Just Another Face in the Crowd: The Evolution and Mechanisms of Primate Facial Processing

Clem Doucette and Daniella Lorman

Illustrations by: Adlai Brandt-Ogman

You are driving down a winding country road when you notice that your car’s “check engine” light turns on. Moments later, you feel the gears in the car’s engine begin to grind before failing completely. As you gently bring your vehicle to a stop along the side of the road, the gravity of your predicament begins to set in. There is no cell service and you are at least ten miles away from the nearest town; your only option is to flag down a passing motorist for help. Fortunately, a driver pulls over and offers to help you. Can you trust them?  

You notice signs that may lead you to trust the driver. Perhaps there is a child in the back seat of the car. Or, the driver may be roughly the same age or gender as you. But, there are other factors that contribute to forming a perception of trustworthiness that you may not be aware of. These cues can be as conspicuous as the individual’s age and gender, or as subtle as the shape and contours of their face. What if we were to tell you that before you can even consciously think through this information, you have already subconsciously decided from the driver’s face alone whether they are trustworthy or not? In fact, it takes just 33 milliseconds for your brain to form a first impression [1]. While trust and first impressions seem to us like innately human experiences, evidence suggests that non-human primates interpret faces very similarly to how we do. In the past few decades, scientists have researched and debated the existence of brain regions specific to facial processing, and have investigated the mechanisms behind how human and non-human primates recognize and judge faces. This research posits that there is an evolutionary benefit to these mechanisms; they may help primates identify threats and effectively navigate complex social environments.

First Impressions Count: Are Some Faces Inherently Trustworthy?

Imagine that your town is in the midst of a local election. Campaign signs dot lawns throughout your neighborhood bearing the names and faces of various different candidates. You don’t know anything about the candidates themselves or their political affiliations. What you may not realize is that you have already formed an opinion of the candidates based on their facial features alone. This is an example of a phenomenon termed the “first impression effect,” or the rapid judgement of one’s character based solely on one’s facial characteristics and structure. 

In a 2008 study, undergraduate student volunteers were asked to assess a series of  computer generated faces on qualities of trustworthiness, dominance, and threat [2]. Researchers found that the faces perceived as being most trustworthy were ones that featured more feminine and youthful structural characteristics. Conversely, faces with more masculine characteristics were perceived as being untrustworthy, dominant, and aggressive. Why? Researchers posited that expressions of dominance and anger subconsciously signal the capability and intent to cause harm, resulting in these faces being labeled as untrustworthy [2]. In many situations we can identify a person’s mood or intentions based on their facial expressions and characteristics alone. We can usually infer that someone who is smiling and laughing is in a good mood, while the person scowling and glaring angrily from the corner of the room, is not. But what happens to our facial perception skills when we are tasked with analyzing the expressions of non-human primates? Can we distinguish a threatening ape from a friendly one?

Unless you are familiar with non-human primate behavior, the answer is probably no. Imagine that you have finally stumbled upon the opportunity of a lifetime: the chance to visit a nature park and see apes in their natural habitat. As you hop into the driver’s seat of your rented camo Jeep, you can barely contain your excitement. Suddenly, a large macaque monkey bounds towards your car, baring its teeth and hooting. He is happy to see you! Or, at least that is what your first impression effect is telling you. On the contrary, if you see a macaque or another non-human primate “smile” at you, you should probably back off. For macaques, the bared-teeth facial expression is one closely associated with a few different emotional states [3]. A wide, open-mouthed grin is typically a threat; the monkey shows off its sharp teeth in an attempt to intimidate [3]. Other expressions that resemble a human smile or grin may signify fear and submission [4]. On the surface, many of these non-primate facial expressions bear close resemblance to our own familiar expressions, such as smiles, frowns, and grimaces. Your first impression of the macaque you encounter is likely based on your anthropomorphic (or human-based) concept of what a friendly and welcoming face should look like. This may make you wonder: does a similar effect occur when a macaque glances at a human face?  

A 2018 study conducted on macaque monkeys hypothesized that the first impression effect is also exhibited by non-human primates [5]. Researchers theorized that faces deemed trustworthy by macaques would feature a low facial width to height ratio (fWHR). Faces with a low fWHR are long and narrow, while those with a high fWHR are short and round. They concluded that the macaques, just like humans, displayed a preference for faces with a low fWHR and a more feminine, youthful appearance [5]. Furthermore, faces with high fWHR are not merely perceived as being more masculine and aggressive, but may indicate hostile personality traits. A 2015 survey of male hockey players demonstrated that sportsmen with a high fWHR often had higher baseline levels of testosterone, and consequently behaved more violently on the ice [6]. Resultantly, macaques and other primates are more inclined to approach individuals with low fWHR faces; we implicitly associate these faces with fearful and weak tendencies, regardless of what feeling they outwardly display [7].

Over time, primates have adapted to their lives in close-knit social groups by developing this common facial perception mechanism [5]. Consider a round of the game “Among Us”; to succeed and ensure your group’s survival, you must quickly identify the dangerous impostor. When scanning the map for clues, you notice that Red is behaving erratically, failing to complete tasks, and is even following and covertly killing your teammates. Immediately, you know that this player is a dangerous outsider. Similarly, in the setting of a primate social group, effectively identifying which individuals are a threat benefits the survival of its members and the integrity of the group.

At this point, you may be wondering how all of these mechanisms actually work. How does your brain even recognize what a face is? Let’s delve into the neuroscience behind facial recognition, and how our brain perceives faces. 

How do our brains perceive faces? 

All faces are pretty similar. Of course, your dad looks different from Theo James, who looks different from your best friend. But most faces have two eyes, a nose and mouth. And, we know how facial features are usually arranged in relation to one another. It turns out that our brains are actually set up to prioritize processing information about faces. The fusiform face area (FFA) is responsible for helping us determine what kind of objects we are looking at [8]. This tiny brain region, located within the brain’s temporal lobe, may partially account for our ability to distinguish between the face of our celebrity crush or best friend. Though scientists are still debating if the FFA is actually specific to faces — or just really good at identifying complex objects that we see a lot —  there is a plethora of evidence suggesting that facial perception gets its own real estate within the brain. In other words, we think the FFA is face specific [8]. 

The FFA’s activity is often studied using fMRI machinery. As research subjects are exposed to images of faces or different facial stimuli, the fMRI machine captures images illustrating the activity levels of different regions within their brains [8]. fMRI machines use magnetic fields and radio waves to non-invasively image bodily tissues, such as the brain [8]. This technique works because brain activity requires blood flow. As blood flow increases, the ratio of oxygenated blood to deoxygenated blood in the part of the brain that is working extra hard also increases. The MRI machine then detects this image. In several studies, fMRI scans showed increased blood flow, or activity, in the FFA when looking at faces compared to other objects [9]. In fact, this result is observed across a wide range of faces, such as front views, side views, cartoons, and even dog faces [9]! 

Ninety-seven percent of neurons in the FFA have been recorded to selectively respond to faces [10]. In a 1997 study, participants were shown various images of random inanimate objects, hands, the “faces” of houses, and, of course, human faces [9]. Scans showed that the FFA responded much more to the pictures of faces than the other objects. This study established the FFA as the most important area of the brain for processing facial stimuli and prompted a slew of further research on the topic [9]. Almost ten years after this groundbreaking study, scientists questioned whether the FFA functions similarly in non-human primates [10]. Researchers tested how macaque monkeys responded to different images, and found that neurons in the FFA responded specifically to images that displayed facial stimuli [10]. 

Furthermore, a recent study found that face-specific areas of the brain will respond to faces even if you have never seen one before [11]. Researchers asked volunteers who were blind from birth, as well as several sighted people, to explore a series of 3D printed objects — such as a chair, a maze, a face, and a hand — by touch. When sighted people looked at the 3D-printed faces, an MRI scan showed activation in the FFA. More strikingly, however, as sighted participants explored objects by touch only, the FFA lit up again, but not as brightly. When blind volunteers explored facial objects, the FFA lit up again. Researchers ran the experience three times to make sure they were really seeing this effect [11]. These results were significant because they helped researchers discredit previous theories of facial recognition. 

One such theory was that face-specific regions developed in response to visual information that comes from the center of your visual field, or your fovea, as opposed to from the periphery of your visual field [12]. Since we normally shift our gaze to center other human faces, it was thought that this brain area might be triggered simply because we happen to put faces in the middle of our visual field [12]. The aforementioned study contradicts this theory because no visual cues were necessary to make the FFA regions of blind participants light up [11]. Another earlier theory was that the FFA may just be sensitive to curved objects rather than rectangular ones, rather than being face-specific [13]. In one study, researchers asked blind participants to handle 3D shapes-- like cubes, spheres, and eggs-- and the FFA did not respond any more to curved objects than rectangular ones [11]. To test for additional variables, researchers played different sounds for blind participants. Some were related to behaviors involving facial features, like laughing and chewing, while others depicted natural scenes such as waves crashing. The FFA was found to respond specifically to face related sounds [11]. In a case study, researchers placed electrodes on the brain of a volunteer, Ron Blackwell [14]. As electrodes move over the brain, their contact acts like a microphone, listening in on brain activity. The researchers stimulated different brain neurons and watched in amazement as Ron lost the ability to recognize faces as neurons within the FFA were stimulated! He said that the “face metamorphosed” [14]. In summary, significant research suggests that the FFA plays a crucial role in our ability to recognize and perceive faces. But how can we explain times when our brains perceive faces in instances where there are none?

Monkey See, Monkey… Don’t: The Phenomenon of Pareidolia

Have you ever wondered why the electrical outlets in your house look perpetually surprised? Or, why, right as you are about to fall asleep, the pile of laundry in the corner of your room seemingly morphs into a menacing face? These visual phenomena, known as pareidolia, have fascinated (and frightened) us for centuries. Pareidolia occurs when we incorrectly perceive an abstract set of visual stimuli as something that we are familiar with [15]. A classic example of this is the infamous Rorschach inkblot test. You’ve probably seen this depicted numerous times in vintage spy dramas and cop shows; the hardboiled detective shows the cold-hearted killer a series of abstract splotchy images, while the killer reports what they think the pictures look like. While pareidolic images can resemble animals, inanimate objects, and a myriad of other easily recognizable forms, facial pareidolias occur most frequently [15]. The theory of gestalt psychology suggests that organisms perceive whole patterns and configurations, and not just their individual components [16]. We are intrinsically aware that a face consists of several pieces, such as two eyes, a mouth, ears, and a nose. For example, when someone texts you a :-) icon, you likely see a simplistic representation of a smiling face and not just a random assemblage of punctuation marks. Why, then, do our minds seemingly trick us into seeing faces when they aren’t there?

Actually, our ability to see silly faces in power outlets and emoticons likely serves an important evolutionary function. In a 2017 study conducted on rhesus monkeys, scientists showed subjects a series of both pareidolic and non-pareidolic images [17]. On average, the monkeys fixated on the pareidolic images for over twice as long as the non-pareidolic ones. Furthermore, after tracking and mapping the gaze patterns of the monkeys, researchers determined that subjects focused on the regions where the eyes and mouth are usually found [17]. So, how does this hypersensitive facial recognition mechanism help us survive? Imagine you are a wild primate, swinging through the trees of a dense jungle. In the dark shadows on the forest floor, you see what appears to be a glowing pair of eyes and a menacing mouth. As you leap to safety, a large leopard jumps out of the bushes and devours your (much less perceptive) peer. In short, primates that quickly spot face-like patterns in nature are more likely to survive and protect themselves from threats than those primates that could not. Our hypersensitive facial detection mechanisms may be heritable, passed down from our well-adapted primate ancestors [18]. 

Have I Seen You Before? 

The ability to perceive faces as a whole also enables us to recognize individual faces, like those of our friends, acquaintances, and particularly famous celebrities. However, this mechanism can sometimes malfunction. Prosopagnosia, commonly known as face blindness, is a cognitive disorder marked by an inability to recognize familiar faces [19]. This is vastly different from the occasional slip-ups many of us have when encountering an old acquaintance or a distant work colleague. Face blindness can manifest in different forms and severities; some people are capable of recognizing close family members, but experience great difficulty telling coworkers apart. In some severe cases, prosopagnosics can even struggle to recognize their own face [19]. 

Most cases of prosopagnosia result from damage to brain regions responsible for facial processing and the amygdala in instances of trauma or stroke [20]. As the amygdala functions in emotional memory, these two structures work together to form a databank of faces a given person has seen during their life and the emotions they associate with each face. When the structures are injured, this databank is essentially wiped. Not only can a patient no longer remember the faces of people they have seen, the brain no longer files new faces away for later retrieval [21]. 

Prosopagnosia and pareidolia have provided important insight into how facial recognition works and may indicate the existence of a specific facial perception mechanism in the brain. For instance, if you see a slender, metallic object with wings and propellers, you know that it is an airplane. However, if someone were to ask you what the exact model of the airplane is, or how it is different from other airplanes, it's likely that you would have no idea unless you are an aviation enthusiast. Similarly, many people with prosopagnosia can recognize an object as a face based on its fundamental layout but are unable to interpret unique facial structures as a whole face. These phenomena demonstrate that facial perception is potentially a holistic, or gestalt, process; we recognize a face as the sum of its different parts, or facial features. 

A Face is Worth a Thousand Words

From spotting dangerous predators on the savanna to determining whether or not a stranger can be trusted, our facial perception mechanisms have helped ensure our survival for millennia. For primates, faces are some of the most essential visual stimuli; our acute ability to monitor and interpret them is fundamental to how we interact, socialize, and identify with one another. Research on the development and evolution of facial processing has boomed in the past few decades, and certain findings — such as the existence of a specific facial processing mechanism in the brain — provide critical and groundbreaking insight into how we perceive faces. Still, facial perception is an incredibly complex and multifaceted subject that intersects with a wide array of disciplines, ranging from neuroscience to anthropology and psychology. Conducting further comprehensive, interdisciplinary research on primate facial processing and its evolution may help scientists better understand and elucidate the many aspects of this intriguing topic. 

---

REFERENCES

1. South Palomares, J. K., & Young, A. W. (2018). Facial first impressions of partner preference traits: Trustworthiness, status, and attractiveness. Social Psychological and Personality Science, 9(8), 990–1000. doi: 10.1177/1948550617732388

2. Oosterhof, N., & Todorov, A. (2008). The functional basis of face evaluation. Proceedings of the National Academy of Sciences, 105(32), 11087–11092. doi:10.1073/pnas.0805664105 

3. Burrows, A. M. (2008). The facial expression musculature in primates and its evolutionary significance. BioEssays, 30(3), 212–225. doi: 10.1002/bies.20719 

4. Waller, B. M., Julle-Daniere, E., & Micheletta, J. (2020). Measuring the evolution of facial ‘expression’ using multi-species FACS. Neuroscience & Biobehavioral Reviews, 113, 1–11. doi: 10.1016/j.neubiorev.2020.02.031 

5. Costa, M., Gomez, A., Barat, E., Lio, G., Duhamel, J.-R., & Sirigu, A. (2018). Implicit preference for human trustworthy faces in macaque monkeys. Nature Communications, 9(1), 4529. doi:10.1038/s41467-018-06987-4

6. Haselhuhn, M. P., Ormiston, M. E., & Wong, E. M. (2015). Men's facial width-to-height ratio predicts aggression: a meta-analysis. PloS one, 10(4), e0122637. https://doi.org/10.1371/journal.pone.0122637

7. Deska, J. C., Lloyd, E. P., & Hugenberg, K. (2018). The face of fear and anger: Facial width-to-height ratio biases recognition of angry and fearful expressions. Emotion, 18(3), 453–464. doi: 10.1037/emo0000328

8. Gore J. C. (2003). Principles and practice of functional MRI of the human brain. The Journal of Clinical Investigation, 112(1), 4–9. doi: 10.1172/JCI19010 

9. Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: A module in human extrastriate cortex specialized for face perception. doi:10.1523/JNEUROSCI.17-11-04302.1997

10. Tsao, D. Y. (2006). A cortical region consisting entirely of face-selective cells. Science, 311(5761), 670–674. doi: 10.1126/science.1119983 

11. Ratan Murty, N. A., Teng, S., Beeler, D., Mynick, A., Oliva, A., & Kanwisher, N. (2020). Visual experience is not necessary for the development of face-selectivity in the lateral fusiform gyrus. Proceedings of the National Academy of Sciences, 117(37), 23011–23020. doi:10.1073/pnas.2004607117

12. Mäkelä, P., Näsänen, R., Rovamo, J., & Melmoth, D. (2001). Identification of facial images in peripheral vision. Vision Research, 41(5), 599–610. doi:10.1016/s0042-6989(00)00259-5 

13. Kosslyn, S. M., Hamilton, S. E., & Bernstein, J. H. (1995). The perception of curvature can be selectively disrupted in prosopagnosia. Brain and Cognition, 27(1), 36–58. doi:10.1006/brcg.1995.1003 

14. Parvizi, J., Jacques, C., Foster, B. L., Withoft, N., Rangarajan, V., Weiner, K. S., & Grill-Spector, K. (2012). Electrical stimulation of human fusiform face-selective regions distorts face perception. Journal of Neuroscience, 32(43), 14915-14920. doi:10.1523/JNEUROSCI.2609-12.2012

15. Voss, J. L., Federmeier, K. D., & Paller, K. A. (2012). The potato chip really does look like Elvis! Neural hallmarks of conceptual processing associated with finding novel shapes subjectively meaningful. Cerebral Cortex, 22(10), 2354–2364. doi:10.1093/cercor/bhr315

16. Sternberg, R. J., Sternberg, K., & Mio, J. S. (2012). Cognitive Psychology. Wadsworth/Cengage Learning. 

17. Taubert, J., Weldon, K.B. & Parr, L.A. (2017). Robust representations of individual faces in chimpanzees (Pan troglodytes) but not monkeys (Macaca mulatta). Anim Cogn 20, 321–329. doi: 10.1007/s10071-016-1054-6

18. Wilmer, J. B., Germine, L., Chabris, C. F., Chatterjee, G., Williams, M., Loken, E., Nakayama, K., & Duchaine, B. (2010). Human face recognition ability is specific and highly heritable. Proceedings of the National Academy of Sciences, 107(11), 5238–5241. doi: 10.1073/pnas.0913053107

19. Biotti, F., & Cook, R. (2016). Impaired perception of facial emotion in developmental prosopagnosia. Cortex, 81, 126–136. doi: 10.1016/j.cortex.2016.04.008

20. Gainotti, G., & Marra, C. (2011). Differential contribution of right and left temporo-occipital and anterior temporal lesions to face recognition disorders. Frontiers in Human Neuroscience, 5, 55. doi: 10.3389/fnhum.2011.00055

21. Corrow, S. L., Dalrymple, K. A., & Barton, J. J. (2016). Prosopagnosia: Current perspectives. Eye and Brain, 8, 165–175. doi: 10.2147/EB.S92838