I say I say I say, have you heard the one about a new study showing that men are funnier than women? Probably, because every media outlet in the universe has picked it up and rehashed it, from the BBC and Science Daily to Malaysia News and Reddit. Which is no surprise, because it’s a gift to the media, a perfect example of ‘soundbite science’. But if you read this blog regularly you will also have heard the one about why it pays to take a closer look at these sex-difference stories. And guess what, I’m going to tell that one again.
I’ll start by summarising the research itself, as described by one of the researchers, Gil Greengross, in Psychology Today. What he and his colleagues did was a meta-analysis: rather than carrying out their own experiment, they combined the statistical findings of 28 previous studies and calculated the overall size of the effect they were interested in, which was the effect of gender on funniness—or as they called it, ‘humor ability’.
In the studies they selected, one set of experimental subjects had been given some cartoons and asked to come up with amusing captions; another set of subjects then rated the funniness of each caption on a numerical scale. Crucially, the raters didn’t know if the caption they were reading had been written by a man or a woman. If they’d known, their judgments could have been influenced by the stereotype of women as less funny than men. The studies had been designed to investigate whether there was any basis for that stereotype. Would women still be judged less funny by people who had no way of knowing they were women?
The short answer turned out to be ‘yes’. As the Psychology Today piece explains,
In statistical technical terms, the effect size was 0.32, or roughly one-third of the standard deviation. In plain English, this means that 63 percent of men score above the mean humor ability of women. This is considered a small to medium difference.
Evidently aware that such findings often attract criticism, and are frequently misinterpreted, Greengross goes on to assure us that yes, they did look for possible confounding factors (e.g., if there were cultural differences which calculations based on aggregated data were masking, or if there were more male than female raters, which might skew the ratings in men’s favour because men prefer the humour of other men); and no, they weren’t saying that all men are funnier than all women. Some women are very funny, and the most successful female comedians are funnier than 99% of men. Their claim was only that men on average are funnier than women on average.
Greengross then turns to the question of why that might be the case, and tells a story I have discussed on this blog before, in fact more than once, because it features in some accounts of the evolution of human language. It’s about sexual selection, the concept developed by Darwin to explain evolved traits like the peacock’s tail, which can’t have been selected because it made the bird better adapted to its environment (it’s hard to fly away from predators trailing that monster); rather it was favoured because it was attractive to potential mates, ensuring that big-tailed peacocks passed their genes to more offspring.
Peacocks are animals that ‘lek’, display themselves at mating time so that peahens can judge their fitness. The idea is that being funny serves a similar purpose for humans: it’s a form of display put on by men so women can assess how good they’ll be as mates. ‘Humor’, Greengross notes, ‘is strongly correlated with intelligence’. He also reminds us that women advertising on dating sites often mention humour as an important quality in men, whereas men are less interested in whether a woman is funny. The principle that ‘males do the courting and females do the choosing’ explains why ‘humor ability’ has been selected for in male humans, but not in females, whose reproductive success is not enhanced by it.
This explanation presupposes that ‘humor ability’ has a significant genetic component; but couldn’t being funny be a learned skill which is more common or better developed in men because they get more opportunity or encouragement to practise it? The researchers, it transpires, did consider that possibility, but they dismissed it on the grounds that
There is minimal evidence to support the view that our society suppresses women from producing and exhibiting humor.
Mate, stop it, my sides are going to split—because this has to be a joke, right? You do realise that there’s a long tradition of discourse, both expert and popular, which ‘suppresses women from producing humor’ by telling them that they’re no good at it, it’s unfeminine, and men don’t find it attractive; and you do understand that YOUR OWN STUDY SITS SQUARELY IN THAT TRADITION? Oh, and while we’re on the subject, it presumably hasn’t escaped your notice that saying ‘men have more humor ability than women’ followed by ‘humor is highly correlated with intelligence’ will lead a lot of people to deduce the proposition ‘men are more intelligent than women’—which you wouldn’t be suggesting in all seriousness, would you?
So far I haven’t questioned the central claim made for this meta-analysis: that it confirms the robustness of the effect it investigated (i.e., sex affects funniness and men are funnier than women). Here are a few reasons why I think that claim needs to be at least qualified, and possibly rejected.
First, the studies used in the meta-analysis were chosen because they investigated the same question using a similar experimental design, which elicited evidence of people’s ‘humor ability’ by asking them to write funny captions for cartoon images. Presumably this task was used because it’s something you can get people to do in a lab, and because a written caption does not reveal the writer’s sex (whereas a recording of them telling a joke or a funny story would). But how good a proxy is captioning cartoons for the general ability to be funny?
I would say, not very good. Most everyday humour is produced spontaneously in the course of interaction, typically with people you know, like and have things in common with, and whose reactions guide your performance from moment to moment. The lab cartoon-captioning task reproduces none of these features. It requires subjects to be funny on command (not spontaneously), in isolation (not in a social group) and in writing (not speech), for an absent audience of strangers whose reactions they can only guess at. To me that makes it a dubious basis on which to generalise about ‘humor ability’. At best, the analysis shows that in lab conditions men write funnier cartoon captions than women—not that they’re funnier across the board.
Second, the finding that men are funnier than women is based on averaging the numerical scores given to subjects’ captions by the judges. But what do these numbers correspond to? Clearly, judges’ ratings must reflect their subjective response to a caption, but how, if at all, does that response relate to the caption-writer’s abilities? The problem is that there’s a lot of variation, individual as well as cultural, in what people find funny. If a judge rates one caption as a 2 and another as a 4, is that because of some objective quality of the two captions—the second is simply funnier than the first–or does it just reflect the judge’s preference for certain kinds of jokes over others? In short, can we have any confidence that these numbers are a useful measure of ‘humor ability’?
I’m sceptical, to be honest. I’m not convinced there’s such a thing as ‘humor ability’, and if there is, as I’ve already said, I have trouble seeing the very specific skill of writing funny cartoon captions as a fair test of it. It’s a test most of us would probably fail: back in 2011, in a Psychology Today piece criticising one of these studies, Ben Hayden noted that the commonest score awarded to captions, regardless of the writer’s sex, had been 1, which meant ‘not funny at all’. ‘That indicates’, he comments, that ‘the experimental subjects were not well-matched to the test—they were just swinging wildly’. They were like the proverbial roomful of typing monkeys, only in the field of cartoon-captioning rather than Shakespearean drama.
Third, even if we do trust the numbers, what the meta-analysis reports is only, as Greengross acknowledges, a ‘small to medium difference’. The effect-sizes reported in meta-analytic studies are generally described using a scale that goes from very small/slight to very large, so this effect isn’t even at the mid-point: to put it in context, it’s smaller than the reported effect of sex on spelling accuracy, and much smaller than the reported effect of sex on the distance someone can throw an object. As sex-differences go, it’s not a biggie. (Yet another ‘men are funnier’ study discussed in Psychology Today in 2011 found a difference so slight that even one of the authors described it as ‘just on the edge of detectability’–and he meant, by using statistical methods, not just observation.) Even liberally sprinkled with the words ‘on average’, the statement ‘men are funnier than women’ gives a misleading impression, because what ‘funnier’ means to most people is ‘noticeably funnier’: it implies a difference large enough to be significant in the ordinary rather than the purely statistical sense.
There are, then, a number of reasons to query the presentation of this meta-analysis as clear confirmation that ‘men are funnier than women’. But my biggest problem with it isn’t the design of the studies analysed, the conclusions drawn by the researchers, or the exaggeration of their significance by the media: it’s the fact that this kind of research is seen as worth doing in the first place.
Think about it: in what other case would anyone design a study to investigate whether there’s a basis for some entrenched negative stereotype of a social group? ‘Is it true that poor people are lazier than rich ones?’ Nope. ‘Are Jews/Scots/Yorkshiremen really meaner than other people with money?’ Nope to the nth power. It would be utterly offensive: no one would fund it and the ethics committee would never approve it. Sex-differences, though, are always potential research fodder. And when the frame, as in this case, is evolutionary psychology, they will always turn out to be explained by the supposed—though in reality often misrepresented–activities of prehistoric hunter-gatherers, and beyond that, by the irreducible difference between the sperm and the egg.
Actually, let me go back a step: it’s not true that all sex-differences are potential research fodder. As Caroline Criado Perez shows in her book Invisible Women, in many areas where women are consequentially different from men, whether it’s the way their bodies respond to drugs or their patterns of travel on public transport, there’s a huge data gap because the differences have not been investigated systematically. Research that would have a positive effect on women’s lives does not get done; yet there’s always room for one more study explaining that women are rubbish at something—maths, map reading, being funny–because biology. (Or alternatively, that women are better than men at something trivial or menial, like nappy-changing or keeping track of people’s birthdays.)
I hope what I’ve just said explains why I think it’s important to challenge claims like ‘men are (on average) funnier than women’. In fact, as I’ve said before, I think the way we talk about who is or isn’t funny matters in its own right: the proposition that ‘Xs have no sense of humour’ has a long history of being used to dehumanise the group in question, and ‘Xs aren’t as funny as Ys’ can also, as noted earlier, be code for ‘they’re not as intelligent’. But even when the difference under discussion really is trivial (for instance, a study once found that female shoppers spend more time browsing than their male counterparts—like their prehistoric ancestors they are natural foragers, whereas men are mighty hunters), it must be seen as part of a larger enterprise, in which the point of cataloguing sex-differences is ultimately to affirm that the sexes are not just different, they are different in ways that make equality a pipe-dream. The meta-message being communicated is ‘women, know your limits’.
Whenever we hear this message we should remember that the whole history of feminism is a story of women refusing to accept the limits imposed by their societies—limits which have always been represented as ‘natural’ rather than man-made. Once, scientists maintained that if women were admitted to universities, the intellectual stress would cause their ovaries to shrivel. Now we recognise that as just self-serving male chauvinist nonsense. The claims made by many scientists today are destined to end up in the same dustbin. Let’s use our intelligence and our ‘humor ability’ to speed them on their way.