, but play on the ground rarely lasts long.

Juveniles often use small branches in play. One individual will pick up a small branch and run away, signaling “try and get it.” With a playmate in pursuit, he dashes around, passes the branch in a baffling way from hand to foot to mouth, and puts it between his thigh and abdomen. Finally, when the branch is snatched from him by force, he changes roles to chase his playmate. I think that the origins of tag and all ball games are very old. Although adolescent females join in these chase games, multi-parous females do not engage in play, except for playing with their own infant. Small branches are not the only “tools” of play; sugarcane also is often used.

When adolescent and adult males join in play, the game is almost always play-wrestling. Play-wrestling occurs not only between juveniles, but also between adolescent males and between adolescent and adult males. As expected, play between adult males is rare. When a senior and a junior are engaged in play-wrestling, the junior is invariably on the aggressive side. His elder squats on a branch or rolls over on his back; then, junior attacks a favorite spot by “play-biting,” grabbing and biting a hand or a foot. Whenever we hear the “play-pant,” a quick “hat-hat-hat,” we know that somewhere someone is play-wrestling.

Juveniles often engage in “food-begging” from individuals other than their mother. The success rate, however, seems to be poor, and the juveniles often give up after a short time.


Figure D. An adolescent male peers at an infant. Photo by Takayoshi Kano.

An adult male may embrace an unrelated infant or carry it on his back. To do this, he approaches a female carrying an infant and stares at the infant while extending his hand. The infant may jump on his back, and if the infant clings to his stomach, he will quickly leave the mother. The mother will tolerate a separation of about 5 m, but if they go farther away or if the infant begins to whimper, the mother will hurry to take it back. No cases of females engaging in this kind of motherly behavior have been reported.

The behavior of elders toward juveniles (2 to 4 years old) may be summed up in a single word, tolerance. There are no reports of severe scolding or violent attacks directed at juveniles. However, aggressive behavior toward juveniles does occur often as they approach adolescence. Elders begin to chase the older juveniles, to take their sugarcane, and to threaten them. In serious cases, they even hold them down and bite them. Because their mothers rarely come to their aid, males and females approaching adolescence begin to spend more time alone on the periphery.

Adolescence (7 to 14 years old) may correspond to puberty in humans, which extends from 10 to 15 or 16 years of age. It is the transitional period from childhood to adulthood. Although there are exceptions, usually a male at this time lives at the periphery of the group. He moves with his mother, but often he separates from all other members of the group. He keeps to himself, especially during times of excitement and enthusiastic feeding, because when he is in the center of the group, he will often be threatened or attacked by other males. No longer a juvenile, he is not tolerated.

Adolescent males keep company with their mothers during the rest period. They finish their meals and lie sprawled, separately scattered about. They make their nests, and when they rest in the shade, the adolescent males casually approach their mothers and often enter into grooming relationships. Almost all of them are inferior to any adult male; they rarely mount and rump-rub with adult males and are still not recognized as “fully adult.”

Among adolescents, dominant-subordinate rank is apparently not yet established. Even in postadolescent males, attacker-victim relationships are not fixed. Also, aggressive interactions among adolescents are not violent and frequently transform to play. An adolescent male may branch-drag, but he does this infrequently compared to an adult male.

Already mentioned in passing, “peering” behavior, a common behavior in adolescents regardless of sex, will be formally introduced here. When peering, a pygmy chimpanzee sits or stands quadrupedally and gazes intently, from the front or side, into another’s face. The actor gets close enough to almost make face-to-face contact and, with a calm facial expression, stares at the other. This stare usually lasts one to two minutes, but one case lasted five minutes.


Figure E. A juvenile peers at an adult male. Photo by Takayoshi Kano.

Peering is typically a behavior that juniors direct toward their age-mates or seniors. The one being stared at usually pretends not to notice, but some kind of social interaction occurs when they are in each other’s line of vision. The effect differs according to age, sex, and rank, but the typical behaviors that are generated include copulation, GG rubbing, play, and food sharing. Peering rarely leads to aggressive interaction, from which we deduce that it is not threatening.

A mother allows her infants and juveniles to come close to her face. While they stare, '‘food-begging” often occurs, and peering may be an extension of this behavior. Peering, however, is not just food-begging, but instead it seems to be begging for any kind of friendly social contact. The frequency of peering behavior increases with age, adolescence being the peak period. It is hardly ever seen in males after middle-age, but even after reaching old age, females use peering to solicit GG rubbing.

Adolescent females are unique beings within the group. As immigrants from other troops, they are newcomers from the point of view of other members. Their behavior is always reserved; they keep their distance from all disputes; and they seldom emit loud vocalizations, even when exchanging calls with another party. They rarely resort to aggressive behavior of their own accord, and if attacked, they just scream and flee without counterattacking.
 
These young nulliparous females are not completely socially inactive, however. They respond to the behavior of males. Because they quickly approach and present when they perceive even a slight sign of courtship display, copulatory frequency is highest in these adolescent females. Also, they eagerly approach older multiparous females to ‘‘peer” and to solicit GG rubbing. After engaging in these kinds of behaviors for a while, the nulliparous females carry their food and retreat to a safe feeding spot in a tree on the outskirts of the feeding site. During the rest period, however, they re-appear, approach the older females, and begin to groom. These adolescent females spend much more time grooming than being groomed. During the rest period, they ‘‘play-chase” with the juveniles for a long time.


Figure F. Teaching good manners.

Thus, the young nulliparous adolescent females avoid aggressive contact on the one hand and initiate friendly contact on the other. They seem to be working to stabilize their own position in the group by establishing friendly social relationships, rather than by building dominant-subordinate relationships. They seem to know well that they have no supporters or guardians in the group. This reminds me of a new Japanese bride who bravely enters the groom’s large household. If she marries the oldest son, she may also have to live with her husband’s parents and the unmarried brothers and sisters of her husband. Because she is picked on by in-laws, and her own family is distant, the new bride is often under great stress. She must be very patient.

Female adolescence may be roughly divided into three stages. In the first stage (7-8 years old), the female has just left her natal group. Her sexual organs are small, unlikely to be penetrated by an adult male penis, and her body also is small. Nevertheless, she actively engages in sexual and social behavior. Sometimes, however, a little childishness remains, and she may become absorbed in playing by herself for a long time. When she enters the middle stage (9-12 years old), the sexual organs increase in size by swelling. During this period, copulation and GG rubbing occur most frequently. The last stage (13-14 years old) lasts about nine months, from conception to birth of the first offspring. Copulation occurs up until about one month before delivery. The sexual swelling gradually slackens up to the time of delivery, at which time the female becomes socially and sexually inactive.

We conclude from the data obtained so far that the inclination toward a specific unit group deepens during the first half of adolescence. By the time a female is 12 or 13 years old, her position within the group seems to be almost settled.
   (Kano)
DISCUSSION
Chimpanzees and bonobos are particularly relevant to our understanding of human social evolution since humans are estimated to have diverged from these two species a “mere” 8 million years ago, compared to 30 million years, or more, for the split between humans and most other primates (Sibley and Ahlquist, 1984; Marks, 1992; Goodman, 1992).

It is rather surprising how much humankind’s two closest relatives differ with regard to the use of sociosexual behavior. From my extensive observations of chimpanzees, I believe that most sexual activity in this species either serves reproduction directly or serves as a preparation for reproductive sex. Thus, juvenile male chimpanzees are attracted to females in estrus, and pubertal females begin to explore sexual contact with older males as soon as their genital swellings develop (van der Weel, 1978; de Waal, 1982; Goodall, 1986). Such intergenerational experiences undoubtedly contribute to the development of adequate sexual skills.

The incorporation of sexual elements into reassurance behavior is rather limited in chimpanzees. Like almost all primates, they present their anogenital region in a gesture of submission or appeasement, often followed by a genital inspection by the recipient. Furthermore, chimpanzees show mouth-to-mouth kissing, and adult males of the species may mount one another during or following aggressive incidents (de Waal and van Hooff, 1981). These contacts lack the intensity and arousal obvious during kissing and isosexual mounts among bonobos, however. Thus, neither GG-rubbing between females nor tongue-kissing have ever been reported for the chimpanzee. All in all, bonobos seem to do with a variety of sociosexual behavior patterns what chimpanzees do with embracing and relatively “platonic” kissing. It is a difference in degree, yet an important one.

This difference cannot be explained as a product of the conditions at the San Diego Zoo. Although a few of my observations have not been reported by others (such as the use of mounting postures during reconciliation), this omission probably is due to the detail in which captive animals can be studied, rather than to fundamental differences with wild conspecifics. The Japanese team of field workers in particular has reported a remarkable variety of socio-sexual behavior patterns in wild bonobos; their largely qualitative descriptions and interpretations are consistent with my quantitative data. Furthermore, that attractive food induces sexual activity rather than competition, has been observed not only in zoos or at the Japanese field site, where humans provide food, but also at a site at which food provisioning is totally absent. In Lomako Forest, also in Zaire, Thompson-Handier et al. (1984) saw wild bonobos engage in sex when they entered trees loaded with ripe figs, or when one amongst them had captured a prey animal, such as a small forest duiker. The flurry of sexual contacts would last for five to ten minutes, after which the apes would settle down to consume the food.

Occasionally, the role of sex in relation to food is taken one step further, a step that brings bonobos very close to humans indeed. It has been speculated by anthropologists, such as Lovejoy (1981) and Fisher (1983), that the reason for the partial separation between sex and reproduction in our species is that sex serves to cement mutually profitable relationships between men and women. The human female’s capacity to mate throughout her cycle, and her strong sex drive, make it possible to exchange sex for male commitment and paternal care, thus tying men and women together in the nuclear family. Instead of a conscious strategy, this is of course thought of as an arrangement favored by natural selection for the simple reason that it allows women to raise more offspring than they would be able to raise on their own. Bonobo behavior seems to fit important elements of this model: female bonobos show extended receptivity and use sex to obtain male favors. Thus, a female who does not dominate a particular male still has sex as a source of leverage.

At the San Diego Zoo, I observed that if Loretta was in a sexually attractive state, she would not hesitate to approach the adult male, Vernon, if he had food. She would present herself to Vernon, mate with him, and make high-pitched food calls while taking over his entire bundle of branches and leaves. She would hardly give the male a chance to pull out a branch for himself, sometimes grabbing the food out of his hands in the midst of intercourse. This was quite a contrast with periods in which Loretta had no genital swelling; then she would wait until Vernon was ready to share (de Waal, 1989). Kuroda (1984, 317) reports similar exchanges at Wamba: “A young female approached a male, who was eating sugar cane. They copulated in short order, whereupon she took one of the two canes held by him and left.” In another case, “a young female persistently presented to a male possessor, who ignored her at first, but then copulated with her and shared his sugar cane.”

It is no wonder that Mori (1984, p. 277) concluded on the basis of his observations of the bonobos at Wamba: “Coexistence of plural males and females without agonistic competition in mating could be guaranteed by changing the character of sexual behavior into affiliative behavior in which all individuals can participate, and by decreasing the reproductive meaning.”

Previously, I have argued that the most logical pathway via which sexual behavior evolved into a general reassurance and appeasement mechanism in the bonobo is that this mechanism was first established in the adult male-female relationship (the original functional context of all sexual behavior), after which it was adopted by other age and sex combinations. In other words, the widespread application of sexual behavior in the bonobo’s social life has its origin in an early emphasis on heterosexual bonding (de Waal, 1987). Field research supports this view in that male-female relationships seem closer and more tolerant in bonobos than in chimpanzees. Chimpanzee foraging parties typically consist either of adult males (sometimes accompanied by females in estrus) or mother-offspring units, whereas bonobos tend toward larger, mixed-sex parties (Badrian and Badrian, 1984; Badrian and Malenky, 1984; Kano and Mulavwa, 1984; Wrangham, 1986). This possible evolutionary background of bonobos’ sexuality is relevant to human sexuality, both because of the shared biological ancestry and because of the characteristic heterosexual bonding of the human species.

Humans go further in the latter respect than bonobos, however. Despite the quid pro quo between the sexes, there are no indications that bonobos form humanlike nuclear families with permanent bonds between a male and one or several females; in the bonobo, the burden of raising offspring appears to rest entirely on the female’s shoulders. In fact, nuclear families are probably incompatible with the widespread use of sex found in bonobos. If we speculate that our ancestors started out with a sex life similar to that of bonobos, the evolution of the family must have brought dramatic change.* The integrity of this social unit requires all sorts of protective moral restrictions and taboos that, among other things, assure males that they are caring for their own, not someone else’s, offspring. Thus, although our species is characterized by an extraordinary interest in sex, there are no human societies in which people publicly engage in it with the “freedom” characteristic of bonobos. A sense of shame and a desire for domestic privacy, which have no place in bonobo society, are typical human concepts related to the evolution and cultural bolstering of the family unit.
      * During the conference from which this volume resulted, some participants indulged in speculation along the lines that “there is a bonobo in all of us.” This may well be true, but we should remember that this ape is not on its way to becoming human any more than humans are on their way to becoming apelike. We share a common ancestor, i.e., are separate branches on the same evolutionary tree. If hominids did indeed know the pansexuality of the bonobo, this would suggest that our common ancestor showed such behavior as well. If so, we would need to explain why it disappeared in the chimpanzee (which shares the same ancestor) and why in modem humans male-female relations became the basis of exclusive reproductive units within the social system.

Sex does probably serve peacemaking and bonding functions within the human family, however, as when a couple seals a reconciliation by making love. In addition, it is obvious that no degree of moralizing can make sex disappear from every nonfamily realm of human society. The bonobo’s behavioral peculiarities may help us understand why that is, and as such have serious implications for models of human evolution.

Acknowledgments
The bonobo study was made possible by the San Diego Zoological Society and a grant from the National Geographic Society. The study and writing of this paper were further supported by National Institutes of Health grants to the Wisconsin Regional Primate Research Center (RR-00167) and the Yerkes Regional Primate Research Center (RR-00165), the latter also supporting the collection of comparative data on chimpanzees at the Yerkes Field Station. The author thanks Katherine Offutt for assistance with data analysis, Frank Kieman for printing the photographs, and the conference participants for discussion of an earlier version of the manuscript.

REFERENCES
   Badrian, A., and Badrian, N. 1984. Social organization of Pan paniscus in the Lomako Forest, Zaire. In R. Susman (Ed.), The Pygmy Chimpanzee. New York: Plenum Press, pp. 325-46.
   Badrian, N., and Malenky, R. 1984. Feeding ecology of Pan paniscus in the Lomako Forest, Zaire. In R. Susman (Ed.), The Pygmy Chimpanzee. New York: Plenum Press, pp. 275-99.
   Becker, C. 1984. Orang-Utans und Bonobos im Spiel. Munich: Profil Verlag.
   Blount, B. G. 1990. Issues in bonobo (Pan paniscus) sexual behavior. American Anthropologist 92: 702-14.
   Coolidge, H. J. 1984. Historical remarks bearing on the discovery of Pan paniscus. In R. Susman (Ed.), The Pygmy Chimpanzee. New York: Plenum Press, pp. ix-xiii.
   Dahl, J. F. 1985. The external genitaha of female pygmy chimpanzees. Anatomical Record 211.24-2^.
   Dahl, J. F. 1986. Cyclic perineal swelling during the intermenstrual intervals of captive female pygmy chimpanzees (Pan paniscus). Journal of Human Evolution 15: 369-85.
   Dahl, J. F. 1987. Sexual initiation in captive group of pygmy chimpanzees (Pan paniscus). Primate Report 16: 43-53.
   Dahl, J. E, Nadler, R. D., and Collins, D. C. 1991. Monitoring the ovarian cycles of Pan troglodytes and P. paniscus: A comparative approach. American Journal of Primatology 24: 195-209.
   Fisher, H. 1983. The Sex Contract: The Evolution of Human Behavior. New York: Quill.
   Furuichi, T. 1987. Sexual swellings, receptivity, and grouping of wild pygmy chimpanzee females at Wamba, Zaire. Primates 28: 309-18.
   Furuichi, T. 1992. The prolonged estrus of females and factors influencing mating in a wild group of bonobos (Pan paniscus) in Wamba, Zaire. In Topics in Primatology: Vol. 2, Behavior, Ecology, and Conservation. Itoigawa, N., Sugiyama, Y, Sackett, G. P., and Thompson, R. K. R. (Eds.). Tokyo: University of Tokyo Press, pp. 179-90.
   Goldfoot, D. A., Westerborg-van Loon, H., Groeneveld, W, and Slob, A. K. 1980. Behavioral and physiological evidence of sexual climax in the female stump-tailed macaque (Macaca arctoides). Science 208: 1477-79.
   Defoe, Gideon. How Animals Have Sex. London: Weidenfeld & Nicolson, 2005.
   Goodall, J. 1986. The Chimpanzees of Gombe: Patterns of Behavior. Cambridge: The Belknap Press of Harvard University Press.
   Goodman, M. 1992. Hominoid evolution at the DNA level and the position of humans in a phylogenetic classification. In Topics in Primatology: Vol. 1, Human Origins. Nishida, T, W. C. McGrew, R Marler, M. Pickford, and F. B. M. de Waal (Eds.). Tokyo: University of Tokyo Press, pp. 331-46.
   Hockett, C. E, and Archer, R. 1963. The human revolution. Current Anthropology 5: 135-68.
   Hübsch, I. 1970. Einiges zum Verhalten der Zwergschimpansen (Pan paniscus) und der Schimpanzen (Pan troglodytes) im Frankfurter Zoo. Zoologische Garten 38: 107-32.
   Jordan, C. 1977. Das Verhalten Zoolebender Zwergschimpansen. Ph.D. dissertation, Goethe University, Frankfurt.
   Kano, T. 1980. Social behavior of wild pygmy chimpanzees (Pan paniscus) of Wamba: A preliminary report. Journal of Human Evolution 9: 243-60.
   Kano, T. 1989. The sexual behavior of pygmy chimpanzees. In P. G. Heltne and L. A. Marquardt (Eds.), Understanding Chimpanzees. Cambridge: Harvard University Press, pp. 176-83.
   Kano, T. 1992. The Last Ape: Pygmy Chimpanzee Behavior and Ecology. Stanford: Stanford University Press.
   Kano, T., and Mulavwa, M. 1984. Feeding ecology of the pygmy chimpanzees (Pan paniscus) of Wamba. In R. Susman (Ed.), The Pygmy Chimpanzee. New York: Plenum Press, pp. 233-74.
   Kirchshofer, R. 1962. Beobachtungen bei der Geburt eines Zwergschimpansen (Pan paniscus, Schwarz 1929) und einige Bemerkungen zum Paarungsverhalten. Zeitschrift für Tierpsychologie 19: 597-606.
   Kuroda, S. 1980. Social behavior of the pygmy chimpanzees. Primates 21: 181-97.
   Kuroda, S. 1984. Interaction over food among pygmy chimpanzees. In R. Susman (Ed.), The Pygmy Chimpanzee. New York: Plenum Press, pp. 301-24.
   Lovejoy, C. O. 1981. The origin of man. Science 211: 341-50.
   Marks, J. 1992. Genetic relationships among the apes and humans. Current Opinion in Genetics and Development 2: 883-89.
   McKenna J.J. 1983. Primate Aggression and Evolution: An overview of sociobiological and anthropological perspectives. Bulleting of the American Academy of Psychiatry and Law 11: 105-130.
   McGinnis, P. 1973. Patterns of Sexual Behavior in a Community of Free-living Chimpanzees. Ph.D. dissertation, Cambridge University.
   Mori, A. 1984. An ethological study of pygmy chimpanzees in Wamba, Zaire: A comparison with chimpanzees. Primates 25: 255-78.
   Morris, D. 1967. The Naked Ape. London: Jonathan Cape.
   Nieuwenhuijsen, K. 1985. Geslachtshormonen en Gedrag bij de Beermakaak (Macaca arctoides). Ph.D. Dissertation, Erasmus University, Rotterdam.
   Rempe, U. 1961. Einige beobachtungen an Bonobos (Pan paniscus, Schwarz, 1929). Zeitschrift fur Wissenschaftliche Zoologie 165: 81-87.
   Savage, S., and Bakeman, R. 1978. Sexual morphology and behavior in Pan paniscus. Proceedings of the 6th International Congress of Primatology. New York: Academic Press, pp. 613-16.
   Savage-Rumbaugh, S., and Wilkerson, B. 1978. Socio-sexual behavior in Pan paniscus and Pan troglodytes: A comparative study. Journal of Human Evolution 7: 327-44.
   Sibley, C., and Alhquist, J. 1984. The phylogeny of the Hominoid primates, as indicated by DNA-DNA hybridization. Journal of Molecular Evolution 20: 2-15.
   Thompson-Handier, N., Malenky, R., and Badrian, N. 1984. Sexual behavior of Pan paniscus under natural conditions in the Lomako Forest, Equateur, Zaire. R. Susman, (Ed.). The Pygmy Chimpanzee. New York: Plenum, pp. 347-68.
   Tratz, E., and Heck, H. 1954. Der afrikanische Anthropoide “Bonobo,” eine neue Menschenaffengattung. Sdugetierkundige Mitteilungen 2: 97-101.
   de Waal, R B. M. 1982. Chimpanzee Politics. London: Jonathan Cape.
   de Waal, E B. M. 1987. Tension regulation and nonreproductive functions of sex among captive bonobos (Pan paniscus). National Geographic Research 3: 318-35.
   de Waal, E B. M. 1988. The communicative repertoire of captive bonobos (Pan paniscus), compared to that of chimpanzees. Behaviour 106: 183-251.
   de Waal, F. B. M. 1989. Peacemaking among Primates. Cambridge: Harvard University Press.
   de Waal, F. B. M. 1990. Sociosexual behavior used for tension regulation in all age and sex combinations among bonobos. In Pedophilia: Biosocial Dimensions, J. R. Feierman (Ed.). New York: Springer, pp. 378-93.
   de Waal, F. B. M. 1992. Appeasement, celebration, and food sharing in the two Pan species. In Topics in Primatology: Vol. /, Human Origins. T. Nishida, W. C. McGrew, P. Marler, M. Pickford and F. B. M. de Waal (Eds.). Tokyo: University of Tokyo Press, pp. 37-50.
   de Waal, F. B. M., and Ren, R. 1988. Comparison of the reconciliation behavior of stumptail and rhesus macaques. Ethology 78: 129-42.
   de Waal, F. B. M., and van Hooff, J. A. R. A. M. 1981. Side-directed communication and agonistic interactions in chimpanzees. Behaviour 77: 164-98.
   de Waal, F. B. M., and van Roosmalen, A. 1979. Reconciliation and consolation among chimpanzees. Behavioral Ecology and Sociobiology 5: 55-66. van der Weel, M. 1978. Sexuele interacties en relaties tussen chimpansees. Unpublished Research Report, University of Utrecht.
   Wescott, R. W. 1968 (1963). In A. Montagu (Ed.). Culture: Man's Adaptive Dimension.
   Oxford: Oxford University Press, pp. 91-93.
   Wickler, W. 1969 [1967]. Socio-sexual signals and their intra-specific imitation among primates. In D. Morris (Ed.). Primate Ethology. Garden City: Anchor Books, pp. 89-189.
   Wrangham, R. 1986. Ecology and social relationships in two species of chimpanzee. In D. Rubenstein and R. Wrangham (Eds.), Ecological Aspects of Social Evolution: Birds and Mammals. Princeton: Princeton University Press, pp. 352-78.
   Yamagiwa, J. 1987. Intra- and inter-group interactions in an all-male group of Virunga mountain gorillas (Gorilla gorilla beringei). Primates 28: 1-30.