Lying Eyes - Part 2
The Science and Mythology of Trusting What We See
The eyes’ role in our sense of who to trust is impossible to overstate.
Previously, we explored how our eyes can deceive us and how similar perceptual biases affect our other senses, including our sense of trust. This relationship between vision and trust operates both internally - how we process visual information - and externally - how we present ourselves to others.
As the saying goes:
The eyes are the window to the soul.
This piece of common sense, or common knowledge, has been extensively studied and replicated to the point that at a minimum, we can say — people really do believe this.
Two studies in particular break down which types of ‘eyes’ are the most trustworthy.
The first from 2008 is by Alexander Todorov and Nikolaas Oosterhof who published their groundbreaking study "The Functional Basis of Face Evaluation" in the Proceedings of the National Academy of Sciences, which systematically mapped how facial features influence trust judgments, in their study:
…using a commercial software program that generates composites of human faces (based on laser scans of real subjects), the scientists asked another group of test subjects to look at 300 faces and rate them for trustworthiness, dominance and threat. Common features of both trustworthiness and dominance emerged. A trustworthy face, at its most extreme, has a U-shaped mouth and eyes that form an almost surprised look. An untrustworthy face, at its most extreme, is an angry one with the edges of the mouth curled down and eyebrows pointing down at the center. The least dominant face possible is one resembling a baby’s with a larger distance between the eyes and the eyebrows than other faces. A threatening face can be obtained by averaging an untrustworthy and a dominant face.
In this particular study, they are looking at the whole face, so you can see the shape of the eyes, as well as the shape of the eyebrows, and the ratio of the nose to the eyes all play a role in our judgments of trustworthiness. Notice also that dominance is largely a matter of jawline in this study. Using a bit of unscientific inference, we can see these traits exemplified in well-known fictional characters. Consider Peter Pettigrew from Harry Potter - a character designed to appear untrustworthy with his small, watery eyes and weak jawline, creating what the study would classify as a "baby-faced" but untrustworthy appearance. In contrast, Santa Claus represents the other end of the spectrum - a dominant face structure softened by features we associate with trustworthiness, like the U-shaped smile and warm eyes.
This intentional use of facial features to signal character is particularly evident in the film adaptation of The Hunger Games, where actress Julianne Moore's appearance as President Coin was carefully crafted to align with these unconscious triggers of distrust. The makeup and styling team deliberately shaped her eyebrows outward and down, and emphasized her sharp jawline - choices that, when compared to Moore's typical appearance, clearly aimed to trigger our instinctive distrust response.
These intentional styling choices become even more apparent when examining another famous character: Daenerys Targaryen, played by Emilia Clarke. In the photo below her appearance was carefully manipulated, particularly her eyebrow shape, nose-to-eye ratio, and chin shape to create a Low and High Trust version. As one Reddit commenter observed, "Low trust gives me predatory animal vibes" - suggesting that low-trust faces appear hungry or threatening. This observation carries complex implications, particularly for women. In the same discussion thread, others noted that low-trust faces on women are both "striking" and create the quintessential "resting-bitch face" - inadvertently revealing how powerful women are often perceived as untrustworthy by default, forcing them to carefully consider how they present themselves.
In the Eyes of the World
While these examples provide compelling insights, they raise an important question: do these facial trust indicators hold true across different cultures? After all, the previous examples focused primarily on Caucasian faces. A 2022 study published in PLOS One by Ce Mo, Irene Cristofori, and colleagues titled "Culture-free perceptual invariant for trustworthiness" took this research further by examining these trust judgments across cultures. What makes this study so interesting is not just that it controls for cultural factors, White vs Asian, but also uses a better method. Instead of composite faces, Mo and Cristofori use noise to augment their different faces. What this means is that the authors can get much more granular about what precisely makes a face trustworthy because the changes are at the pixel level. Additionally, the secondary control is for attractiveness, not dominance, as in the first study.
What the study finds is that across cultures, the eyes themselves are the most important factor in determining trustworthiness. The researchers used heat map visualizations to show their findings, where blue coloring indicates features that correlate positively with perceived trustworthiness. As shown in Figure 5, the eye regions consistently appear in blue across all conditions, indicating their universal importance. Interestingly, while facial features like smiles may help when you're a "stranger in a strange land" (as seen in the out-group rating panel), the eyes remain the primary trust indicator whether you're being judged by people from your own cultural group or by those from another culture. The right side of the figure shows attractiveness ratings for comparison, marked as "N.S." (not significant), demonstrating that trustworthiness judgments are distinct from simple attractiveness.
Specifically, the authors state:
The specific facial feature for trust that emerged as a “universal” cue was high contrast eyes characterized by large dark pupils. Chinese and French participants implicitly darkened the eye region of Asian and European descent faces, to make the face more trustworthy.
In other words, it was pupil dilation that the test subjects focused on when determining which faces to trust regardless of gender or race. In addition to bright light, people’s pupils dilate both when seeing a dead body, and when having an orgasm. Pupil size is such a reliable indicator for arousal that it even has its own name, pupillometry, which has been used to assess everything from sleepiness, introversion, sexual interest, race bias, schizophrenia, moral judgment, autism and depression. The infamous psychologist Daniel Kahneman used pupillometry to show that the degree of pupil dilation is in proportion to the difficulty of the task at hand. Multiply 25 x 40, and your pupils dilate a little bit. Multiply 56 x 42, and your pupils will widen until you find the answer.
In the case of trust, people are clearly using pupil dilation to infer a person’s state of mind and science would seem to back up this intuition. As researchers explain:
Stimulation of the autonomic nervous system's sympathetic branch, known for triggering "fight or flight" responses when the body is under stress, induces pupil dilation. Whereas stimulation of the parasympathetic system, known for "rest and digest" functions, causes constriction. Inhibition of the latter system can therefore also cause dilation. The size of the pupils at any given time reflects the balance of these forces acting simultaneously.
So what underlying story do we construct when we associate dilated pupils with trustworthiness? One part seems to be around the perception of surprise and uncertainty. For example, when a stranger looks lost and asks if they could look at my phone for directions, I would let them. However, if someone calmly walks over and asks for my phone, I am leaving. Ironically, it’s the calmness of the sociopath that we find most unnerving and feeds our instinctual sense of (dis)trust. This is the second part of the story we read into people’s eyes — their emotional arousal. We can not always tell which emotion, but we can definitely tell when the emotion is absent.
This phenomenon is perhaps most starkly illustrated by what combat veterans and photographers call the "thousand-yard stare" - a vacant, unfocused gaze characterized by constricted pupils, often observed in soldiers who have experienced intense combat. The consistency of this expression across wartime photographs from different eras is striking, with the pinpoint pupils being a common feature.
But is it true?
As noted at the beginning, while scientific evidence confirms that people strongly believe they can judge trustworthiness from a person's eyes in first impressions, and while pupils do indeed reflect emotional states, the reliability of these judgments is questionable. Research has consistently shown that humans are absolutely awful lie detectors. Even trained professionals like law enforcement officers and FBI agents perform no better than random chance when attempting to detect deception in controlled studies. Ironically, those who relied most heavily on physical cues like pupil dilation tended to perform worst of all.
I know this is a frustrating conclusion but it is an important one. People will and do judge other people’s trustworthiness based on their eyes, but it’s a terribly faulty measure. It is not that it tells us nothing, it’s that we give it an unwarranted level of certainty than it deserves.
In the end, the eyes are not so much the window to the soul as they are the curtains. You can see the shadows of who is home, but you do not see them, you do not see inside, even if you think you do.
References:
Bond, C. F., & DePaulo, B. M. (2006). Accuracy of Deception Judgments. Personality and Social Psychology Review, 10(3), 214-234. https://doi.org/10.1207/s15327957pspr1003_2
Eggleston, A., Tsantani, M., Over, H. et al. Preferential looking studies of trustworthiness detection confound structural and expressive cues to facial trustworthiness. Sci Rep 12, 17709 (2022). https://doi.org/10.1038/s41598-022-21586-6
Hartwig, Maria & Bond, Charles. (2011). Why Do Lie-Catchers Fail? A Lens Model Meta-Analysis of Human Lie Judgments. Psychological bulletin. 137. 643-59. 10.1037/a0023589.
Konrad, K.A., Lohse T., Qari S. (2014) Deception choice and self-selection–the importance of being earnest J. Econ. Behav. Organ., 107 , pp. 25-39
Mo, C., Cristofori, I., Lio, G., Gomez, A., Duhamel, J.R., Qu, C., & Sirigu, A. (2022). Culture-free perceptual invariant for trustworthiness. PLOS One, 17(2), e0263348. https://doi.org/10.1371/journal.pone.0263348
Ockenfels, A., Selten, R. (2000) An experiment on the hypothesis of involuntary truth-signalling in bargaining Game. Econ. Behav., 33 , pp. 90-116
Oosterhof, N., & Todorov, A. (2008). The Functional Basis of Face Evaluation. Proceedings of the National Academy of Sciences, 105, 11087-11092. https://doi.org/10.1073/pnas.0805664105