Undoubtedly the scariest time of year (besides maybe tax season) is Halloween – as October rolls around, we pause our perpetual avoidance of fear and instead embrace it, reveling in haunted houses and scary costumes. But how do we know what scary is, and why are many things universally frightening? Part of the answer could be cultural: for example, the mask from Scream or the hooded, scythe-wielding Grim Reaper. There could also be evolutionary influences on what we find frightening, such as sharp teeth and grotesque facial features.
Some scientists have approached this question from the field of face perception. They ask a more specific question: What makes a face scary? Previous research on face perception has examined the ability of humans (and, amazingly, some animals) to detect distorted faces, and has gauged preferences for attractive faces over unattractive ones. Sinnott et al. (2012) took this latter line of research further to investigate how animals and humans perceive scary faces – in particular, scary Halloween masks.
They began with three hypotheses about how non-human animals might react to the masks:
1. Human Cultural Hypothesis – Faces are scary not because of any particular way they look, but because of the cultural meaning associated with them. So a skull is scary because we associate it with death, the Scream face is scary because of its associations with the horror film, etc. If this hypothesis is true, we would expect animals to be unaffected by the masks, compared to humans.
2. General Biological Hypothesis – Many scary masks possess features commonly associated with predators, such as sharp teeth and large, angry eyes. This hypothesis predicts that animals would therefore perceive the masks as predators and be frightened of them.
3. Primate Biological Hypothesis – Primates have more complex faces than other animals – they can move facial features independently of each other to make various facial expressions, and they can use those expressions as way to communicate emotions like aggression and fear. Nonhuman primates thus may be more likely than other animals to perceive the frightening nature of the masks. If this is the case, then non-human primates, but not other animals, should be frightened of the masks.
In order to test these hypotheses, Sinnott et al. studied the avoidance response latency for animals to take food from a masked experimenter. In each trial, an experimenter wearing a mask offered an animal some food, then measured the amount of time the animal hesitated before taking it. If an animal is frightened, it will be more cautious and hesitate longer before taking the food. On the other hand, an animal unaffected by the mask will take the food with little or no hesitation.
Sinnott et al. tested 13 different Halloween masks, ranging from politicians to vampires to aliens. They also conducted control trials, where the experimenter was unmasked, to rule out the possibility that the animals were frightened by humans in general. They tested a wide variety of animals, including several primate species, lions, a bear, a camel, and macaws.
They found that primates had significantly longer response latencies than non-primates for all masks. There was no difference between primate and non-primate latencies in the control trials, ruling out the possibility that the primates were just more afraid of humans. Since only the primates were affected by the masks, Sinnott et al. rejected the General Biological Hypothesis, which posits that all animals should be frightened by the masks.
The fact that the non-human primates had longer response latencies suggests that the Human Cultural Hypothesis may also be incorrect, but Sinnott et al. wanted to directly compare non-human primates and humans by investigating whether they found the same masks to be scary. They asked humans to rate the masks on a scale of 1 (not scary) to 7 (very scary). When they compared those ratings to the non-human primates’ response latencies, they found a significant correlation – in general, non-human primates were more afraid of (had longer response latencies for) the masks that humans rated as scarier! This supports the Biological Hypothesis – the primates perceive the frightening nature of the masks, probably due to their greater sensitivity to facial expressions.
There were a couple interesting differences in how non-human primates and humans perceived the masks. For example, the non-human primates had longer response latencies for the politician masks, which the humans rated as not very scary. Sinnott et al. suggest that this is because the politician masks have big, toothy smiles, which the non-human primates likely perceive as “fear grimaces” (a smile-like baring of teeth that indicates fear).
The results of this study suggest that some higher-level cognitive process occurs in the primate brain when perceiving the scary masks (and probably faces in general). The presence of predator-like facial features (like big teeth) alone isn’t sufficient to elicit fear, or else all the animals would have been affected by the masks. Rather, primates may interpret what they perceive at a higher cognitive or even emotional level, which causes them to fear a face (or not).
So although humans may find some faces scary due to cultural associations, there’s also likely an evolutionary influence on the scariness of a face, based on facial features and their configuration, and requiring higher emotional and cognitive processes.
After all that talk about fear, I’ll end this post on a fun note: here’s a video showing how the animals at the London Zoo celebrated Halloween this year. Zoos often use holidays and birthdays as opportunities to provide the animals with themed enrichment (new and interesting things to explore and eat!).
Sinnott, Joan M., et al. “Perception of Scary Halloween Masks by Zoo Animals and Humans.” International Journal of Comparative Psychology 25 (2012): 83-96.