Once upon a time, someone had the bright idea of making sets of fridge magnets for young children learning to read and write. All the children were following the same school curriculum, but since the designers knew they came in two distinct varieties–some were girls and some were boys–they decided to make two different versions of the product. The girls’ version featured words like ‘make-up’, ‘bunnies’ and ‘love’, while boys were given words like ‘money’, ‘car’ and ‘dirt’. To make sure everyone would know which words were suitable for which children, the designers mounted the magnets on colour-coded pink and blue card.
Parents expressed their gratitude in the reviews they posted on Amazon. ‘Thank goodness for this product!’ wrote one:
For a while now I’ve been concerned about my little girl – she has been showing an increased interest in things which are clearly just for boys, such as monsters and climbing. I have even seen her on occasion use money, ride a bike or go swimming. This product has been a godsend as it has allowed me to say to her once and for all: “These are boys’ things and they do not concern you.”
Another declared himself ‘relieved that the [boys’] set excludes any words that might relate to any form of intellectual pursuit or emotion (other than fear)’.
Not all the reviews were so sarcastic, but almost none of them were positive: most people who left comments were critical of the magnets, and some called on Amazon to stop selling them. The crude stereotyping struck many as particularly out of place in a product that was meant to be educational. As one commenter put it, ‘Words are universal. Vocabulary is not gender-specific unless we make it so’.
But in reality, of course, we do make it so. By repeatedly using certain words about certain kinds of people, we create patterns which are more or less strongly gender-marked. The words are not ‘gender-specific’ in the sense that they can only be used by or about girls and women or boys and men. It’s more that we’ve learned to associate them with either femininity or masculinity. The adjectives ‘feisty’, ‘petite’ and ‘shrill’, for instance, are so strongly coded as ‘feminine’ words, applying them to a male may be taken as a comment on his (lack of) masculinity. In most cases the gender-coding is subtler, but it’s still part of our tacit knowledge.
You can test this out for yourself by looking at the wordlists I’ve reproduced below:
active, adventurous, analytic, assertive, battle, boast, challenge, champion, confident, courage, decision, decisive, defend, determine, dominant, driven, fearless, fight, force, greedy, headstrong, impulsive, independent, individual, intellect, lead, logic, objective, opinion, outspoken, persist, principle, reckless, self-confident, self-reliant, self-sufficient
agree, affectionate, collaborate, commit, communal, compassion, connect, considerate, cooperate, depend, emotional, empathy, enthusiasm, feel, gentle, honest, inclusive, interpersonal, interdependent, kind, kinship, loyal, modesty, nurturing, pleasant, polite, quiet, responsible, sensitive, submissive, support, sympathetic, tender, together, trust, understand, warm
There are no words on either of these lists which could not, in principle, be used in reference to either sex. But the words on List 1 have more masculine associations, while the ones on List 2 are more associated with femininity. If I described some gender-unspecified person as ‘dominant, driven and fearless’ you would be likely to imagine a man; if I described them as ‘nurturing, pleasant and polite’ you would be likely to imagine a woman.
One striking difference between the two lists is that a lot of the ‘masculine’ words seem to be describing leaders, achievers and rugged individualists, whereas most of the ‘feminine’ words describe helpers, supporters and carers. This contrast figures so prominently, you might suspect me of taking the words straight from the pages of Men are from Mars, Women are from Venus. But in fact, I took them from a webpage explaining a piece of software called the Gender Decoder for Job Ads. And the source from which the software designer took them was a 2011 article in a psychology journal, entitled ‘Evidence that gendered wording in job advertisements exists and sustains gender inequality’.
The authors of this study began by looking for gender-marked vocabulary in the job ads on two popular Canadian listings sites. Their sample included ads for both male-dominated occupations like plumbing, engineering and computing ,and female-dominated occupations like nursing, early childhood education and HR. Their analysis showed that the male-field ads used significantly more masculine-coded words.
So far, you might think, so unsurprising: but the kicker is in the second stage of the research, which involved presenting male and female subjects with ads for various positions (they included male-dominated, female-dominated and ‘neutral’ fields) which had been manipulated to make the wording either strongly ‘masculine’ or strongly ‘feminine’. For instance, one version of an ad for an administrative assistant stated that the company was looking for someone ‘dependable and reliable’, while the other specified that the applicant should be ‘independent and self-reliant’. Subjects were asked to say how appealing they found each position, and whether they felt they belonged in the role.
The main finding was that women saw jobs as less appealing, and were less likely to think they belonged, when an ad relied heavily on masculine-coded vocabulary. (Men’s perceptions were less affected by the choice of words: they did find ‘feminine’ ads less appealing than ‘masculine’ ones, but the effect was very slight.) The researchers concluded that the wording of job ads is a factor affecting women’s willingness to apply. The issue isn’t just that women see themselves as unsuited to particular kinds of work: even when they have the right qualifications, the perception that they won’t fit in is reinforced by ads that use masculine-coded language (e.g. ‘we are looking for a self-reliant individual who is driven to achieve results’), and can be countered by ads that substitute more ‘feminine’ terms (e.g., ‘we are looking for a committed, responsible team-member who is sensitive to clients’ needs‘).
This finding prompted the development of the Gender Decoder for Job Ads, a tool designed to help organisations avoid gender bias in recruiting. It works rather like the Gender Genie, which I discussed in an earlier post: if you paste the text of a job ad into it, it will calculate the relative proportions of ‘masculine’ and ‘feminine’ words, and on that basis tell you whether your ad has an overall bias. I came across it on a blog maintained by the UK Parliamentary Digital Service, which published a guest-post earlier this year entitled ‘Breaking the bro code‘. The writer argued that ‘removing masculine words from job adverts is a quick and easy step to attract more women’. This view seems to be gaining ground: Iris Bohnet, for instance, the author of an influential book called What Works: Gender Equality By Design, describes the wording of job ads as ‘low-hanging fruit’ for those who want to reduce bias and build diverse, inclusive workplaces.
My own feelings about this approach are mixed. I certainly don’t dispute that there are bits of the ‘bro code’ which we could and should dispense with: they hang so low their knuckles are dragging on the ground. For instance, according to the Harvard Business School’s recruitment blog, the use of ‘ninja’ as a job title in the tech sector increased by 400% between 2012 and 2016. By all means let’s stop advertising for ‘ninjas’ (unless they’re being employed as role-players in an exhibit about feudal Japan). And while we’re at it, we could cut out the kind of meaningless guff which so many job ads are full of–corporate clichés like ‘we strive to be competitive in a demanding global marketplace’, which increase the masculine vocabulary quotient without adding anything of substance.
But while I’m all for getting rid of what’s unnecessary and offputting (or in the case of ‘ninja’, idiotic), I’m always wary of approaches to sexism which treat changing language as a panacea. Language is rarely the root cause of the problem: it’s the outward and visible symptom of a deeper cultural disease. In this case, for instance, the problem that has to be tackled isn’t just that the language of job ads is inadvertently alienating women. The deeper problem is the gender-code itself: it’s the fact that words like ‘analytic’ and ‘logical’ are generally understood (by women as much as men) to denote ‘masculine’ qualities. That’s got nothing to do with the words themselves, and everything to do with our cultural beliefs about what men and women are like (‘these are boys’ things and they do not concern you’).
Just substituting ‘feminine’ for ‘masculine’ words in job ads does nothing to address this deeper problem. Even if it persuades more women to apply for jobs in male-dominated fields, it does so in a way that leaves the codes themselves intact. It says to women, in effect, ‘you may think you don’t belong in this job, but actually you do, because it isn’t really about leadership and competition, it’s about stuff women are good at, like teamwork and collaboration’. Is that challenging gender stereotypes or is it pandering to them?
Iris Bohnet, the author of What Works, might respond that I’m missing the point here. The evidence suggests that changing the language of job ads ‘works’: it helps to diversify the applicant pool for jobs. So what if people still mentally put words, and the attributes they denote, into pink and blue boxes? ‘It’s easier’, Bohnet says, ‘to de-bias organizations’ practices and procedures than to de-bias mindsets’.
As I said before, my feelings about this are mixed: it’s not that I can’t see the force of Bohnet’s argument. But in the end I think feminism does have to be about changing mindsets rather than just devising procedures to work around them. And while I realise there’s no quick fix for sexist thinking, I’ve been alive for long enough to know that change is possible. Back in 1962, when I was learning to read, no parent would have objected to those pink and blue fridge magnet sets. Today, many parents find them objectionable. It’s been a long, slow process, and it isn’t finished yet. But if researchers 100 years from now discover that ‘logical’ is still a blue word and ‘compassionate’ is still a pink word, my ghost will be seriously disappointed.