Sunday, October 10, 2010

Bias

Would you treat a little Connor or Chelsea differently from an Alexander or Elisabeth on their first day at school? Or be impressed by a male violinist's virtuosic playing in comparison to his female collegue's 'smaller technique' even though they performed with equal skill?

You'd probably like to think you wouldn't. Most of us don't mean to be irrationally prejudiced, and would be upset and offended at the suggestion that we harbour such ideas about people based on their gender or name, never mind more politically sensitive attributes such as race and skin colour. But we are, and we do. Orchestral auditions, while possibly not the most critical area in which we need fairness, are a good illustration of the problem - it was shown that when musicians performed their auditions behind a screen, so that judges were unable to know their gender or other personal characteristics, the probability of women being hired increased significantly.

On a similar note I wonder how many of the choirmasters who insist that young boys' voices have some elusive 'special quality' that girls just can't emulate would find themselves able to distinguish accurately between boys and girls in a screened audition. Just for amusement value I'd like to see if coffee snobs could really tell their Nescafe from Starbucks from Proper Italian Coffee in a controlled test.

And then there's my own experience, in the corporate world, on both sides of the interview process. I remember being told in feedback for an interview that my voice did not sufficiently covey 'confidence and authority' in the group discussions and presentations, compared to that of the other candidates (all male). Because the City is no place for feminist analysis, I politely accepted the feedback rather than questioning whether social conditioning and centuries of cultural baggage might just cause the interviewers to unconsciously associate those traits with the deeper sound of the male voice. And from the other side, hearing my colleagues discuss a candidate they'd just interviewed, technically very strong, but they wondered whether she would be a good fit for the team, would she be up for the banter? Easier maybe to just hire someone just like themselves. No one explicitly said they didn't want to hire a woman, but that, whether they realised it or not, was effectively what they meant.

I recently wrote an article for The F-Word about the need for women-only literary prizes. I want to be clear that I don't think the judges in such competitions are intentionally judging women differently, but that all of us are human, which means we are not nearly as objective and rational and free of prejudice as we would generally like to think. What to do about it is a more difficult question.

Does segregating women in their own competitions make the problem worse, or risk being seen as patronising to women, implying female authors are in need of special treatment or unable to compete fairly? Possibly, and acclaimed literary author AS Byatt for one would say so. But for me the obvious response is to ask why, centuries after the Bronte sisters and George Eliot had to hide behind their male or ungendered pseudonyms, in a world supposedly now free of gender prejudice, she chooses not to publish her work as 'Antonia Byatt'.

This is by no means a problem limited to the literary world. I am unlikely to ever be nominated for a literature prize, however I would definitely hope to publish some research papers in the next few years. So what about academic peer review, the process of criticism and approval or rejection by other academics that a paper has to go through before it can be published in a reputable journal? But unlike the Orange Prize question, this seems to me a problem with a simple and easy solution.

In science we recognise our inability to be objective in other situations, for example in evaluating the effect of a drug or medical treatment, which is why randomised, controlled and double-blinded clinical trials are considered the gold standard. Otherwise we can't know we're not subconsciously cheating ourselves. Even outside of clinical trials, in scientific experimentation we place great importance on knowing whether what we really know what we think we know.

So it's remarkable to me that even in science we send papers to be reviewed with the authors names clearly indicated. There's no good reason for this not to be an anonymous process, like orchestra auditions, and in fact similar studies have shown a similar increase in women's representation when 'double-blinded' review is used. But it still isn't universally or even widely adopted.

For now, one possible alternative solution is to take a leaf from AS Byatt's book and use initials only, which cleverly disguises the gender of the author, and indeed this is common practice anyway. But that only works until the reviewer looks you up on your institution website and you're betrayed by your girly photo and the feminine pronouns in your bio. And gender is not the only issue. Both given and surnames tell us a certain amount about ethnic and cultural background, and social class. In the 'naughty children's names' survey, for example, it's notable how many Irish names appear in the list of boys names viewed negatively by teachers, though I'm sure most of them would vehemently deny any anti-Celtic feeling.

Double-blinded peer review seems the only practical way to overcome these problems, and no one should be offended at the implication that they are racist or sexist or some other -ist, unless they are offended at being considered human.