A scientific basis for "gaydar"?


Gaydar--the supposed ability to identify a person's sexual preference just by looking at them--may be more than just a bad pun. A couple of studies have been done on this recently, as part of broader research into the general ability of humans to very quickly make correct judgments about other humans' skills, personalities, age, etc.Turns out, people asked to guess whether a man is gay or a woman is a lesbian, just by looking at the subject's face, got it right at a rate higher than that which could be explained by mere chance. And the trick worked even when the images were cropped to remove all trace of tip-offs like hair style or fashion accessories.

So far, the studies have been very small and they certainly don't show that gaydar, if it does exist, is foolproof. The guessers did statistically significantly better than chance, but they weren't exactly sexual-orientation sniffing bloodhounds. There's also some open questions about whether the gender or sexual orientation of the guesser makes a difference on the rate of accuracy. What the research does do is add to considerable body of evidence showing that humans are evolutionarily programmed to pay extremely close attention to the facial features of other humans, and it tells us that we still have a lot to learn about what that programming means for ideas like "instinct".

Cognitive Daily: People Identify the Sexual Orientation of Strangers in As Fast as 50 Miliseconds
Miller-McCune: Subjects Correctly Identify Lesbians Solely on Facial Features


  1. I always thought “gaydar” meant being able to accurately guess a person’s sexuality based on their overall appearance, voice and mannerisms- not from a single photo of a face alone.

  2. Learnt about a study in psych class (sorry, don’t have the reference on me) that showed participants short video footage of gay and hetero people talking to the video. Heads were cut out of the shot, so just the torso and arms were seen. I believe the clips were around 10 seconds long. People guessed the sexual orientation of the speakers very accurately (again can’t quote exact results), which was based on only a small sample of the person’ upper body movements.

  3. Me too Brainspore. And I haven’t heard of straight people having it (except ones who’ve hung around with gay guys a lot).

    My own gaydar no longer works on guys under 25. They all seem gay to me. Fortunately I realized this before relying on it!

    1. Interesting, Xopher… maybe “gaydar” is more of a measure of how attracted we are to the subject…

      I agree that this calls for another study that looks at full impressions, not merely face pics.

  4. Of course people got the “right” answer. It’s a 50/50 chance of guessing correctly. On top of that, humans evolved an innate ability to pick a suitable partner based upon looks, clothing, scent, pheromones, etc… The basic idea being, you can’t mate with someone not interested in you, so you learn to pick up on subtle cues subconsciously.

    1. The point here is that people got significantly better than 50/50 guessing. That means there was a systematic difference not likely to be due to chance alone. And saying that “of course” something happens does not cut it for science, hence why we actually test things rather than just accept that what we reckon must actually be the truth.

      1. No, I was not trying to say that all the students tested “guessed” their way to the correct answers, I was implying that guessing was part of it. Their own personal experience was the other part influencing their choices. According to the first link, the results were 57%, just barely above chance then they qualify it with this:

        “When the results were corrected using signal detection analysis (to compensate for the fact that fewer than 50 percent of men are gay in real life), accuracy was 62 percent at 50 milliseconds, and as high as 70 percent when self-paced.”

        So given unlimited time, and using “corrected” results you have 70% accuracy. using the same “signal detection analysis” you could also “correct” the 50/50 guesses to a higher than 50% accuracy rate.

        So no, I don’t think you can boil “gaydar” down to an exact science any more than you can boil down your instincts to an exact science.

        1. I know only too well that we cannot boil down instincts to an exact science. I am doing a PhD in the social sciences ffs! That is not the point. The point is that we should try to do these things, not give up because we think they are too hard to measure and rely instead on what we reckon.

          Your statement “Of course people got the “right” answer” is not valid in a scientific context, even if we are talking about social sciences.

          1. I never intended for my initial statement to be anything near scientific, just merely trying to illustrate the ludicrousness of defining “gaydar” as a science. The study does illustrate that, as several posters have already mentioned, there is more to being homosexual than a “choice” in gender preference. Its more than getting too much or not enough of something as a kid.

          2. Science is a method of inquiry that can be applied to just about any question. Whether or not the “gaydar” is a real phenomenon can be addressed scientifically just like anything else.

    2. Of course people got the “right” answer. It’s a 50/50 chance of guessing correctly.

      Given that only around 10% of people are homosexual, doesn’t that complicate the odds somewhat?

      And given that I studied stats only a couple of years ago, I’m disappointed I can’t remember how to get my head around such a basic problem : (

      1. Well as an impartial body, if I was running this test, I would get a 50/50 split of gay/straight photo subjects then show the photos to the students to see if they could discern the difference. The article makes no mention of a control group, say a group shown all photos of either gay or straight people to create a “baseline.” I can only assume a control group was included in the study, as any good scientist would include one.

        So as I was telling apoxia, my reasoning for say that was that between guessing and peoples innate ability to find a suitable mate that the results of the study would be easy to get. I would be interested to know where you got your figure of 10% for the ratio of homosexuals in the populace, I would think it would be higher than that.

        I would like to see this study repeated, one group of straight subjects, one group of gay, one 50/50 mix group, then split them up in the same groupings by gender and repeat the same photo experiment. It would be interesting to know if one gender could be better at identifying a compatible partner better than the other. If say, women were better at identifying homosexuals than men, then you could say it might have something to do with female physiology that gives them this advantage over men. Often it is said that homosexual men bear traits similar to those of women, so it could be said that it might be hormonal in aiding homosexuals to identity each other, thus aiding their ability to find a compatible partner.

        But having not performed any of these studies myself, most of this is just speculation on my part.

  5. My own anecdotal experience is that gaydar is a real phenomenon but doesn’t work across cultural or class boundaries very well. We simply learn to pick up subtle, possibly subconscious, cues from the people around us.

    It also suffers from a serious amount of confirmation bias: you remember when your gaydar was accurate but forget when it wasn’t… unless, of course, you act on your miscalibrated gaydar and the guy bashes your head in with a baseball bat when you ask him out on a date.

    In that case you remember.


  6. This kind of reminds me of a Neil Gaiman short story named “Changes”, where a miracle drug invented to cure cancer inadvertently causes people to “reboot” into a different gender. After it’s use becomes predominately recreational (At least in western culture) the story explores what effects this would have on society. One example was a sort of night club bouncer who could sense what your born gender was, and how this was likely a skill humans had but was unnoticed due to it’s lack of relevance. Just a story, a fantasy, but it raises an interesting idea.

    Maybe because homosexuality is becoming more prevalent in modern culture, at least in that people are more open about it and it’s featured in entertainment in a different light then years past, we will begin to notice the difference like we already do between men and women?

    Maybe not, but who knows right?

  7. “Gaydar–the supposed ability to identify a person’s sexual preference just by looking at them…”


    /Sorry to be “that guy”, but my friends in “the family” always corrected me when I made that mistake

  8. “Gadar” is a relic from the transitory time between when gays had to hide it like they were secret pagans in the middle ages to the modern time when they can do a “Mr. Garisson and Mr. Slave” act in public and people are conditioned to believe that they are scum Nazis if they are disgusted by them. It was just a joke about telling signs so you wouldn’t waste a friday/saturday night trying to pick up/attract someone not interested in you.

    Judy Brown, in that brief 90s boom of good shows did a wonderful “Women with no Gaydar” skit. Might look for it on YouTube.

  9. isn’t the more interesting implication of this that, if it’s confirmed, we could infer that there are physical genetic markers of homosexuality?

    If you can make a better than chance guess as to someone’s sexuality just by looking at their face, then there is probably some underlying genetics that express themselves in people’s physical features. In other words, another drop in the “homosexuality is genetic” bucket.

  10. It reminds me of Raphael Carter’s short SF story “‘Congenital Agenesis of Gender Ideation’ by K.N. Sirsi and Sandra Botkin”. Written in the form of a scientific paper, it describes what first seem to be people incapable of perceiving gender differences, but who later turn out to be so finely tuned in to gender that they actually perceive over 20 different categories, and are incapable of mapping these categories onto the general culture’s “male” and “female”.

  11. I read an article a few years ago about gay voice, and whether it exists. What I remember from it was that the conclusion was that gay voice doesn’t exist, because not all gay men talk with gay voice, which made me incredibly angry, because the test shouldn’t have been “Do all gay men talk with a gay voice?” but rather “Are all men that talk with a gay voice gay?”

    If a man is efeminate, I’m going to think he’s gay until told otherwise.

    1. I’m pretty sure people are born with homosexual tendencies, but not with the ‘gay’ voice. That’s something a select minority of gay people pick up along the way, maybe to identify with peers, or to show pride. Most gay guys aren’t particularly effeminate. This is, overall, a complex subject though, and these interesting studies needs more support.. maybe they do have validity but jumping to a conclusion from a couple studies is bad science..

      1. I don’t know about your theory of “gay voice” being acquired out of a sense of identification. The people I knew who grew up to be effeminate gay men had distinctive speech patterns before they were even old enough to know what “gay” was.

        David Sedaris had a great story about being forced to go to a speech therapist in elementary school because of a lisp, and how the sign on the door might as well have read “Future Homosexuals of America.”

        Any effeminate guys care to weigh in on that?

      2. I’m pretty sure people are born with homosexual tendencies, but not with the ‘gay’ voice.

        I’m pretty sure that everyone is born with the gay voice and straight boys just get it knocked out of them.

      3. I agree. I think it’s cultural, an “in-group” dialect and accent. While some guys talk that way all the time, most code-switch, as you say. While I’m not the butchest thing on two legs even now, I went through a brief period in early college when I was very fem. My voice, my word choices, my walk, my…well, other stuff. I picked it up from the older gay men I was hero-worshipping, and my peers in the bars.

        I got over that, and now I’m much butcher than I was. Yes straight people, ‘butcher’ can be an adjective. Partly this is because I decided being butch was sexier (and I needed all I could get in that department!) and partly because I just didn’t hang out with that many gay people any more.

    2. Then you might be interested in these studies:

      The influence of sexual orientation on vowel production: Munson, Benjamin, Pierrehumbert, J. & Bent, T. & Bradlow, A. & Bailey, J., Journal of the Acoustical Society of America, 2004.

      The acoustic correlates of perceived sexual orientation, perceived masculinity, and perceived femininity: Munson, Benjamin, Language and Speech, 2007 .

  12. A side note – a very interesting topic for a post would be studies on ‘code switching’ – a phenomenon that happens in all cultures, but is really noticeable in subcultures such as Gay or Urban African American, etc. People tend to talk differently around their ‘own’ than around everyone else. Okay, someone post about that.

  13. greengestalt – “Gadar” is a relic from the transitory time between when gays had to hide it like they were secret pagans in the middle ages to the modern time when they can do a “Mr. Garisson and Mr. Slave” act in public and people are conditioned to believe that they are scum Nazis if they are disgusted by them.

    Funny, I don’t see gay bashing going down in terms of crime statistics, nor do I see men decked out in BDSM gear outside of pride parades or kink events. Are you living outside of some leather store in the Castro?

  14. This study seems to have been done using static stimuli. Research has pointed to humans being much better at recognizing dynamic stimuli, such as videos. In a study where motion capture was used (so there would be no identifiable artefacts) recognizing male/female based on movement proved trivially easy. Even when trying to simulate the motion of the opposite sex it turned out they were easy to spot.

    We’re just really good at identifying patterns, especially patterns with a time component. Anything out of the norm is picked up really fast. Why would sexual orientation not affect the body/movements etc in some way?

  15. I’m pretty sure people are born with a voice that can only go “ggu ba ba ba”, before they learn more from their parents, family, and friends. Someone will have to clarify if those syllables are gay or not, because I can’t tell.

  16. I don’t know if that proves the phenomenon of “gay-dar” so much as it might prove the phenomenon of “gayface” (in the interest of disclosure, I myself am gay (though I don’t know if I’d say I have gayface)).

  17. I have to second bbonyx. I don’t like word police, but the difference between “preference” vs. “orientation,” actually has consequences.

    The question of how essential effeminacy is in the gay men who exhibit it has always interested me, but I’ve never come to any strong position except that there seems to be both environmental and innate determinants. Like other commenters, I’ve certainly known guys who only flamed out when they were around other gay people, but we’ve all also probably known guys who were super femme from the time they were little boys.

    As for physical markers, I always thought one of the most interesting was otoacoustic emissions in lesbians. Your ears actually make very faint sounds on their own, and apparently you can predict if someone’s a lesbian by going on only the frequency/patterning of their otoacoustic emissions. Baffling.

  18. Kind of off-topic, but I couldn’t help thinking about this nonetheless:

    I recently had a date with a fellow who informed me that he (a straight man) had been on what turned out to be four dates with a gay man before he found out that they weren’t just “hanging out” at those hockey games (he found out the orientation of his “buddy” when the buddy informed him that he didn’t feel “the relationship [was] going anywhere”).

    Apparently, there are at least two people in Dallas – one homosexual, the other heterosexual – who have wildly malfunctioning gaydars…

    1. That sounds almost like a sitcom plot…except that if you wrote it as one, no one would believe that a gay man could be so clueless (the cluelessness of straight men is legendary). So your date thought it was just a “bromance,” huh?

      That’s a hilarious story.

  19. Gaydar is not detecting deviance in sexual orientation, it is detecting deviance in gender expression, e.g. nellie males and butch females. Visible deviance in gender expression may be a good predictor of deviance in sexual orientation; however, most people who are bisexual or homosexual are not observably deviant in their gender expression. People who conflate these things seriously misunderstand both gender expression and sexual orientation, preferences and practices. Human beings are MUCH more diverse than they appear, and most interesting deviance is not casually apparent.

  20. “Deviance” takes on some very negative connotations. “Variance” would be more appropriate, and not carry those negative connotations.

  21. What the research does do is add to considerable body of evidence showing that humans are evolutionarily programmed to…

    I doubt very much that the results of any of these studies say anything at all about evolution. You could come up with a pretty compelling story about how we pay attention to faces because we see so many of them and they are so important for so many aspects of our lives. To th extent that evolution plays a role, it just needs to provide the hardware capable of learning that faces (or whatever) are worth paying close attention to.

  22. On the topic of ‘gay’ voice: I am queer, I am not effeminate at all (either in looks or actions), and I also have a high voice with the breathy ‘s’ lisp sound.

    Now given that a large part of the sound of your voice comes from the length and thickness of your vocal chords, and that both those factors are heavily effected by hormones, I don’t think it is a stretch to suppose that some gay men have ‘gay voices’ for primarily biological reasons rather than sociological or cultural ones.

    I consciously work to suppress what I consider to be my natural voice, because you aren’t taken as seriously with it as without it (ie. people assume that you are a ditzy or bitchy queen the minute they hear it. Or worse, that you are going to be their new best ‘girlfriend’ – ugh!). I like to use my authority voice (shorter words and sentences, lower, flatter, slower and louder) when I’m trying to get people to do things.

Comments are closed.