Passing and Failing the Naipaul Test

V.S. Naipaul thinks he can tell within a paragraph or two whether a piece of writing is by a woman. And he thinks no woman—especially not that chit Jane Austen—can write as well as he does.

Interviewed at the Royal Geographic Society on 1 June, the Trinidad-born Nobel literary laureate opined that women’s inherent inability to match his brilliance comes about because of their ‘sentimentality, their narrow view of the world … And inevitably for a woman, she is not a complete master of a house, so that comes over in her writing too.’

Naipaul is a silly sausage to whose opinions we really shouldn’t give much credence. Notwithstanding his tendency to be curmudgeonly for the sake of it, no one author’s work should reasonably be taken as a touchstone for literary quality in general.

Still, Naipaul’s comments infuriated people because there’s already a groundswell of anger at the way literary culture is so heavily skewed in favour of male authors. Despite the fact that an overwhelming majority of publishers, editors and readers are female, women are startlingly diminished in everything from children’s book protagonists to the ways books are marketed, reviewed and awarded.

Enter the Guardian newspaper’s ‘Naipaul test’. Taking Naipaul at his word, it challenged readers to determine an author’s gender based solely on a paragraph of his or her writing. The test contained ten paragraphs—five by men, five by women—chosen from well-known works. The idea of determining gender by analysing text isn’t new. In 2003 a team of scientists led by Moshe Koppel of Bar-Ilan University, Israel, and Shlomo Argamon of Illinois Institute of Technology developed an algorithm that could predict the author’s gender with 80 per cent accuracy.

As the New York Times reported at the time, ‘women are more comfortable talking or thinking about people and relationships, while men prefer to contemplate things’. The researchers suggested women writers prefer personal pronouns such as ‘I’, ‘you’ and ‘she’, while men opt for ‘determiners’ including ‘a’, ‘the’, ‘that’ and ‘these’, plus numbers and quantifiers such as ‘more’ and ‘some’.

The ‘Israeli method’ has been adapted for an online test called the Gender Genie. When I had it chew over some of my recent work—a review, a news story, an opinion piece, this essay—the Gender Genie deemed me to be a man every time. But curiously, when I pasted in several posts from my various blogs, the Gender Genie always picked me as a woman.

I also subjected my personal blog, A Wild Young Under-Whimsy, to testing by the Gender Analyser, which was 54 per cent sure it was written by a woman, ‘however it’s quite gender neutral’. My fashion blog Footpath Zeitgeist was, perhaps unsurprisingly, also deemed to be by a female author (64 per cent), while my online professional portfolio is allegedly the work of a man (82 per cent).

It’s only natural to ask what such tests ‘mean’, since humans are subjective, social creatures calibrated not so much to read but to read into. We constantly—even subconsciously—weigh a text against others we’ve read, experiences we’ve shared, ideas we’ve encountered and prejudices we’ve acquired. In this way, we create personal ranking systems for the esteem in which we hold particular kinds of writing.

Widely circulated on Facebook and Twitter, the Naipaul test inevitably led to conversations about what low or high scores ‘meant’. Those who did ‘poorly’ were comforted by the idea that this meant they were virtuously ‘gender-blind’. Meanwhile, having ‘passed’ the test with six out of ten, was I more sexist?

The test sets out to challenge the accuracy of literary stereotyping strategies. I got my score by thinking explicitly about these stereotypes, then second-guessing the likelihood that I was encountering one against the likelihood that the Guardian would anticipate my strategies and deliberately pick authors or texts to defeat them. Perhaps this author was successfully evoking a character of the opposite gender, or working in a counter-intuitive genre. For instance, Nicholas Sparks writes sentimental love stories, while Zadie Smith favours incisive, analytical essays. And for every Philip Roth novel about Philip Roth, there’s an empathetic Ian McEwan portrait of a woman, or a Lionel Shriver portrait of a man.

My first strategy was genre. Based on my own previous reading, I assigned each paragraph to a category such as ‘epic postcolonial literature’ or ‘romantic memoir’. Then I considered whether I associated those genres with men or women. (I decided ‘epic postcolonial literature’ was male. Naipaul himself turned out to have written that paragraph. How droll.)

Next I weighed stylistic considerations. Is bold action or authoritative declaration ‘male’, and introspective, equivocal, reflective writing ‘female’? I also considered voice—of both the protagonist and the narrator. Are narcissistic protagonists or smug, detached narrators the avatars of a male author? And how valid is the assumption that empathetic voices spring from a feminine imagination?

Finally, I evaluated word choices. Were the Israeli researchers correct in assuming that male authors favour terse, verb-packed, object-oriented language, while female authors are more descriptive and relationship-oriented? Or perhaps male authors display a certain self-conscious, arrogant stylisation, while women express themselves more simply and forthrightly. Does ‘I’m a sight this morning’ sound feminine and ‘she was planning one of her fucking headaches’ appear masculine?

You mightn’t agree with my strategies—and they weren’t always correct—but the Naipaul test is valuable because it prompted me to make explicit the kinds of judgements I might have previously internalised. The test reveals how we interpret gender at the capillary level of a text, not just through the prism of the literary industries. It spotlights how every reader is implicated in making gendered value judgements, even as it reveals the arbitrariness of such judgements. Perhaps the more widely we read, the more source material we accumulate for our stereotyping strategies, and therefore the subtler they are likely to be. But unlike Naipaul, we should be humble enough to admit we can’t know everything.

Copyright Mel Campbell 2011