Why scientific research can’t determine beauty: the appeal to popularity fallacy
People often treat scientific research as an authority on beauty, which is interesting given the surprising reality that scientists aren’t necessarily experts on fashion, makeup, or aesthetics.
Scientists can still have meaningful things to say about beauty even if they’ve never read Teen Vogue. However, much research on what features make a face beautiful is severely limited by a flaw in the researchers’ premises.
The philosophical flaw: Argumentum ad populum
Millions of people love the latest pop album; only a dozen prefer a more obscure work. Does that make the pop album the better album?
Logically, no. The fact that the majority of people like or dislike something doesn’t say anything definitive about the inherent merit of that thing.
Majority opinion also doesn’t determine truth. So if you survey people about how many planets are in the solar system, and most people tell you 9, that doesn’t make 9 the correct answer (apparently it’s 8 now).
Philosophers have a name for the logical thinking error that occurs when we assume something to be good or true simply because lots of people think so. It’s called the "appeal to popularity” (argumentum ad populum) fallacy.
The appeal to popularity fallacy in beauty research
In scientific research on beauty, study participants typically provide their opinions on faces’ attractiveness. Then, researchers (as well as some organizations; Qoves is the one I’m familiar with, but I’m sure there are others) sometimes use the results of those studies to make claims about which facial characteristics are more and less beautiful.
These claims typically rely on the appeal to popularity fallacy. The flawed logic used in these instances is something like, “The majority of people rated this face as more beautiful, so therefore it is objectively more beautiful.”
This is the same flawed logic as saying, “The majority of people preferred album X to album Y, so therefore album X is objectively better and more aesthetically pleasing than album Y.”
Any time research on beauty relies on participants’ subjective opinions and treats those opinions as if they can determine what’s objectively beautiful, then that research succumbs to the appeal to popularity fallacy. Given that research on beauty is ultimately reliant on participants’ subjective opinions, research on who is “most beautiful” is doomed to succumb to this fallacy.
So, scientific research on beauty, because it inevitably relies on people’s subjective opinions, can never tell us what faces or features are or aren’t beautiful.
Appearance-based discrimination
Critically, while the majority view can’t determine objective beauty, real and extreme consequences can stem from having a face that happens to be very preferred, or not preferred, by a majority. As just one example, appearance-based discrimination may lead to someone being hired or not hired.
Given this, I don’t condemn scientists’ aim to define what the majority considers beautiful. This knowledge could potentially aid in understanding and reducing appearance-based discrimination.
However, advocating for people to change their faces to better conform to faces that were rated as most attractive in scientific research doesn’t seem to be an effective or healthy solution to appearance-based discrimination (more on why below).
Identifying an ideal solution is difficult. A partial solution may be to more consistently acknowledge, including in scientific contexts, that beauty is subjective, and to show greater representation of different types of beauty in popular culture and other areas.
Maybe ideally we’d rewire human brains to care less about beauty, but to totally eradicate humans’ preoccupation with physical appearance probably isn’t realistic.
Other reasons to be skeptical of science-based beauty advice
A summary of just some of the reasons why trying to conform to specific science-based beauty standards (like how symmetrical you should be, or what shape your jaw should be, or how far apart your eyes should be, or how wide your nose should be, etc.) may not be an ideal approach:
Beauty research can be biased
Scientific research on beauty is rife with bias, whether because of participants’ age, culture, race, socioeconomic status, etc. For example, if the study sample was mostly white American college students, the results of the study may not reflect what the majority of people, who aren’t white American college students, thinks about beauty (not that the majority opinion has any kind of objective truth, as just discussed). Cultural factors in particular have enormous influence on what is generally deemed beautiful.
Beauty research can have methodological flaws
Scientific research can exhibit methodological flaws that compromise the validity of the results, and this may be particularly true when the research topic is something as subjective as beauty. As just one example of many, a difference can be found in a study that a researcher deems meaningful but that ends up being pretty meaningless in real life. For instance, say Face A was found on average across participants to be more attractive than Face B, and the researchers conclude that this was a significant finding and that Face A is in fact more attractive. But Face A may have been rated on average a 4.5/5 on the attractiveness scale, while Face B was rated a 4.3/5. If you have enough participants, that small difference could end up being statistically significant (in statistical analysis, a larger sample size increases the statistical power of a study, which means the study can detect smaller differences). But in real life, a person might not even be able to distinguish between what they rate as a 4.5 and 4.3. Scientists aren’t always great about saying whether the results of their studies are meaningful in real life.
Beauty research can obscure the important influence of individual preferences
Research suggests that people have individual preferences for what they find beautiful. For example, people may tend to be attracted to people who resemble themselves, their families, or other faces they saw a lot of growing up, whether in real life or virtually. Importantly, the research that revealed those findings may also be methodologically flawed, but one advantage that this research has over a lot of studies on beauty is that it takes individual differences into account. Meaning, it proceeds from the accurate assumption that people differ in what they find attractive.
Statistical techniques can create misleading results
Much research on beauty uses statistical techniques to find what “on average” is most attractive, which may end up obscuring important data points. For example, maybe most people rated Face A as a 7/10 on attractiveness; Face B was more polarizing and some people rated it as 10/10, but many others rated it as 5/10. Face A might end up being deemed more attractive by scientists than Face B, because of the way statistics work by averaging participant responses. But there was a decent-sized subset of participants who thought Face B was considerably more attractive, it’s just that the study obscured this result.
People’s opinions on beauty can be influenced by fleeting trends
Beauty trends come and go, and something that’s popular now (like a super-defined jawline) won’t necessarily be popular a few years or even months from now.
Finally, it sends an extremely negative message about human behavior to find a hyper-specific set of attributes that are most popular and to subsequently recommend that everyone try to better embody those attributes. Saying there’s a certain specific set of facial attributes that’s most attractive is like saying there’s a certain specific set of personality traits that’s most attractive, and therefore everyone should try to change their traits to better approximate the most attractive personality. You should strive to be this percentage extraverted, this percentage agreeable, this percent adventurous, etc., to have the ideal exact set of personality traits that appeals to the widest array of people.
Just as we don’t want a society of psychological clones who all think and act the same, we don’t want a society of visual clones who all look the same.
Is style analysis also based on the appeal to popularity fallacy?
Not inherently. But depending on how it’s presented, it could be. For example, if we say something like, “Dressing based on your style type will make you look harmonious, and the majority of people find harmony beautiful.” I admittedly probably have made statements like this.
The crucial difference, though, is that dressing and doing your hair and makeup harmoniously is so much more realistic than changing the shape of your eyes or size of your jaw or whatever other recommendations may come from a science-based approach to beauty. Style analysis also ultimately validates individuality and uniqueness rather than advocating conformity to majority view. It’s a healthier alternative to the philosophically-flawed idea that there’s a certain specific set of traits that are inherently most beautiful.
So I don’t think style analysis succumbs to the appeal to popularity fallacy, because the idea isn’t that you should look more harmonious because it’ll make other people see you as more beautiful. Ultimately the idea is that if you look more harmonious, you may—you won’t necessarily, but you may—see yourself as more beautiful.
Many people may not see style analysis as an effective way to enhance beauty, and that’s fine. I admit that my perspective on facial aesthetics is subjective, and that not everyone shares it—and that’s part of the beauty of beauty. We can have unique views.
