The Evolution of the Eye Gaze and its Role in Social Behavior (2006)

As the complexity of social interactions in primates increased, the role of the eye gaze developed from being exclusively for threat signaling to conveying information about the environment such as the social interactions among conspecifics (Tomasello et al, 1998). The duration and direction of the eye gaze provides insight into the mental and emotional states of primate individuals as well as the social contact taking place among them. Thus, to elicit appropriate social responses, it is essential for individuals to accurately recognize gaze stimuli. Researchers agree that the extent to which eye gazing is used as a source of information directly relates to the complexity of the social interactions. However, within primates there are two main hypotheses regarding differences in gaze behaviors, in particular the ability to follow the gaze of conspecifics in search for social information. Some researchers suggest that the main difference in gaze behavior can be observed among great apes, who can follow the gaze, and all other primates, who fail to exhibit this behavior (Povinelli and Eddy,1996 as cited in Emery, 2000, Itakura, 1996). Other researchers hypothesize that there is a gradual transition within primates: while prosimians do not follow gaze, monkeys will follow it but cannot understand its meaning and finally, great apes are capable of following as well understanding gaze and its meaning (Tomasello et al, 1998, Emery 1997). It would follow that this gradual transition continues with humans in a much more complex way that can be studied in detail because of familiarity with human social processes in normal individuals as well as individuals with underdeveloped social skills. In this paper, the methodology and the results testing both of these main hypotheses will be discussed, focusing on the repercussions in social behavior, and the changes in the environment that might have brought about this evolutionary adaptation.

For non-primates, gaze is strictly used in agonistic encounters, where direct eye contact is a threatening signal while an averted gaze indicates submissiveness (Emery, 1997). In primates, gaze is used in a wider contextual range. It provides information about where an animal’s attention is focused at a particular time, which potentially provides insight as to what is affecting its current mental state. Thus, following the direction of the gaze of an individual could provide information about events in the environment that require attention, such as food, predators, and social interactions among group-mates (Tomasello et al, 1998). In terms of social behavior, the gaze indicates dominance and submissive behavior, particularly in the main facial expressions characteristic of primates (Perrett, 1985). Furthermore, the level of attention placed on monitoring an individual’s gaze is indicative of its position in the social hierarchy (Chance, 1967 as cited in Emery et al, 1997), which means that submissive individuals are more likely to monitor the gaze of higher ranking animals in search of social information.

It is natural to ask if there are any relevant differences in the gaze-following behavior of primates, which would suggest dissimilarity in their social organization. To accomplish this, Itakura (1996) studied the gaze behavior of apes, monkeys, and prosimians. The experiment consisted of trials where the primate had to discriminate the direction of the gaze of a human under two conditions: when the gaze consisted of eye movement only, and when the head also moved slightly. After the experimenter approached the cage and tried to make eye contact, the experimenter’s gaze was fixed on a hidden object located behind the animal on one of its sides. It was found that in 70% of the trials, orangutans and chimpanzees correctly followed the gaze of the experimenter, while non-apes did not differ significantly.

At first, it appears that primates that are more closely related to humans place more attention on gaze as a source of social information. The performance of the greater apes confirms the observation that in their natural habitats chimpanzees routinely follow the visual gaze of others to interesting and useful objects and events (Plooij, 1978 as cited in Carpenter et al, 1995). However, when evaluating these findings, it is important to consider that the experimental methodology could have influenced the results. Most importantly, the direct eye contact between the primate and the experimenter could result in the animal feeling threatened. Furthermore, the finding that non-apes do not follow the human gaze does not directly show that they lack the ability, since they might lack motivation to follow social cues by humans, but not from conspecifics (Tomasello et al, 1998) perhaps because of the relevance of the cues. Finally, it is essential to consider that the subjects involved had extensive participation in psychological studies, which can predispose them to pick-up indirect cues from the experimenter, perhaps not related to the direction of gaze.

In an effort to rule out these alternative explanations, an experiment was carried out with rhesus monkeys (Emory, 1997). Instead of direct human interaction, a videotaped image of a monkey was used, and the direction of gaze of the subject was noted under three main conditions. In the first condition, the videotaped monkey was looking towards one of two objects located in opposite spatial directions of the screen. In the second condition, there were no objects, to test for the possibility that the monkeys’ attention to the objects was caused exclusively by their physical characteristics. Finally, in the last condition, objects were presented alone, to test if the videotaped monkey influenced future preference of the object even when the monkey was not present. It was found that rhesus monkeys look in the same direction as the gaze of the videotaped monkey, regardless of the presence of an object. However, this preference did not persist when there is no stimulus monkey. These results suggest that rhesus monkeys have the ability to follow the gaze but they don’t seem to understand the mental significance of it (Emery et al, 1997) as shown by the lack of future preference towards the observed object.

This study strongly suggests that the experimental procedure employed in Itakura’s cross species experiment involving humans affects the performance of non-apes due to motivational deficiencies and the possibility of eye contact being interpreted as a threat. This was further confirmed in a follow-up study involving chimpanzees and monkeys searching for hidden food. Both species used the direction of the gaze of conspecifics comparably to locate the food (Tomaselli et al, 1998). This finding corroborates the observation that there are no noted differences in the social behavior or cognition of these two species (Tomasello & Call, 1997 as cited in Tomasello et al, 1998). The only significant difference between them seems to be a lack of spontaneity in rhesus monkeys in following the gaze when compared with chimpanzees from Itakura’s study (Emery et al, 1997).

The previous studies strongly suggest that the differences between great apes and other primates lies in the attribution of meaning to the gaze, and not in the actual ability to follow gaze, as suggested by Itakura. From this finding, the main hypothesis that develops is that there is a progressive increase in the complexity of eye gaze behavior. The first step in its evolution would be to posses the ability to follow the gaze, which is independent from the ability to understand its meaning (Povinelli and Eddy, 1996b as cited in Emory, 1997). This ability might simply be an extrapolation from the behavior that emerges from direct eye contact, as the submissive individual look at the same location that more dominant individuals are looking. As the need to integrate information about the social world increases, gazing behavior evolved further to include attribution of meaning. This implies that the primate not only follows the gaze to an object, but also tries to understand why it is the source of attention of others. Finally, in the last stage, eye gazing would enable primates to infer information about the motivation behind the actions of others and their current mental state (Emory, 1997, Tomaselli, 1998). This gradual progression directly ties with the proposed evolution for the theory of mind, or the ability to infer information about the mental state of conspecifics.

Given that our knowledge about social interactions in primates is limited, researchers started investigating the human gaze in relation to more familiar social contexts to further understand the evolution of the social gaze. Since human gaze-following behavior develops between the 6th and 18th month of age (Emery, 2000), a procedure paralleling primate studies does not provide a relevant source of comparison. Because of this, experiments have focused on more complex social interactions, such as those involving language.

In humans, gaze is used to gain information about the mental and emotional states of others, such as their level of involvement in communicative behavior (Vertegaal et al, 2001), but it also informs about what situations a human is currently focused on. Gaze also has a very strong influence in the social interactions of humans. For instance, moderate amounts of gazing can influence people’s liking of each other (Vertegaal et al, 2001). In addition, it has an important function in controlling human interaction. It has been found that people gaze more while listening than while speaking, using the gaze to synchronize the conversation and determine when it is appropriate to communicate (Kerdon, 1967 as cited in Kleinke, 1986). On the other hand, gaze is also related to dominance and submissiveness, with dominant individuals establishing longer eye contact and gazing less often for obtaining social information. Similarly, gaze is related to persuasive and deceptive behaviors in humans. Individuals that sustain direct eye contact are more likely to be rated as more credible, while those that are attempting to be persuasive or deceptive increase the frequency of their gaze (Kleinke, 1986). These findings suggest that the ability to control an animal’s own gaze and also be more aware of the gaze of others is a significant advantage in effectively establishing and maintaining social interactions.

All of these correlations corroborate the notion that there is a strong tie between gaze and social interactions in humans. It is relevant to ask how this relationship would be affected if gaze behavior diverts significantly from the accepted social norm. Theoretically, it is expected that the inability to decipher information contained in the gaze would result in marked difficulties in social interactions. To further understand this relationship, autistic individuals were studied for patterns in their gaze behavior. This population is of particular interest since autism is marked by underdeveloped social skills but also by difficulties in understanding the mental states and the intentions of others (Pelphrey et al, 2005). Thus, studying autistic children could corroborate the hypothesis that the relationship between gaze and social interactions encompasses the evolution of the theory of mind.

The study designed to test the gaze behavior in autism consisted of presenting the child with a picture of a cartoon that is looking at one of four different types of candy. Children were then asked which candy the cartoon prefers, with the expectation that they would use gaze cues to base their decision. As expected, there are distinct differences in the gaze-following behavior of autistic individuals. It was found that both normal and mentally retarded children correctly indicate that the preferred candy is the one where the cartoon’s gaze is focused. However, autistic children ignored the direction of gaze of the cartoon when making the decision, and did not significantly point at the candy that the cartoon was looking. This conclusion corroborates that the theory of mind is not as developed when autism is present, since these individuals are unable to gain access to the mental states of others simply by observing their eye gaze.

If the hypothesis that the role of eye gaze progressively evolved in primates, it is relevant to study what physiological changes might have been brought about to support this adaptation. In general, changes in the morphology of the eye among non-human primates has been very subtle, which suggests that even though there are differences in attribution of meaning, in general, the eye gaze behavior did not require physiological changes. The most important change that took place in the transition between non-human primates and humans occurred in the sclera, the outer layer of the eyeball. This adaptation occurred since the color and size of the sclera can conceal information about the gaze to an observer. In primates, it is important to conceal the direction of gaze due to predatory risks, which explains a dark colored sclera. In contrast, the human sclera does not have any pigmentation and it is greatly exposed (Kobayashi, 2001). Changes in the sclera suggest that the role of the eye gaze is increasingly more important in humans, but also support the notion that the social behavior is involved, once predatory risks are greatly diminished by cooperation.

From an evolutionary perspective, gaze-following ability developed as a response to more complex social organizations and the increased need to understand the mental states of others. Based on the gaze behavior of primates, the ability to follow the gaze and understand its meaning appears to have evolved progressively: from prosimians, who do not follow the gaze, to monkeys, which follow it without understanding, then to great apes who can follow it and understand its meaning, and finally to humans, who follow the gaze from an early age, understanding its importance and even studying its relationship with social interactions. The evolution of the role of gaze parallels the evolution of a theory of mind, which suggests that gaze discrimination was an adaptation to gain access to essential information regarding the mental states of primates who are members of social groups. If an animal has the ability to control its own gaze while being perceptive to the gaze of others, it has privileged access to the mental states of others, which provides an advantage in understanding the motives behind their actions. Further understanding of the gaze, also provides valuable skills in establishing and maintaining effectively social interactions. Fully understanding eye gaze behavior promises to be the key to deciphering the evolutionary background to the development of a theory of mind and also to unravel the secret of how social interaction was the key to the evolution of primates.

Carpenter, M. et al. 1995. Joint Attention and Imitative Learning in Children, Chimpanzees, and Enculturated Chimpanzees. Social Development, 4, 217-237.
Emery, N.J. 2000. The eyes have it: Neuroethology, function and evolution of social gaze. Neuroscience and Biobehavioral Reviews, 24, 581–604.
Emery, N.J. et al. 1997. Gaze Following and Joint Attention in Rhesus Monkeys (Macaca mulatta). Journal of Comparative Psychology, 111, 286-293.
Kleinke, L. 1986. Gaze and eye contact. Psychological Bulletin, 100, 78-100.
Kobayashi & Kohshima. 2001. Unique morphology of the human eye and its adaptive meaning: comparative studies on external morphology of the primate eye. Journal of Human Evolution, 40, 419–435.
Pelphrey, K. et al. 2005. Neural basis of eye gaze processing deficits in autism. Brain, 128, 1038–1048.
Perrett, D.I. et al. 1985. Visual cells in the temporal cortex sensitive to face view and gaze direction. Proc. Research Society of London, 223, 293–317.
Tomasello, Call & Hare. 1998. Five primate species follow the visual gaze of conspecifics. Animal Behavior, 55, 1063–1069.
Vertegaal, R. et al. 2001. Eye Gaze Patterns in Conversations: There is More to Conversational Agents Than Meets the Eyes. CHI, 3, 301-308.