How do humans learn what’s beautiful in a face? Can chickens be trained to find beauty in human faces the way humans do? A group of researchers trained chickens to be attracted to human faces and found that the chickens had the same exact preferences for faces as college students. Lu and Tirth discuss this incredible finding on S01E02 of the Recreational Science podcast (timecode 9:56):
Lu: In this study, scientists trained chickens to be attracted to average human faces and then asked these chickens what they thought of uglier or more beautiful faces. So, the subjects of this paper are six chickens, four female and two male.
Tirth: Okay.
Lu: The chickens were deprived of food for 12 hours. The scientists then showed them photos of average human female and male faces. If the chickens pecked at the correct face, they received a food reward.
Tirth: Wait, what does it mean an average face?
Lu: So, you’ve seen those faces where it’s the average face of a country?
Tirth: Yeah.
Lu: Basically, they took pictures of 35 faces, female or male, and averaged the pixels.
Tirth: And then what does it mean if the chicken pecks at the right face?
Lu: Great question, Tirth. Great question. So, these scientists, remember I said they have four female chickens and two male chickens?
Tirth: Yeah.
Lu: They trained the female chickens to preferentially peck at the male human face and the male chickens to preferentially peck at the female human face.
Tirth: Of course they did. They’re making a lot of assumptions about chicken sexual orientation, by the way.
Lu: Exactly. So let’s just pause there for a second and think about what that means exactly. These people care more about heteronormativity than conspecific relations.
Tirth: Yes.
Lu: They’re more concerned about homosexuality than bestiality. Okay? Let’s think about that.
Tirth: Yeah…
Lu: Okay, so if the female chickens peck at the average male face, they get a food reward. If they peck at the female face, they don’t get a reward.
Tirth: Okay.
Lu: Then they took these two average female and male faces and generated a set of seven faces that range from most feminine to most masculine. The first thing they did was average the average female and the average male face to generate an androgynous face.
Tirth: Okay.
Lu: And then they extrapolated from the average androgynous face to the average female face to make a more feminine face. And then again to make an even more feminine face (hyperfeminine face).
Tirth: Okay.
Lu: So, you have the average androgynous human face, average female face, more feminine female face, and hyperfeminine female face. They then did the same thing on the male side. So you have a total of seven faces ranging from most masculine to most feminine.
Tirth: Okay.
Lu: And they ask the chickens to peck at these faces to register which ones they preferred.
Tirth: Oh my god.
Lu: As control, as sort of a calibrator, they got 14 undergraduate students, human students, seven male, seven female, and asked them to rate these seven faces on a scale of 1 to 10 in attractiveness.
Tirth: This is basically a precursor to Facebook. This is how Facebook was started. It was hot or not, essentially.
Lu: Mhm. So I’ll tell you the human results first. How the human students rated the faces. The male and the female responses are actually pretty much the same. So for the sake of simplicity, I’ll just describe the female results. The female students found the three female faces as basically not at all attractive.
Tirth: Okay.
Lu: The androgynous face they found slightly more attractive than the three female faces, but a lot less attractive than the three male faces. The more masculine face was the most attractive and was slightly more attractive than the average masculine face and the hypermasculine face.
Tirth: Okay.
Lu: The female chickens pecked at these faces in the exact same frequency.
Tirth: No, they didn’t.
Lu: Yes, they did. The two curves, when graphed on top of one another, look exactly the same.
Tirth: Both the chickens and the humans prefer the same person, same face.
Lu: Exactly. So, the average male face for both of them is pretty attractive.
Tirth: Uh-huh.
Lu: The average androgynous human face is much less attractive, and the more masculine face is slightly more attractive. It’s the exact same pattern. Isn’t that crazy?
Tirth: It’s crazy. Well, it’s something.
Lu: It’s certainly something. What do you think that something is?
Tirth: I… nature. Human nature, I guess.
Lu: Human… or chicken nature.
Tirth: Chicken nature. Or just nature. Maybe this thing should have been in just Nature because it covers humans and chickens.
Lu: Could not agree more. You might be wondering now what’s the significance of this paper?
Tirth: I’ve been wondering that from the first, the title, really.
Lu: Well, this paper suggests that human preferences arise from like a general property of nervous systems, rather than like face-specific adaptations. At least that’s what they claim, which kind of makes sense.
Tirth: Right.
Lu: It suggests that humans didn’t evolve only to decode information from faces. There are a couple of theories on why we find certain faces more attractive. The first theory is the mate quality hypothesis.
Tirth: Okay.
Lu: The mate quality hypothesis is that humans receive a signal from the face about the health or the quality of the mate. And we’ve evolved/adapted to respond to that and that’s why we find certain people more attractive. The competing hypothesis is something called the bias hypothesis, which is that attractiveness may be due to biases that are inherent in the nervous system. And the bias hypothesis would predict that really any nervous system can develop preferences for human faces, which they showed here. Sort of.
Tirth: Okay.
Lu: What do you think of the paper overall?
Tirth: I liked it a lot.
Lu: A lot to think about.
Tirth: A lot to think about. I’m going to be pondering this over as I sleep tonight. You know, I’ll think about attraction and…
Lu: As you stare into the face of your lovely fiance, you’ll be wondering if I peck at her, will I get a food reward?
Tirth: If she was a chicken. You know, that’s the other thing they should have done, man. They should have taken these undergrads and they should have shown them pictures of chickens and see if they would have rated chickens… I mean, that would have been the perfect experiment.
Lu: Can you train humans to judge chicken faces… That’s actually a great idea. I’m not even kidding. That experiment absolutely should have been done.
Tirth: This has a lot of practical value, right? For example, if you’re a big chicken company, let’s say you’re Chick-fil-A or Tyson Farms or whatever, you want to pick the best chickens, right? So you can get the best eggs, you can get the best meat. There’s a lot of utility to this.
Lu: And you think more attractive chickens produce more attractive eggs, better tasting eggs?
Tirth: This is a one hypothesis. I mean, we’d have to test it. We don’t know.
Lu: We’ll test it and we’ll talk about it on this pod.
Article citation
Ghirlanda et al., 2002. Chickens prefer beautiful humans. Human Nature. https://link.springer.com/article/10.1007/s12110-002-1021-6
Leave a Reply