Go to Lecture One or Three or Four or Five or Six or Seven or Eight

Go to Cyber Semiotic Institute home page
Go to Course Outline


Human Communication as a Primate Heritage

Instructor: Anne Zeller

Lecture Two: Codes in Communication

In lecture one we discussed the various modalities and channels of communication used by primates. These modalities and channels are also used by humans in a wide variety of ways. The emphasis I placed on Edward Hall's work in the first lecture was intended to alert you to the cultural differences in interpretation of signals that humans exchange in these ways. These cultural differences are fundamental in organizing the message patterns that humans use to communicate with each other. For example, in some societies looking another person straight in the eyes while you talk with them is considered a way to express sincerity and frankness. The sender is a person to be believed, an honest person with nothing to hide. In other cultures, this would be a very aggressive thing to do. You would be challenging the receiver not to believe you, and to look out for the consequences if they did not. There are also gender related differences in the appropriateness of this behaviour. In current western culture women and men strive to act in the same way in board rooms, which means meeting peoples' gaze squarely, and not appearing to be evasive. In other situations, such as a bar or entertainment spot, the initiation of such a forthright gaze by a woman to a man might be interpreted as a sexual signal. In other cultures, men are not culturally permitted to gaze directly at women's faces even when doing business with them, and women, even when in positions of economic influence do not engage the direct gaze of men. In some situations face to face contact between unmarried men and women is culturally forbidden, whether or not it actually occurs in private. This is only one aspect of the visual code. Body posture, and movement patterns are other subtle gestures which convey culturally explicit information to humans which are also common modalities with primates. In addition, humans have patterns or codes of clothing and jewellery which are appropriate for particular activities and are also markers of social class. Sumptuary laws and church rules in the Middle Ages were very clear about who could wear what kind of material (lace, velvet, cloth of gold, etc.) and who could not. These were markers of social class which were clearly understood by all members of society, even if they were not always obeyed. This type of coding is only an extension of the kind of coding that is found in the visual and vocal modalities of primates.

Basic communication modalities in primates (messages which are not under the control of the individual sender) still present information in a coded format, or else it would be very difficult for primates to learn the system. For example, facial colouration patterns in guenons are quite complex. Distinct patterns of red, brown, black, white and blue spots, stripes and eye rings decorate the faces of the12 to 15 gueneon species. These are complemented by crests of hair on the top of the head, the eyebrow region, the 'side burns' area or a beard like growth of hair on the chin. Since these species of small mainly arboreal monkeys are closely related and frequently share the same trees, although eating different diets, and utilizing different levels, they have plenty of opportunity to meet one another. Since sexually mature animals, particularly males, frequently transfer from their group to improve their social rank and mating prospects, it is necessary for them to be able to recognize conspecifics when looking for a new group. Females also must be able to recognize the appropriate species of male if they are in estrus and sexually available. Therefore, the recognition of hair pattern and face colouration is quite important, especially since mixed species troops do exist. Another situation of potential misunderstanding is when a stranger lone animal begins to associate with a troop for its own safety since living alone as a primate is a very risky business.

Within group communication between females and males about sexual readiness is also important and can occur at a basic level. This does not mean that the appropriate response is innate. I have seen a female baboon with a huge estrus swelling presenting repeatedly (standing with her perineum directed at the male's face and looking over her shoulder at him) to a young male who was obviously sexually aroused because he was masturbating his erect penis, but he wasn't even looking at the female. Males must learn what the change in colour and size of a females's perineum means, if the appropriate transfer of information is to occur.

Face colours can also be indicators of sexual receptivity or arousal in both genders. Among Japanese macaques who have rather hairless faces the female's face colour becomes bright red when she is in estrus. Male mandrills who have brightly coloured red and blue faces become positively incandescent when consorting with a breeding female. These markers of sexual readiness and receptivity clearly transmit information not only to intended partner, but to the other members of the group.

It is not only the face which can be affected in the transmission of basic sexual signal. As mentioned above the perineal region (around the anus and vulva) of mature female monkeys often swells and changes colour when she is sexually receptive. This is an important visual cue to primates who may live in large groups or in a forest environment where visibility is somewhat occluded. Chimpanzee males have been observed to spot a flash of bright pink sexual swelling on the other side of a valley and hurry over to investigate. In species such as baboons and some macaques, males form longer term (1 to 3 or 4 days) consort relationships with females and remain in close proximity while grooming and mating. In many cases younger or lower ranking males will consort with females when their swellings are not at maximum size (or when they are young and the swelling is small) but a female with a large estrus swelling is usually the recipient of a lot of sexual interest by males. This is because the swelling usually begins several days before ovulation, and increases until ovulation occurs at which point it begins to reduce. Thus mating with a female at her maximum swelling size will most probably result in conception.

There are additional visual indications of sexual receptivity which change in some primates. These include swelling of the scrotum in males of seasonally breeding species, enlarging of the pink fleshy vesicles on the chests of female geladas and the behaviour of animals who are ready to mate. When discussing behaviour, we move from basic level communicative markers to the interactive level where animals have some control over the messages that they send. In these cases, the decisions to approach or not to approach, to respond or not to respond can be communicative or can reflect the arousal level of the animals. Some individuals simply will not mate with each other and this may indicate close relationship such as brother/sister or mother/son. In some cases one is willing and the other is not. This can lead to conflict in which one animal tries to flee from the other, vocalizes, tries to sit or lie down or refuses to stand in the correct mating posture. This can occur for both males and females, and clearly indicates unwillingness to all conspecific in the vicinity. Some females have been seen trying to pull exhausted males to their feet and get them to mount, especially in a seasonal breeding situation in which all the females are receptive at the same time. Other visual signals include a set of what are called proceptive behaviours by females which are postures and movement to encourage the males to mate. These include presenting the hind quarters, head flagging (shaking the head back and forth) slapping at the male, grooming him and manual stimulation. In addition, some females use particular vocalizations to indicate mating readiness. Males have their own sets of proceptive behaviour as well which vary from species to species. These include genital inspection of females either by sniffing the perineum or touching it with their hands, approaching the females to groom (especially if this is not a common male/female activity as in orangutan), consorting with a female (ie. following her around) touching her hip to indicate that she should stand in a mating posture, and 'bird dogging' which is a posture of Japanese macaque males which looks very much like a well trained hunting dog indicating the presence of a pheasant. In addition are the more overt signals of 'snag crashing' (pushing over trees) dropping branches and loud vocalizations of orangutan males who must attract females from distant locations because they live in fairly solitary lifestyle, and the herding behaviours of chimpanzee males attempting to sequester females during estrus so other males will not find them (termed 'going on safari' in the chimp literature, or 'safari behaviour'). Most monkeys living in groups do not need to use such loud attention getting behaviour, but frequently must be persistent in their signal transmission to get a successful response to their message.

Visual Gestures

A number of the signals mentioned above can be considered visual displays. Displays are highly coded behaviours which are defined by their replicable form and the type of stimulus causing their occurance. Threat displays, mating displays and appeasement displays are only a few of the sets of behaviour which bring together various modalities such as body movement, gesture colour and vocalization to send one coherent message. As Jolly (1985) defines the term, displays are made up of ritualized behaviour. "Ritualized behaviour can be recognized by just those properties that resemble ritual. First, it is exaggerated. Second, it is stereotyped and stylized. A gesture is less ambiguous if it is always performed in a single form, at a single level of intensity (Morris 1957). Third, it may often be repeated" (Jolly 1985:196). Therefore parts of many displays evolve out of attention movements; movements which previously had a particular non-communicative function which has been abbreviated to a short form in which it indicates the motivation underlying the message. For example, the hand slap on the ground and head bob forward seen in a threat display are usually classified as intention movements - indicators of a possible chase. In the same way, the lips forward open mouth is often seen as 'iconic' for a bite. None-the-less, a whole set of face and body movements and vocalization can go together in a repetitive stylized form which indicates 'threat display'. For a long time it was considered that this type of display indicated an aggressive motivation but more recent work suggests that this may be simplification, and that at least some displays may be emancipated from an emotional tie. Animals may threat in aggression, or fear or bluff or in situations requiring social control. Charging displays by male gorillas are a frequently used example of a display using up to 9 components including false feeding (where the male bites a leaf off as he goes by) standing on the hind legs, open mouth, eyes staring, chest beating, bipedal run towards, kick out to the side, head shake and vocalization. Not all of these must be included for the activity to be considered a charging display but the activity itself is quite unmistakable and serves as a waring to the intruder to back off. Very seldom does a bluff display like this lead to attack unless the response is inappropriate. These displays can be repeated several times but the intensity is changed by adding more components, rather than by increasing the volume of the calls, or the length of the bipedal run.

In spite of their impressive nature, displays are not the most common level of visual communicative interaction by most primates. Usually communicative episodes include more low key glances, approaches and shifts in body posture. None the less, these can also be quite complex given the level of differentiation possible in parts of the body and degree of movement. My original work on macaque communication (Taub and King 1986, expanded in Fa and Lindburg 1996) was an effort to describe the movements of the face during threat interactions in Macaca sylvanus, the barbary macaque of Gibraltar. A generalized list of movements of macaque faces had been drawn up before this, but the list was not specific to a particular species and did not consider the possibility of individual variability or variability based on differences due to age of the animal, sex or kin group. Instead, such variability was regarded as noise in a cybernetic system of information transfer. I argued that if there was regularity in this variability then it was not noise, but information added to the message at a metacommunicative level. Briefly stated, my results indicated that Macaca sylvanus used a total of 32 of 34 particular components when making threats. These components were defined as part of the face coupled with a movement, such that eyebrows raise and eyebrows lower were two components. Of these 32 components each individual used between 9 and 20 depending on the individual and the situation. The most interesting discovery was that no components were used in all episodes of sending a threat, which indicates that there is no irreducable core of components that means 'threat'. This discovery has been confirmed by expanding the data base to include Macaca fasicularis (long tail macaques) and Macaca fuscata (Japanese macaques). Only one component in M.fuscata was seen in 100% of the cases. Otherwise, although every individual used the four main components at some time, the patterning of the gesture indicated considerable variability. I divided the components into 3 categories - constant components, variable components and unique components. The idea was to distinguish between those components used by all animals, those used by some and those used only by one animal. However, I found that only one animal used a unique component, so it was not possible to differentiate individuals on the basis of a signature component. When I sorted the components on the basis of use frequency with constant components showing regular use by all member of a species, I discovered that there were only 6 such components in the 3 species examined and only two of these overlapped. In other words, the basic components of threat differ noticeably between barbary, Japanese, and long tail macaques. Not only that, but in barbary macaques all of the constant components were involved with the upper face or piloerection, while both the long tail macaque ones were mouth related components. This indicates that different parts of the face carry different aspects of the message in the two species. When I began to divide the animals in terms of sex and age, I found that some components were used consistently by males which differentiated them from females, and some by females, which differentiated them from males. This occurred in all three species. In Barbary and Japanese macaques the larger proportion of these consistently used components were from the non-mouth region thus reinforcing the pattern of less variable components being non-mouth and more variable components being from the mouth region. This is not, however, the pattern seen in longtail macaques. The same distinction occurred when the animals were divided into old adult, young adult, and subadult categories. The barbary and Japanese macaques showed consistant use of non-mouth discriminators, while long tails discriminated age sets by frequent use of mouth components. The actual components involved are discussed in Chapter 24 in Fa and Lindberg's 1996 'Evolution and Ecology of Macaque Societies." The importance of this level of description and differential is the discovery that the information coding system for threat not only sends ' the threat message' but also operates at a meta-communicative level to send information about the social class (age and sex) of the sender. Even more informative is the fact that in Barbary macaques (which is the only species which I had the data to test for kin group) the three matrilineal kin lines could be distinguished by differential component use, and the one non-kin male distinguished from them. This level of social distinction not only suggests the importance of learning the correct code for your species, but for the social group within the species. The level of information coded included the main message of threat, and the meta level messages of species, age, sex and kin group. This is a very complex coding system and one which was not envisioned when primate communication studies began in detail some forty years ago.

Vocal Communication

Understanding of the complexity of vocal communication coding has also mushroomed in the last 30 years since high quality portable sound recording devices became practicable for use in the field. Up until that time lab work had concentrated on gathering a repertoire of sounds from each species in an effort to document the range of call types and discover the stimuli which elicited them. In 1967 Stuart Altman edited a book "Social Communication among Primates" which detailed experiments concerning neural control of vocalizations, and research attempting to assess the level of innateness in primate vocalizations by use of isolation experiments. In addition, was a paper by T. Struhsaker who wrote that free ranging vervet monkeys had different types of calls which elicited different responses from group members when they were approached by snakes, mammalian predators (leopards) or avian predators. This idea suggested a symbolic or referential aspect to primate communication which was quite different than the paradigms accepted up to that point. Later, other researchers published on differential predator alarms in brown lemurs, sifakas, ring-tail lemurs and some new world monkeys. In 1975 Green published material on 'coos' of Japanese macaques which indicated that although the vocalizations sounded the same to unaided human ears, they were distinguishable by the monkeys, and that this is based on whether they peak early or late on the spectrographic readout. This is not a graded phenomenon, but one which is responded to as a distinct difference, in the same way that humans distinguish between the consonants 'p' and 't'. When the computer graduated coo calls are played back to rhesus macaques and stumptail macaques, however, they are not better at distinguishing early and late peak 'coos' than humans are. Later research has distinguished seven types of 'coos' in Japanese macaques, all used in different situations and discriminably fine level sound spectrographic variability. In 1984 the same type of discrimination was discovered in vervet monkey grunts by Seyforth and Cheney. They found four situationally distinctive vervet grunts which were indistinguishable to human ears but would elicit different responses when played back to monkeys from a hidden speaker. They also began a series of playback experiments to test Struhsaker's findings of differential predator calls in vervets and confirmed them. Not only did they observe monkeys responding appropriately to the different calls in the presence of predators, but also during playback experiments, which controlled somewhat for the level of fear and location of danger which it had been suggested, might have influenced either the vocalization differences or the response. After a 'snake chutter' the monkeys group together and look around on the ground for the large pythons that prey on them. After a leopard alarm they all run up the trees and out to the ends of small branches, and after an avian alarm they run in close to the tree trunk or down to the ground. Thus, these are mutually exclusive behaviours which would be dangerous if the animal responded incorrectly since that would put them in very vulnerable positions. Cheney and Seyforth continued their vervet research on vervets from 1977 until 1990. During that time they also conducted playback experiments which indicted that infant vocalizations were recognized as being associated with the mothers of those infants by other females. Female vervets would respond to aggression towards their offspring or sisters by attacking offspring or sisters of the aggressive female at a later date. In other words, their vocal communicative system not only coded referentially for various predators, differentially for various grunts, individually for infants of particular females, but was used to refer to an understanding of an individual's social relationship. The entire background of how all this information is coded is not yet clear, but considerable work is being expended on individual, kin group and situational recognition in a wide variety of animals from Mouse lemurs (Zimmerman), Marmosets (Snowdon), and vervets (Seyforth and Cheney) to Japanese macaques (Sugiura and Masataka in Zimmerman et al 1995).

I have only begun to touch on the levels of complexity in coding of visual and vocal signals in primates in this lecture to make it clear that these are not simplistic one or two variable systems, but extremely complex, information rich, multi layered systems of information transmission. Human non-verbal coding may not actually be as complex in terms of layering, but certainly provides a very rich context for our pervasive verbal system.

References for Lecture #2

Altmann, S. 1967. Social Communication among Primates. University of Chicago Press. Chicago.

Fa, J. and Lindburg D. 1996. Evolution and Ecology of Macaque Societies. Cambridge University Press Cambridge.

Jolly, A. 1985. The Evolutionary Primate Behaviour. 2nd ed. Macmillan Publishing Co. London.

Taub, D. And King, F. 1986. Current Perspectives in Primate Social Dynamics. Van Nestrand Reinhold Co. New York.

Zimmerman, E., Newman, J. and Jurgens, U. 1995. Current Topics in Primate Vocal Communication Plenim Press. New York.

Readings for Lecture #3

Byrne, A. and Whiten, A. 1988. Machiavellian Intelligence. Clarendon. Oxford.

King, B.J. 1996. The Information Continuum. SAR Press. New Mexico.


copyright 1999 Anne Zeller
Send comments or question to Anne Zeller: azeller@artspas.watstar.uwaterloo.ca
Go to Lecture One or Three or Four or Five or Six or Seven or Eight
Go to Cyber Semiotic Institute home page
Go to Course Outline