Men, Women and Nigel Short
(See also the second, more technical follow-up: Men, Women and Nigel Short 2: An academic response)
Much to my amazement, chess has hit the front pages of the mainstream media for the third time in a fortnight. This time, however, it's more a case of old wine in a new bottle. The always controversial English GM Nigel Short has come under the spotlight for claiming that male and female brains are differently hardwired when it comes to chess. Never mind that the article for New in Chess magazine was published three weeks ago; today was the day, for whatever reason, that the story went viral.
As you might expect, the English tabloids had a field day, covering angles from claims that Nigel said that women shouldn't play chess at all, to claims that women have lower IQs than men, as well as branching out to the issue of general sexism in chess. While Nigel's article didn't hint at any of these claims, there are admittedly several strong players who believe the first two of these points, while the third - a male-dominated chess culture - is undoubtedly true.
I don't want to get into these issues too much. I'm an academic, and for anyone familiar with scientific publications on gender in chess, the issue has really been done to death: After accounting for sample size - the fact that far fewer women play tournament chess than men - there is no significant evidence whatsoever that men are better than women. This has been shown in countless academic studies (although not a single one was quoted in any of the media reports today). I don't claim that there aren't relevant gender differences to professional chess - for one, men have been found to be on average more competitive than women - but this specific question, at least, has been answered some time ago.
If one really wanted to definitively test nature effects, the ideal hypothetical experiment would go something like this:
- Get some twins - one male, one female - and whisk them off to a desert island
- Raise them in identical conditions with no exposure whatsoever to gender influences
- Teach them both chess in identical training environments
- Test their chess strength
- Repeat steps 1-4 with a thousand other sets of boy-girl twins
Not such a convenient experiment to run. But for many people, this is not the real question anyway. For most parents, what they really want to know is whether the answer to "Should I teach my daughter chess?" differs from "Should I teach my son chess?" Many parents are likely worried that the environment for success in chess is more difficult for girls than for boys - and to some extent, this is true. On the one hand, there are fantastic opportunities for female players in today's chess society, with many more lucrative female-only competitions than there used to be. On the other hand, there remains a lot of sexism within the world of chess, as there seems to be in many gender-homogenous communities.
I'm not a parent, and I'm not qualified to give any advice on this. All I can say is that my future children will be given the chance to take up chess as soon as they are able, regardless of whether I sire little Smurfs or Smurfettes.
One final remark on the issue, specifically related to the common human fallacy of underestimating 'non representative samples'. It sounds like a lot of techno mumbo-jumbo, but bear with me. I heard a comment with regard to today's gender issue that "Of course men have higher IQs than women. If you saw a man and wife walking down the street and someone offered you a 50-50 bet for $100 over which one had a higher IQ, would you bet for the man or the woman?"
Interesting bet, but it's crucial to realise that this is not the same question as if we had randomly chosen a man and a woman from the broader population, say the national census. We've been given extra information that restricts our sample: the man and woman in question are a couple. Why does this matter? Well, social scientists have well established that women traditionally value intelligence highly in a mate (either directly, or because they value wealth, which is strongly correlated with IQ). That's not to say that men don't like intelligent women; on average, however, men place higher priority on...other factors. So our man and woman are not representative of the broader population.
Of course, it would be possible to work this out if one really wanted, just as it would be possible to study whether male ballet dancers have lower IQs than female ballet dancers, or whether a gay hairdresser is better at parking cars than a straight hairdresser. But honestly, at some point the question we really need to ask is: who cares?