Why Men Won't Dance

All around the world, people dance. While it is true that not every individual person likes dancing, it is true that all cultures have dances. People dance largely because they enjoy it. There seems to be some mental reward, some joy to be had, from moving in a rhythmical and rather pointless way. That humans have an innate ability to enjoy dancing means that we presumably evolved this. The things that give us pleasure are not random. We enjoy good food, sex, success, hunting, warmth, company, and the like because these all help us to survive and reproduce. People in the past did not evolve to like sex because it gave them pleasure. They evolved to like sex because a liking for sex helped them pass on more genes. People who hated sex did not become our ancestors.

So, there must be some good reason why humans like to dance. I have a theory which I here put forward to explain why people dance. In short, I believe that women dance in order to encourage men to dance, and that they do this in order to learn about the men.

The Grandparents Observation

If you ask very old people how they met their spouses, the answer will often be that they met on the dance floor. In the time of our grandparents, there was no television, and one of the principle social gatherings was the dance. In Britain, even little villages had dances every week. If you ask a grandmother why she fell in love with her husband, the answer will often be that he was such a good dancer. Conversely, if you ask a grandfather, he is likely to say “Well, she was such a sweet little thing then. The prettiest girl in the room.” It is very rare for a man to fall in love with a woman just because she is a good dancer. I confess that if I saw two women dancing, one of whom danced like a goddess but looked like a hag, while the other looked like a goddess and danced like a limpet, I’d still probably go for the limpet.

The Night Club Observation

Go to a night club in Britain today, and the chances are that if you go early enough, you will see a dance floor populated entirely by women. These may be formed up in small circles, perhaps around piles of handbags, and they would be dancing, and enjoying dancing. They would be getting simple pleasure from the act of moving to the music. They will almost certainly be dancing in a manner which requires little skill. To an observer, they would not be putting out a message which said “Look how skilfully I can dance”. Instead, the message would be either simply “I am enjoying doing this,” or, if they are putting on a show for anyone looking on, it might be “Look what a great body I have.”

Some considerable while later, the floor gets some men on it. This only happens when the floor is so crowded that the men are able to hide on it, and/or when the men have had a great deal to drink, and are too sozzled to be able to dance with any skill.

If a man does get up early on, and walks onto the dance floor and starts dancing, then he risks looking a fool. If he dances on his own he looks odd. If he tries to break into one of the rings of women, then he will have a tough time of it. If he dances with enthusiasm and great skill then onlookers are likely to conclude that he is, to use British slang, “a wanker”, that is, someone who thinks that he is marvellous, but isn’t.

Something else one might observe is a rowdy group of males, dancing in a manner to show strength and solid group-inclusion. To the women these men are displaying that even though they can’t dance well, they are each accepted as one of the lads, none of them an outcast. A man can also dance badly, but knowingly badly, and with gusto and humour, to show that he is a man of good character. These are strategies used by men who are obliged to dance, but who have to compensate somehow for not being able to dance well.

Another observation that supports my theory is that homosexual men seem to use dance to select other homosexual men. Homosexual men have tastes in men similar to heterosexual men’s taste in women. They go for looks, youth, beauty, and physique. Perhaps homosexual men are using the same tactics as women. They go to clubs where it is obligatory to dance, and pick out the best dancers.

That there is a strong sex difference in this behaviour is clear to anyone visiting modern night clubs in Britain. My guess is that this goes a lot deeper than current cultural norms.

The Fighting Hypothesis

I noticed that at my local jive club, all the best male dancers did or had done martial arts. Patrick Swayze, the heart-throb star of the ultimate chick-flick Dirty Dancing is a karate black belt. Bruce Lee, hailed by many as the finest martial artist ever, was Hong Kong’s Cha-cha champion. There seems to be a link between male dancing and male fighting ability.

To dance well, one needs good health, poise, co-ordination, stamina, strength, athleticism, rhythm, balance, suppleness, speed, an ability to predict and react to the movements of others (especially for partnered dancing), and style. Interestingly enough, all these things are also what make a good fighter. Yes, even style is useful, for it intimidates opponents.

If a woman were to see a man dance, then she would have a good idea of how naturally talented at fighting he might be. One might argue that a better thing to do would be to watch the man fight. However, what opportunities might a woman ever have of seeing a man fight? Very few. She does not want to provoke him into fighting her, for that is bad for her health, and she doesn’t want to provoke him into fighting another man, because he might get injured, and she wouldn’t make herself popular. Furthermore, I suggest that the ideal man a woman would want for a husband, is a man who can fight very well, but who doesn’t fight. A man who can fight and does is a bully who is likely to end up getting jumped on by several other guys one day, and to wind up dead in a ditch. A man who cannot fight well is less able to protect his wife and children. Ideally, a woman wants a man who could, if the need arose, defend her and her children effectively, but who would not seek out fights without cause.

How then, can a woman find such a man? If he fights only in defence of his wife and children, then until he has these, he will never fight, and she will never see him fight. She could encourage him to dance. If she sees him dance, then she is able to judge his talent for fighting, and no one gets hurt. Some cultures have dances which are a lot like fights. Often a tribal dance re-enacts a mythological fight between some Great Spirit and a tribal ancestor. In Brazil, the art of Capoeira was developed, which was both simultaneously a dance and a martial art. Dances are public events. At one dance, all the women in the tribe could see how talented the men are at dancing, and thus judge their fighting ability. The only opportunity to see every man’s fighting ability directly would involve mass bloodshed, and would give a very inaccurate answer. Some men would be difficult to see in the fray, and some good fighters would be defeated by superior numbers or bad luck.

Dancing is a woman’s way of asking “How good are you at fighting?”

The Salsa Effect

I have done a fair bit of salsa and latin dance, but my real love is for the swing dances: jitterbug, jive, shim sham, shag, balboa, and Lindy hop. I have noticed one interesting difference between salsa and Lindy hop, which fits my theory rather well.

At the end of a salsa dance which goes well for me, I might get a favourable reaction from my partner. Often, this consists of her opening her eyes as if waking from a nice dream, and saying “That was nice.” She had closed them during the dance, and had been imagining that I was a young Robert Redford. While dancing with me, she had surrendered control of her body to my lead and her auto-pilot. She had contributed nothing to the choreography of the dance, and had instead been content to go wherever she was led. She was moving to the music, and the music was constant. She was a slave to the rhythm and to the lead. She had enjoyed the dance for the simple reason that she happens to get a kick out of moving to music. It is a sort of Bacchanalian revelry.

At the end of a good swing dance, things are very different. We both build up for the big finish, and then jump and land perfectly together and on the beat, each having predicted what the other would do. The reaction I see is that she looks at me with a look of joy and satisfaction, as if to say “We did it!” During the dance, she had contributed a great deal to the choreography, and the music had not been constant, but punctuated by pauses and complicated changes of rhythm. All the time she had been alert to what was going on, and improvising to suit.

I suppose that for modesty’s sake I should say that not all my dances go this well.

The relevance of this contrast is that I know many people who do both salsa and Lindy hop. Every single example of someone who prefers salsa, the dance, to Lindy hop, is female. Women seem to enjoy dancing for different reasons. Women seem to get an innate joy from moving to music, like those women dancing round their handbags at the night club. Men by contrast, seem to get joy from dancing well, but not nearly so much from the mere act of dancing. A good Lindy hopping man will therefore prefer Lindy to salsa. When asked why, the women who prefer salsa have all given me the same answer: “It’s a more sensual dance.”

Some women with whom I have corresponded on this topic have reported feeling a “romantic rush” once they have been dancing for a while. It seems that dance can put women “in the mood”, so to speak. Any good male dancers in their vicinity at this point would be in the best position to take advantage of this.

So Why Don’t Men Like To Dance?

If what I have written is all true, then surely men would love dancing, as it is an opportunity to show off to the ladies, and attract mates. No. In any population of men, half the men will be above average at dancing, half will be below average. There is nothing to be gained from advertising to the world that one is below average. Half of all men will shy away from dancing, therefore. Men who are only slightly better than average will risk dancing badly on a bad day, and might not have practised or learned enough of the current cultural dance forms, and so will also be very reluctant to dance. If all the below average men danced not at all, then the worst of the above average would be the worst dancers on the dance floor. It is only the very good dancers who will benefit from dancing. They will enjoy dancing, because they have evolved to get a kick out of dancing well. Men will have inherited from their male ancestors a terror of dancing badly.

One aspect of fighting I have not mentioned is bravery. One might argue that it takes little bravery for a man to get up and dance. Very few men have died dancing, whereas many have died fighting. Men often use Dutch courage (alcohol) to work up the nerve to dance, just as if they were working up the nerve to fight. Men do risk a great deal too by dancing. In a fight a man risks losing his ability to pass on genes. This is the thing of which evolution has made us most terrified. If a man dies, he passes on no more genes. If a man fights badly, he might die. If a man dances badly, he will show the women that he is not a good catch, and so he will pass on few genes. Bravery is the over-coming of fear. Fear evolves to stop us from doing things which might harm our ability to pass on genes. For a man, therefore, dancing does require courage. For a woman it requires almost none. A woman has little to lose by dancing badly.

In modern Britain, and in similar countries, dancing is no longer the standard courtship. Very few men have taken dance lessons. A teenage boy with a reputation for toughness, but no skill at dancing, would have the greatest vested interest of all to tell people that only sissies dance. Perhaps unsurprisingly, such people are indeed those who are the most vocal in their scorn for dance. If people believed otherwise, then these tough boys would lose out to their dance-educated rivals.

Dance Education

Posh women dance with refined etiquette, not wild abandon. I suggest that a woman cannot make herself attractive merely by how well she dances. She can, however, put herself into the world of high status men, making herself an appropriate match for a well-connected man, by learning to dance in the current correct manner. A rich man might send his daughter to learn ballet, or whatever is the current dance which only people who can afford the best teachers can learn, and she can learn to dance within the strict confines of current etiquette.

Conversely, if a woman wants to make herself sexy, then she could learn to dance with extreme sexiness. Such vulgar gyrations would not win the approval of the refined girls learning ballet, but they wouldn’t be complained about very much by the local men. While the upper classes might go to ballets or the equivalents, the majority of women professional dancers would be employed dancing the sexier stuff. I have not made a thorough study of the history of this, but I suspect that I am guessing rightly when I say that the past matches my theory.

For men, the situation is different. A rich man might do well to send his son to learn to dance. In a world in which dances are common, men will often find themselves on display at such important social gatherings, and could do themselves a great favour by learning the latest dances. Especially in a monogamous society, rich men will spread more genes by seeing that their sons can dance well. Rich men have the opportunity to spread many genes, and so should do all they can take advantage of this. Having a few sons who dance well is a good start. It was part of a European nobleman’s education until very recently to learn to dance.

At the opposite end of the social scale, very low status men might, if very talented at dancing, use this to find a way out of the ghetto. This seems to fit observable fact. Many modern dances started in the poor areas, amongst young men. Charleston was supposedly first danced by black dock workers. Twentieth century dances from the birth of jazz to high-tech hip-hop started amongst dedicated and talented young men in poor places. These men can win great respect in their locales, but very few become truly high-status through dance alone.

The Smell Hypothesis

Recent studies have shown that women are very sensitive to the smell of men. Sub-consciously, it seems, women can tell a good-quality man from a low-quality one, and their decisions are influenced by smells. One such decision is how highly they rate the attractiveness of a photograph of a man, while smelling the smell either of a healthy attractive man, or an unhealthy unattractive one. Thornhill and Gangestad (1999) showed that women prefer the smell of handsome faces and symmetrical male bodies (but not when on the pill), and that the effect of women’s smell on men is much lower. Women were most smell fussy at fertile peak of their menstrual cycle (2). When a woman goes out dancing, she will judge the men, subconsciously at least, by smelling them. Since men don’t rate by smell much, they can live without the opportunity to smell all the women. Most modern women at a dance hide their smell with perfume anyway.

However, in order to use this skill, the women would have to get a good smell of all the men in their tribe. This is not always possible or convenient, nor easy to do politely. Also, the smell of old sweat is often bad, and not so informative. Ideally, a woman would get to smell all the men under the same conditions, and would get to smell fresh sweat on all those men.

One way a woman might get to smell fresh sweat on all the local men, would be to play vigorous sports with them. While this would be less dangerous than fighting them all, it would have its hazards, and most men can run rings round most women, and so this is unlikely to work very well. Since the purpose of discerning which men are good and which bad is to decide which men to have sex with, getting the men sweaty by copulating with them is a very bad idea. By the time he’s broken sweat, it might well be too late. Dancing offers a woman not just an ideal opportunity to smell fresh sweat on all the men of her tribe, it probably offers her the only opportunity for this.

The supposed reason for under-arm hair is for increasing the ability for others to smell one’s sweat. It is interesting to notice that a movement common to almost all partner dances is for the man to turn the women under his raised arm. In passing, the woman’s nose wafts past the man’s armpit every time.

The Antiquity of Dance

Dance is not a modern creation. An archaeologist, Dr. Yosef Garfinkel of Hebrew University in Jerusalem has established what he says is an illustrated record of dancing from 9,000 to 5,000 years ago. This comes from looking at finds from the Balkans and Middle East, which depict dancers, and it coincides with a period in which farming was taking over from hunter-gathering. Nearly all the depictions showed dancers in groups. Garfinkel’s conclusions, however, are not those which an evolutionary biologist would devise. He writes that dancing developed as a means of social bonding and communication in pre-state agricultural societies. He tells us that dance was used to co-ordinate a community’s activities, such as harvesting. I think that this is wrong.

For a start, modern hunter-gatherer societies dance. If dancing dated back to the start of agriculture, why do hunter-gatherers dance? Also, we humans enjoy dancing. This is a human universal, just like a love of music is. Evolution is unlikely to have created this human trait all round the world, if dance first started in the Middle East a few thousand years ago. What’s more, theories which attempt to explain evolved human traits by showing what is good for groups, rather than individuals, need to believe in what is known as “group selection”, which today is rather out of favour with evolutionists. Also, there are many other forms of ritual and of communication which could serve the same end. Why would all societies solve the problem the same way? Why mightn’t some sing, or chant, or talk, or recite poems, or build buildings, or do some other group ritual for the same purpose? Indeed, if dance served this purpose, why did societies develop other rituals for the same function?

I see no reason to believe that dancing is not very ancient indeed. I am confident that hunter-gatherers of the Pleistocene danced. Dancing may well be more ancient than proper spoken languages. Dance can be seen as a virtuoso extension of body language. A man who could gesture with style and grace might win more of the chicks that a man who just flailed his arms about(1).

Partnered Dance

A lot of dance is done in couples. This is not universal, and some see it as possibly a recent European innovation. Nevertheless, I shall write a few words about it. Much of this section is still relevant to other types of dance.

In partnered dances, the man leads and the woman follows. The lead part is more difficult, requires greater strength, and it helps if the leader is larger than the follower. To dance well in unison, the man has to look after his partner. Women do not enjoy men who seek to prove how strong they are by leading with unnecessary force. A good leader leads with enough force and no more, and leads the moves his partner enjoys. It accords with records of female sexual fantasies that they would enjoy dancing with a man who is tall, strong, gentle, and looks after them. By dancing with a man, a woman is in the perfect position to judge both how physically talented the man is at dancing, and how well he treats women. A man who is determined to dance impressively, may ignore the wishes and reactions of his partner, and execute an extra-ordinary series of whirls and jumps, which jars his partner’s shoulder painfully.

Partnered social dance is improvised, and so a man cannot disguise a lack of dance ability by practising a set routine over and over again. This means that it gives a woman a more accurate picture of his true talents. Partnered dances often mimic romance, with the woman testing the bond with feigned flight, and the man having to show an ability to dominate without cruelty. Partnered dances are often described as vertical expressions of horizontal desire. A woman can try out her partner’s touch, to see if “chemical electricity” flows. To some degree this may tell her how compatible a sexual partner he might be.

An evening of social dance is not overtly competitive. If men knew that they were all being judged, then they may be affected in a hundred ways that would make them act unnaturally, which would be bad for the women, who want an accurate picture of the men.

A woman might be motivated to become a good dancer, so that she could get to dance with the best men. Only when she has danced with a man can she get a true picture of how good he is. Interestingly enough, the women at social dances, in my experience, have quite a different view from the men of which men are the best dancers. I might see a man using many complicated moves, and point him out to a woman, who says that his lead is very weak, or wildly too strong, or that he knows some fancy moves, but does them over and over again to the point of tedium. The women often refer to the “nice” dancers rather than the “good” dancers. They can only find this out by dancing with the men, or by talking to those who have.

Some years ago, I was with about 600 people attending a salsa dance. Half of them were people who attended dances often, and who had taken lessons in dance, and the other half were people who were there because it was a big party with a beer promotion. The people who were used to going to dances were also used to the culture and etiquette of such dances. They were asking each other for dances, and accepting such requests (it is bad form to refuse). I went up to a girl to ask her for a dance. She, and her boyfriend next to her, were not from the dance-going culture. She pointedly looked away from me in a definite "I heard you but I'm pretending not to have heard you. Do not ask again" - sort of a way. Her boyfriend shot a murderous look at me. I backed off and asked someone else.

The simple fact that I had dared to ask the girl for a dance presumably suggested that I was a competent dancer. If she had accepted, then she would have learned that I knew all of three salsa moves.

This incident got me thinking. The boy's reaction was perfectly rational. If he judged me to be a better dancer than himself, then he was right in being hostile, as I was a threat to his possession of a mate (my intentions were honourable of course, but I take it that the Darwinists reading this will understand what I mean). Understanding this, and not being part of the dance culture (and possibly finding me repulsive, but let's skip round that), the girl was acting rationally in signalling to her boyfriend her loyalty to him. Shortly afterwards, I devised this dance theory.

Birds and spiders often dance as part of their courtship, and are presumably judged by how well they perform. In salticid spiders – the Fred Astaire of spiders - when another male tries to get into the act, they break off the dance, take up an aggressive stance, and display fangs. I can’t help feeling that this is closely analogous to what happened to me at that party.

A Recent Study

I posted this theory of mine to the Human Behaviour and Evolution group on the first of June 1998. One member of that group at Stirling University, suggested to one of his psychology students that she carry out a study to test my theory.

She had two parts to her study. The first was a questionnaire in which people were asked to rank how important various characteristics of a potential husband were. She had sixteen characteristics. A willingness to stand up for self/other came out at number 1. Physical attractiveness came out at number 2. Other results were: good health 3, creative 4, high earning capacity 7, tall 9, good athlete 13. Good dancer came at number fifteen, just below “owns own car” at 14. The only thing that dance was rated higher than, was “good footballer” which came last at 16.

The above is far from fatal to my theory. For a start, it shows that women might prefer good dancers to good footballers, which alone is reason for football-loving men to learn to dance. Also, people do not admit to, nor even are they consciously aware of, their true desires. That women marry uncreative and unhealthy men who are very rich is an observable fact, and contrary to the ratings given in this study. Few women, though, would admit to rate wealth above many other things, and few really understand the magic that makes rich men seem so much more handsome and charming than poor ones. Almost no women would rate the smell of a man’s sweat very highly, even though experiments show that it is a definite factor. Also, if my theory is correct, then women are judging indirectly such things as athleticism, and creativity, which rate higher than dance, by watching men dance. The fact that men dance would allow them to display their physical prowess and creativity, even if the dance itself were unimportant.

The second part of her study involved getting women to rate the attractiveness of men after showing the women videos of men involved in dancing and other forms of exercise. She didn’t make her own video tapes, and only had a sample of six men on tape. Five of the six men were rated as being more attractive as potential dates when dancing, and a couple of them very significantly more. Unfortunately, the sample size was so small, that nothing very significant was shown. To do this sort of study thoroughly, one would need videos of many men, some dancing well, and some dancing badly, for contrast. Those same men could vary in other ways, such as in age, height, wealth (shown by expensive clothes), and the like.

The student tried another experiment, which involved asking people about their own relationships, and seeing if the ratings of their partners went up or down when the subjects were watching dancers. No effects showed up. It seems that the noise caused by people being either in serious relationships or not, affected their ratings more than exposure to dance.

This was a very minor study, and far more would have to be done to test my theory properly.


I believe that my theory is strong. I shall make some predictions, all of them testable.

There will be no society anywhere on Earth where all dancing is done in secret. Dancing will be a public phenomenon everywhere. Whereas a man or woman might practice alone, the end product will always involve witnesses.

There will be no society on Earth where all the dancing done by men is away from the eyes of women.

Women, far more than men, will find the skill with which a potential mate can dance far more of a factor which influences their choices of mate. This will be true of all societies.

Women will get the most pleasure from dancing (except perhaps when taking the contraceptive pill) when they are at the fertile peak of their menstrual cycle.

It is commonly noted that in many primitive tribal societies, dances by men are often endurance events, and the men dancing often dance in public for a very long time, and by this method display their fitness. I predict that in societies where this is the primary form of dance, that dancing will not be something which is taught formally, and that the dances will be simple. In societies where the dances are creative, varying, requiring skill, and very vigorous, men will be far more willing to learn and practice dances, and there will be men who teach. The men who dance the endurance dances will be using dance to display stamina, partly because they are unable to use it to show much else.

Men who are good dancers will have more affairs than men who are bad dancers. If this single prediction is well-supported, then this will form a strong pillar for my theory. Anything which improves a man’s chances of passing on genes, is likely to end up forming part of innate human nature.

Women will rate the likely future marital fidelity of men who dance well as lower than that of men who dance badly. One lady with whom I corresponded on this theory told me that her mother had forbidden her to marry a good dancer, because such a man would “be trouble”. The instincts of women are finely honed to be watchful of the straying attentions of men. If women rate good dancers as sexy, then they will also rate good dancers as more able to have affairs.

Men who consider themselves to be bad dancers will be less likely to dance. When asked to dance by women, the chemicals released in their bodies will be similar to or the same as those associated with fear. They will not just be choosing not to dance because it is not their cup of tea, actual fear will be part of the reason for their refusal.

There will be differences between the traditions of dance in monogamous and polygynous societies. Monogamous societies would be ones in which women would seek to find out the best dancers, for affairs, or for guarding against affairs. Dance would be a way to counteract the drawbacks to women of monogamy. Men, deprived of the opportunity to have more than one wife, would, if good dancers, be more keen on partner dancing. In such societies, I would expect almost all women to be interested to some degree in dancing, and for only the minority of men to be interested in dancing. I would expect them to be more likely to have dances in which men and women partner each other. I would expect men who are less good dancers to be particularly scathing about men who dance.

In polygynous societies, the reverse might be true. Dance might be used by men to counteract the drawbacks to men of polygyny. Men might conspire to lessen the advantage of the best dancers. Most men will be interested in dancing, and men will commonly dance in groups together. The dances will be less creative, and more formalised, with set choreography which allow all the men to dance in unison. Mixed-sex partnered dancing will be uncommon, because this would give the women the most information on individual men. The women will still be able to judge the men to some extent, because they will see which men can remember the moves best (like the notes of a male songbird's song), and which dances them with the greatest style. The very highest status women will not encourage dancing, because such women will already have husbands, and the attentions of those husbands is wayward enough as it is. For women, dancing will be more popular amongst those of lower social status.


Women have an innate liking for dance. In the past, the women who danced learned more about the men around them, and so passed on more genes, and created a species in which the females have an innate liking for dance. Some things, the women could already see: they knew which men were oldest, which tallest, and which most muscular, but by getting the men to dance, they could learn a great deal more. All information is good. Finding out which men were bad dancers was just as important as finding out which men were good. If a woman gets a man to dance just once in his life, then she has perhaps learned that the reason he doesn’t dance, is that he has two left feet and no rhythm. These women did not have to marry the good dancers, but they might use their knowledge to get the best sperm by having an affair. For most women, the evolutionary jackpot is getting the best sperm in the district, and then fooling some poor sap into helping her bring the resultant child up, thinking it to be his own. Dance helped women achieve this jackpot.

A dance is a bit like a lek. In a lekking species such as grouse, all the males display to the females, perhaps doing a sort of dance, because this is the only way to get selected by the females. Very few males win mates in a lek, and most males go lonely. Humans are not perfectly monogamous, and do go in for lekking on a lesser scale. Stronger more talented lekkers would benefit from lekking, but for most men, leks are a great risk, or even a certain disaster. Since there are other ways to win a woman than lekking, for most men, it is best to avoid comparing oneself with rival mates on the dance floor. Women do not rate dance ability above all else. It is but one of many factors. They also like intelligence, kindness, a handsome face, and of course a fat wallet. Every little helps, though, and evolution will favour any little advantage that one creature can get over another.

It is well recognised that women are more choosy about sexual partners. Natural selection has armed women with dance to improve their powers of choice. Natural selection has also equipped men with a useful fear of this weapon. That’s why men won’t dance.


(1) See Donald, Merlin Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition. Cambridge, MA: Harvard U Press, 1991.

(2) Thornhill, Randy and Gangestad, Steven W. (May 1999) The Scent of Symmetry: A Human Sex Pheromone that Signals Fitness? In Evolution and Human Behaviour, Vol 20(3): 175-202.