Chimpanzee Gaze

A series of papers has come out recently from the Primate Research Institute in Kyoto, Japan, using gaze-tracking technology to compare how chimpanzees and humans view photos and videos of both species—and whether there are any major differences in their patterns of viewing.

First, in a study by Kano & Tomonaga (2009a), chimpanzee subjects were presented with full-body images of chimpanzees, humans, and other mammals. Chimps viewed the face more than the body or limbs in each image, but human subjects looked significantly longer at faces than chimp subjects did. This study has shown that chimpanzees shift their gaze faster than humans do, and shift a greater distance across the presented image than humans. However, they tend to have longer fixations on the images.

Of key importance is how chimpanzees view faces, in particular, whether they pay more attention to the eye region than other regions, which is a common pattern in humans—humans gain a lot of semantic and emotional information from looking at others’ eyes. In a second study, Kano & Tomonaga (2009b), chimpanzee subjects were shown images of the faces of chimps and humans, with species-specific face expressions; each face was divided into areas of interest, such as eye, nose, mouth, and other. Humans consistently viewed the eye region, across species and expressions. Chimpanzees looked at the mouth area more than humans, and looked at the mouth more than the eyes in four of eight expressions. These results suggest species-specific strategies for facial communication.

In another study, Hirata et al. (2010), chimpanzee subjects were shown images of chimpanzee faces in either an upright or upside-down position, and with open or closed eyes. The chimps looked longest at the upright faces with open eyes, demonstrating that they process open and closed eyes differently. While chimps are sensitive to eyes, it may be that the eyes might not play as important a role in normal social interaction as in human interactions.

Not surprisingly, Hattori et al. (2010) showed that chimpanzees pay more attention to social cues given by images of conspecific models, rather than human models (human subjects paid equal attention to both chimp and human models). Using gaze-tracking, Hattori demonstrated that when the model looks or reaches to a target object (with an equivalent distractor object on the opposite side), the chimps looked longer at the target in every condition except when the human model looked, which suggests that chimpanzees gain more referential information from conspecific models.

Across studies, humans consistently pay most attention to the eyes, and it is interesting to note the unique morphology of human eyes compared to all other primates: humans have white sclera, and our eyes are horizontally elongated. Kobayashi and Kohshima (2001) studied and compared 88 species of primate eye, and suggest that pigmented sclera is an adaptation to hide gaze direction from others (including predators), and that highly visible human eyes are adapted to enhance gaze cues.

Tomasello et al. (2007) hypothesized that humans’ white sclera evolved to improve mutualistic cooperative and communicative interactions, such as joint attention and pointing. They compared how human infants and adult apes follow the gaze of a human model, who looked up to the ceiling with either his eyes only, head only, or both. Apes followed the model’s gaze more with the head cue, whereas human infants looked up most often when the model gazed using only his eyes. Tomasello writes, “Individuals in constant competition with onlookers should not evolve morphological characteristics to help these others follow their gaze direction.”
While head orientation cues may be more salient than eye-direction cues, Bethell et al. (2007) found that chimpanzee eye movement does have potential signal value to onlookers, and that gaze direction is often incongruent with head orientation, especially during different activities such as feeding, but not grooming.

It is slightly baffling that chimpanzees do look at the eyes in images of faces, but tend to have little success in recognizing attention, and fail at “guesser-knower” and object choice tasks that require the subject to use a human’s gaze cue to find hidden food (see Povinelli et al., 1990, Povinelli & Eddy, 1996, Reaux et al., 1999, Call et al., 1998, and others). However, this pattern makes sense when one considers that chimpanzee social cognition evolved in chimpanzee society, and may not generalize to interactions with humans; furthermore, chimpanzees tend to be more competitive, whereas humans are more cooperative, so it would be unusual for a chimpanzee to share information with a partner about the location of a hidden food, as was done (with a human model) in the failed object choice experiments (See Hare, 2001, Hare & Tomasello, 2004, Hare et al., 2000, Hare et al., 2001 for more on chimpanzees’ success in finding food in competitive conspecific paradigms).

With this improved understanding of chimpanzee ecology and behavior, along with the new data from the Primate Research Institute’s studies, it is important to realize that chimpanzees have different strategies for viewing faces and eyes than humans do, and to take this into account in designing future studies. Some interesting questions to address when comparing chimpanzee and human eyes include the evolutionary “cost” of sclera pigmentation, and whether chimps’ quick and broad scans with short fixations are a different adaptation for a similar purpose as humans’ elongated eye for horizontal scanning, which is presumed to be less energetically costly than a head turn. And, most importantly, what information can a chimp gain from a conspecific competitor’s gaze?

–Katie Hall