Which of these images is the most attractive?

Figure 01


1. Cross-cultural evidence
2. From the woman's point of view
3. From the man's point of view
4. Fantasies as evidence
5. Waist-hip ratios and physical attractiveness
6. Assessment Questions

Beauty is claimed to be in the eye of the beholder. And it could also be held to be very variable across cultures and history. Compare an image of womanhood by the painter Peter Paul Rubens from 1613, Jupiter and Callisto, and a contemporary photographic image of the 'supermodel' Claudia Schiffer.  One of these tends to be picked as 'more attractive' by contemporary white, western males than the other. And then there are likely to be all kinds of cultural differences in the choices made, aren't there? And physical attractiveness rates highly for men when asked to judge the attractiveness of women, as opposed to their personality characteristics, according to the social psychological literature.

Go and take a look at a typical experiment in this paradigm. There is one being run online at Faceprints by Victor Johnson of New Mexico State University; and another by Ronald Henss at the Universität des Saarlandes in Germany.

By contrast, the social psychological literature on what women find attractive in men consistently puts their physical attractiveness below their personality characteristics and status (see, for a compilation, Gould, J.L. and Gould, C.G. (1989) Sexual selection. New York: Scientific American Library). The one physical character in men that has traditionally been given a high rating by women is height: tall men are universally selected as more attractive than short men. Recently, research has begun to show that facial and bodily symmetry is also an attractive feature in males, and an aspect of 'beauty'.

The relation between symmetry and what we find attractive is really quite remarkable in evolutionary terms. There are two sides to this. First, many biological forms are built up of parts that relate to each other on what is termed 'the Golden Ratio': 1 to 0.618. This relationship is so ubiquitous that it appears to represent a kind of plan of nature. That plan has to be executed in development everytime a new organism matures, and many events in nature can upset the exact nature of the realisation of the plan. Departures from the plan thus act as an index of how susceptible any organism has been to such disruption, and give an indication of how robust that organism is to developmental perturbance. This might act as an indicator of the organism's 'quality', and it would thus make evolutionary sense to find symmetric and proportioned individuals attractive to us, would it not?

Similarly, birds called swallows have forked tails. Females find males with symmetrical tails more attractive than those with asymmetric tails, which leads us onto, second, the notion of 'handicapping' as an indicator of evolutionary fitness, because these females prefer mates with longer tails. Why 'handicapping'? Because if, say, a male peacock can grow a 'perfect' tail, he must be a pretty fit individual, and more, if he can lug the thing around and keep it and himself in good condition, then he really must be a fit individual, and thus a worthy potential mate. Make life more difficult; show you can cope; demonstrate fitness.(For more details than you probably want on this, go check this lecture at napier University, UK)

The differences in the way in which men and women value the physical attractiveness of the opposite sex is supported by many 'pop' psychology notions. Men are visually stimulated; women are romantically inclined: the kind of advice found in magazines like 'Cosmopolitan' is predicated on this assumption. And pornography is clearly directed at the male 'gaze'.

Let's get back to humans. There is some evidence that men and women worldwide have different 'mate preferences' from each other. A pioneering study is that of David Buss and his associates.

1. Cross-cultural evidence

A major study by David Buss and his colleagues (Buss, D. M., et al. (1989) Sex differences in mate preferences: Evolutionary hypotheses tested in 37 different cultures. Behavioral and Brain Sciences 12: 1-49) provides an overview of human preferences from 10,047 individuals aged between 14 and 70 living in 37 different modern cultures. The following brief summary of this research is taken from The Times Higher Education Supplement, June 2, 1995, p. 15:

The basic findings of the survey can be grouped in three distinct "clusters". The first cluster covers universal desires which show no sex differences. The key finding here is that everyone wants a mate who is intelligent, kind, healthy, dependable and a relationship where there is love or mutual attraction.

The second set of results deals with desires that are universally sex differentiated. Men place a greater emphasis on physical attractiveness or good looks. They also universally desire women who are younger than themselves. The size of the preferred age difference varies, however. In largely polygynous cultures, men preferred women who were dramatically younger than they were by seven or eight years. But in Britain, Norway and Sweden, for example, the ideal age difference was less. In all cultures though, men in general deslre younger partners. For women, the key concern in choosing mates revolves around the male havlng good financial prospects as well as having the qualities that lead to those prospects being fulfilled - such as a man's ambitions, industriousness and social status. Women also universally desired men who were older than they were with older age being linked to greater resources.

The third cluster of findings focuses on cultural variability, with the desire for chastity being the most culturally variable finding. Cultures like the (mainland) Chinese viewed virginity as indispensable in a mate, while people in countries like lreland and Japan placed intermediate value on virginity. For the Scandinavians, virginity was an irrelevance. '

For further details, read an interview with David Buss. Here I'll unpack some of this scenario from the two different perspectives.

2. From the woman's point of view:

'Hogamous, higamous;
Men are polygamous.
Higamous, hogamous;
Women monogamous.'

(The psychologist William James is reputed to have woken one morning with this 'ditty' ringing in his head.)

The notion of mate value was introduced into work on human relationships by Symons in 1987. The argument goes like this: consider two men from the point of view of a woman, and in the light of what we know happens to genetic material over time in terms of the differential reproductive success of individuals. Who has the higher 'mate value' (ie: who is more likely to contribute to 'increased reproductive success'):

Consider, for example, a woman who can choose between two husbands, A and B. Husband A is young, healthy, strong, successful, well liked, respected by his peers, and willing and able to protect and provide for her and her children; Husband B is old, weak, diseased, subordinate to other men, and unwilling and unable to protect and provide for her and her children. If she can raise more viable children with Husband A than Husband B, then his "mate value" can be said to be higher. Over evolutionary time, ancestral females who had psychological mechanisms that caused them to find males of high mate value more sexually attractive than males of low mate value, and acted on this attraction, would have outreproduced females with opposite tastes. This differential reproduction would continue until such mechanisms became universal and species-typical in women. This logic leads one to expect that a man's sexual attractiveness to women will be a function of traits that were correlated with high mate value in our natural environment: the environment of a Pleistocene hunter-gatherer. Natural selection should have designed evaluative psychological mechanisms (information-processing rules or algorithms - see Lecture 16) in women that assess such traits and give rise to sexual and romantic attraction in response to them. (Ellis, 1992: 267).

Note the line of reasoning that Ellis then works within:

The central premise - that women will respond preferentially to men displaying traits indicative of high mate value - does not imply that women consciously appraise men through the sharp eye of matemal pragmatism. When a woman experiences feelings of sexual attraction, she is not, at an unconscious level, "plotting" a reproductive strategy designed to maximize the representation of her genes in future generations. Rather, she is probably simply experiencing desire for the man in question; this desire may or may not enhance reproductive success in the milieu where it is experienced. But underlying the nature and intensity of that desire is a complex host of psychological mechanisms, and these mechanisms should have been designed by natural selection to detect and prefer mate traits that in our natural environment were reliably associated with (a) the ability and willingness to provide economically, (b) the ability and willingness to protect a woman from physical attack or intimidation, and (c) the ability and willingness to engage in direct parenting activities such as teaching, nurturing, and providing social support and opportunities. Taken together, these preferences form a coherent, integrated system that throughout our evolutionary history presumably had the effect of causing women to choose men of high mate value (ibid: 284).

The evidence is not inconsistent with this view (see readings)

3. From the man's point of view:

Here the picture has been portrayed as being very different! Just think of dramatic American soap operas. Think specifically of the 'mate' relationships portrayed: the older man has the ravishing young wife or mistress. Do you think this is a fair assessment? Well, ask yourself this:

What is your stereotype of the kind of wife you would expect of a 55 year-old producer of Hollywood movies?

Some of the really interesting issues here come into view when we start asking some further questions: why do you have this kind of stereotype? Have you any empirical data to back this up? I haven't any real database to work from (maybe we could do some research on this as a class project). But I think she will be younger than him. Why is she 'ravishing'? What does that mean? Ravishing to whom? How do you know?

But these are not questions we can explore directly here: they're partly philosophy of science, and you need to go elsewhere for that; and they’re partly social psychology. Now, there are other relationships than these portrayed in soaps. However: these can be seen as more or less successful 'playings' with this fundamental theme of [male] power as an aphrodisiac. Take gender out of this stereotype. Powerful women beyond reproductive age can have 'toy boys'. This is playing with the theme of power. It's 'usually' the other way around. And the claim in this line of psychological theorising is that we are 'built' to see such things as men being older than their partners as 'usual' (for revision material on this question of how 'structure' plays a role in perception, etc, go back to lecture 1), and hence can grasp different plot devices as sensible in relation to our biases. And this is something else I won't pursue here, since we're getting into the territory of literary theory, and how people make sense of plots, themes, narratives and stories.

So what are the differences proposed between men and women? Here is a summary paragraph that provides one possible answer:

Men, for whom the reproductive stakes are higher, are likely to be more competitive with each other, and therefore... men are more likely to end up wielding power, controlling wealth and seeking fame. Consequently, women are more likely to have been rewarded for seeking power, wealth or fame in a husband than men are in a wife. Women who did so therefore probably left more descendants among modern women. So it follows from evolutionary thinking that women are more likely to value potential mates who are rich and powerful. Another way to look at it is to think of what a woman can most profitably seek in a husband that will increase the number and health of her children. The answer is not more sperm, but more money, or more cattle, or more tribal allies, or whatever resource counts (Ridley, 1994: 258-9).

This view is in line with the findings on preferences from Buss's cross-cultural study, already mentioned.

4. Fantasies as evidence

It is also backed up by a study by Ellis and Symons (1990). The reasoning they employed was that it is difficult to measure preferences in real life, since what happens there is generally a compromise, if only at the level that any arrangements between two people usually are. But you can get away from this problem if you look at people's fantasies:

if, as we argue, men and women differ in their innate sexual psychologies, sexually dimorphic psychological mechanisms should be revealed more sharply and dramatically in sexual fantasies than in sexual activities, since real-life heterosexual interactions must inevitably compromise, and hence blur, male and female desires and dispositions (ibid: 527-8).

Their subjects were 182 female and 125 male college and University students in California, who were given a questionnaire. Ridley (1994: 261) summarizes their results as follows:

Ellis and Symons found that two things showed no sex differences at all. The first was the students' attitudes to their fantasies: guilt, pride and indifference were each as common among men as among women. And both sexes had a clear image of their fantasized partner's face during the fantasy. On every other measure there were substantial differences between the men and the women. Men had more sexual fantasies and fantasized about more partners. One in three men said they had fantasized about more than one thousand partners in their lives; only eight percent of women had fantasized about so many partners. Nearly half of the women said they never switched partners during a sexual fantasy; only twelve percent of men never switched. Visual images of the partner(s) were more important for the men than touching, the partner's response or any feelings and emotions. The reverse was true of the women, who were twice as likely to focus on their own responses rather than on the partner. The women overwhelmingly fantasized about sex with a familiar partner.

The patterns of these differences appear to fit with the lines of argument that were introduced in Lecture 3 on mate choice: men and women have different psychologies because, from an evolutionary perspective, they have different pay-offs for their actions. The actions are not predetermined, but the preferences that predispose possible courses of action are structured into the baseline of how we perceive one another.

Now, there is one further set of work, by Devendra Singh of the University of Texas, that I want to introduce you to.

5. Waist-hip ratios and physical attractiveness

Little has come out of the social psychological literature concerning what women find physically attractive in men. But there was one finding noted above: height. Could the finding about height determining the relative attractiveness of men to women give us a clue about looking at the determinants of attractiveness in a different way? We've already noted that men tend to be bigger than women, and that this difference in size probably has been selected for in our evolutionary past. Could it be that both men and women have particular preferences for the 'ideal' physical characteristics of the opposite sex? And could these differences be accounted for by evolutionary theory?

I turn here to two papers by Devendra Singh, a psychologist at the University of Texas. Singh's work first focussed on men's perception of the attractiveness of women (1993), and produced a line of support for the above 'traditional' view of male vs. female differences in the perception of physical attractiveness. It was only subsequently that he extended his work (1995) - to do the control study for the 1993 study by looking at women's perception of men's attractiveness in the same framework. So, here I'll start with his 1993 paper.

'Fat' has become a loaded and politically sensitive term in recent years, but the substance itself is pretty useful to us. Fat is the way our bodies store energy reserves. Animals that hibernate lay down deposits of fat to tide them through long periods of not eating. In a world in which your next meal cannot be guaranteed, being able to store your next surrogate meal from the remains of your last feast is a great bit of evolutionary engineering. And it's hygienic as well: body fat doesn't go off in the way the contents of your fridge do. Fat has evolved.

Fat is essential to human reproduction. Women who are pregnant are literally trying to feed two people when they eat. Pregnancy creates a large demand for extra energy. Some of that is provided for by fat. Once a woman's fat deposits go below a certain percentage of her body weight, then she will generally stop ovulating, or at least show changes in the regularity of her menstrual cycle. This occurs in anorexics, body builders and marathon runners. It also tends to be the case that mothers who breastfeed infants tend to begin re-ovulating later than mothers who bottlefeed their infants.  Any fat deposits left after pregnancy now either aid in sustaining milk production, or are not rebuilt as quickly as usual, since current food intake is used in milk production without leaving much over to be laid down as fat. And a certain percentage of fat has to be built up for ovulation to resume. OK, so you should by now be thinking that if fat plays a role in reproduction, then hormones are going to be coming in here somewhere, and we've already seen how hormones are implicated in one aspect of the psychology of human behaviour. It's the steroid hormones that are involved in the regulation of fat deposition and usage, and they contribute to that awful problem of dieters, that fat can be used up from one region of the body while being simultaneously laid down at another (Pond, 1981). Now let's turn to another 'fat' issue.

It probably hasn't escaped your attention that men and women tend to get fat in different ways. Us guys get 'beer guts'; women don't, but tend to put it on in other parts of their anatomy. But only at certain ages: the sexes are similar in infancy, early childhood and old age; differences in fat distribution are most marked between early teenage years and middle age (Vague, 1961). Singh puts the situation in a way I can't match:

The most striking gender-specific difference in the physiology of fat accumulation and utilization are observed in the abdominal and gluteofemoral (buttocks and thighs) regions. Simply stated, testosterone stimulates fat deposits in the abdominal region and inhibits fat deposits in the gluteofemoral region. ... estrogen, by contrast, inhibit[s] fat deposits in the abdominal region and maximally stimulate[s] fat deposits in the gluteofemoral region more than in any other region of the body (1993: 294).

A consequence of this is that men and women have a different waist-to-hip ratio (WHR), where 'waist' is defined as 'narrowest portion between ribs and iliac crest' (ibid), and 'hip' is defined as 'the level of the greatest protrusion of the buttocks' (ibid). So, to determine the WHR, measure the waist and divide it by the hip measurement. And you will appreciate from what has just been said that women will tend to have smaller WHRs than men, because . . . well, work it out. In fact, the typical range of WHR for healthy women between early teen years and middle age is between .67 and .80 , whereas for men it is between .85 and .95 (sources are cited in Singh, 1993).

Now, there is a lot of evidence (see Table 1, p.295 in Singh, 1993, which I can't include here because of copyright restrictions, but it is in the Readings) that

WHR reliably signals female reproductive status (pre- or post- pubertal and menopausal), reproductive capability, and, to a certain degree, health status, as inferred from risk for major diseases (ibid: 295).

And so we come to the central hypothesis that motivates Singh's study:

If the attributes of good health and reproductive capability are critical in mate selection as posited by evolutionarily based theories, then men should possess mechanisms (conscious or unconscious) to detect these features in women and assign them greater importance than other bodily features in assessing female attractiveness (1993: 295).

That is, men should rate women who have a waist-hip ratio in the range .67 to .80 as more attractive than women whose WHR falls outside this range. In addition, Singh included three groups of women amongst whom the ratings were made: underweight for their height; normal weight for their height, and overweight for their height. His results indicate that normal weight women are rated as more attractive than those under- or overweight, and those with a WHR of .7 are more attractive than those with any other WHR included in the experimental materials. Thus:

The present findings suggest that body fat and its distribution play a critical role in judgements of female attractiveness, health, youthfulness, and reproductive potential. All of these attributes are associated with a female figure of normal body weight and low WHR. Neither body weight nor WHR alone is associated with female attractiveness. Highly attractive women must have a low WHR; yet deviation from normal body weight, either lower or higher, reduces attractiveness and perceived healthiness (Singh, 1993: 303).

Check these findings out for yourself:

Figure 02

Figure 03

Figure 04


When Singh (1995) repeated this experiment, but with women judging the attractiveness of men by their WHR in the different weight ranges, the most attractive figure was of normal weight and a WHR of .9, right in the middle of the predicted range of .85 to .95. In addition, Singh included information about the financial status of the men being rated, and found that this is interactive with attractiveness as measured by WHR. Thus, contrary to most of the previous literature:

Women prefer men who are attractive and have higher financial status. Thus, physical appearance and financial status (ability to support and defend?) both influence females’ mate preferences, as predicted by Darwin (1871) (ibid: 1099).

Another aspect of these findings that strikes me as significant is that they relate to a part of our anatomy that is uniquely human: none of the other apes have a waist (as you will see in the following figure, which I also include here since they just don't publish figures like these anymore in modern journals: this one comes from 1926). Thus, our perceptual preferences are not a part of our common inheritance shared with our closest relatives, but are something uniquely human.

For additional information, see this update.

6. Assessment Questions

List the arguments that explain:

  • 1. Why men prefer women with a WHR of between .67 and .80; and
  • 2. Why women prefer men with a WHR of between .85 and .95.

Web resources

Three different but related readings are directly relevant here. The first, by Jared Diamond , asks questions about the decline in women's fertility when they reach menopause. As he introduces it, the problem is this:

Human females undergo a steep decline in fertility from around the age of 40 and within a decade or so can no longer produce children. While some women continue to have regular menstrual cycles up to the age of 54 or 55, conception after the age of 50 was almost unknown until the recent advent of hormone therapy and artificial fertilization. Human female menopause thus appears to be an inevitable fact of life, albeit sometimes a painful one. But to an evolutionary biologist, it is a paradoxical aberration in the animal world. The essence of natural selection is that it promotes genes for traits that increase one's number of descendants bearing those genes. How could natural selection possibly result in every female member of a species carrying genes that throttle her ability to leave more descendants?

The second item reports a study of ' lonely hearts ' advertising, by David Waynforth, a graduate student in anthropology at the University of New Mexico, and Robin Dunbar, a primatologist at the University of Liverpool. How do people represent themselves and specify what they are seeking in others when they seek to find a partner through a newspaper advertisement? The material you have looked at here suggests men should be looking for women who are younger than themselves, and . . . well, figure out what you might expect, and see what they found.

And the third item reports some quite unexpected results about human body odour preferences that again seem to fit into an evolutionary perspective.

A fourth source is given here as a way of having you consider the extent to which the kind of material presented thus far in this course can provide the basis for a comprehensive view in psychology. It confronts you with an ethical dilemma: would you use any of the material we have looked at thus far in justifying your answer to the question posed at the end of this paragraph? It also asks you to think about humans, not as organisms that have evolved, but as human beings who have rights, self-consciousness, and difficult choices to make within the framework of legal systems that are often the result of an historical accident, prejudice and uncritical acceptance of a status quo. The question? Should gay or lesbian people marry ?

And finally, some light relief. One thing that men and women do is kiss one another. Back in the 1960s, Desmond Morris speculated in his Naked Ape book that kissing as a gesture had its origins in the weaning of infants from breast milk to solid food. There weren't food-processors way back when, nor cans of baby food. So food was chewed by adults, and passed from their lips into the mouths of babes. Reasoning on the lines that behaviours that involve infants are often transferred into the adult realm (see lecture 3), he came up with his explanation. Well, who knows? Perhaps you'll have a better idea if you bone up on the history of the kiss. Or what about knighthood, the invention of romance in the West, when troubadours sang of devotional love? How do you measure up on the 31 Rules for Lovers listed by Andreas Cappelanus in his The Art of Courtly Love ?


Darwin, C. (1871) The descent of man and selection in relation to sex. London: John Murray.

Ellis, B.J. (1992) The evolution of sexual attraction: evaluative mechanisms in women. In J.H.Barkow, L. Cosmides and J. Tooby (Eds.) The adapted mind: Evolutionary psychology and the generation of culture. Oxford: Oxford University Press. Pp. 267-88.

Ellis, B. J. and Symons, D. (1990) Sex differences in sexual fantasy: An evolutionary approach. Journal of Sex Research, 4: 527-555.

Pond, C.M. (1981) Storage. In C.R.Townsend and P. Carlow (Eds.) Physiological ecology. Sunderland, MA: Sinauer Associates.

Ridley, M. (1994) The red queen: Sex and the evolution of human nature. London: Penguin.

Singh, D. (1993) Adaptive significance of female physical attractiveness: Role of waist-to-hip ratio. Journal of Personality and Social Psychology 65: 293-307.

Singh, D. (1995) Female judgement of male attractiveness and desirability for relationships: Role of waist-to-ip ratio and financial status. Journal of Personality and Social Psychology 69: 1089-1101.

Symons, D. (1987) Can Darwin's view of life shed light on human sexuality? In J.H. Geer and W.T. O'Donohue (Eds.) Theories of human sexuality. New York: Plenum.

Vague, J. (1956) The degree of masculine differentiation of obesities: A factor determining predisposition to diabetes, arteriosclerosis, gout and uric calculous disease. American Journal of Clinical Nutrition 4: 20-34.


The morphed images of women with different waist-hip ratios come from a collection of experimental resources developed by Ronald Henss at the Psychologisches Institut, Universität des Saarlandes, Germany.

Back to Course Outline