Why does the Oxford Dictionary of English portray women as “rabid feminists” with mysterious “psyches” speaking in “shrill voices” who can’t do research or hold a PhD but can do “all the housework”?
The Oxford dictionary he was talking about was the one that comes with Apple devices (Macs, i-Pads, i-Phones), and his question was about the examples that follow the definition of a word and illustrate its use in practice. The ones he reproduced included the phrase ‘a rabid feminist’ illustrating the metaphorical usage of ‘rabid’; the sentence ‘I will never really fathom the female psyche’ exemplifying the use of the term ‘psyche’; and a series of examples featuring women and female voices in entries for ‘shrill’, ‘grating’ and ‘nagging’. He also reproduced entries for the words ‘doctor’ and ‘research’ where the examples referred to doctors/researchers as ‘he’.
The point of this intervention was not just to criticise a few specific entries, but rather to draw attention to a pattern of sexist stereotyping in the dictionary’s illustrative examples. But when Oman-Reagan tweeted to Oxford Dictionaries, citing the ‘rabid feminist’ example, whoever was running their Twitter account that day chose not to acknowledge the deeper point. Instead he was told that (a) the ‘rabid feminist’ example was authentic, and (b) that ‘rabid’ isn’t necessarily a negative term. In the ensuing arguments (first on Twitter and then in lengthier pieces like this and this) the main issue became whether Oxford was endorsing a view of feminists as mad fanatics, and then compounding the offence with its dismissive responses to criticism.
Eventually Oxford apologized for its ‘flippant’ tweets, and promised to review the example in the ‘rabid’ entry, noting that in its corpus (the collection of texts which examples are drawn from) the commonest words found alongside ‘rabid’ are actually ‘fan’ and ‘supporter’. In one way that’s a positive outcome, but in another it’s frustratingly limited: revising a single entry which has been criticized for overt political bias does not address the much larger problem of covert sexism in the dictionary as a whole.
I use the word ‘covert’ for two reasons: first, because most of the sexist examples are incidental, appearing in entries for words which are not specifically ‘about’ women; and second, because much of the sexism will remain invisible if you only look at single entries in isolation. There’s nothing obviously sexist about an entry for ‘research’ where the example sentence uses the pronoun ‘he’; what’s covertly sexist is if there’s a systematic preference for ‘he’ over ‘she’ in all the entries for words denoting intellectual pursuits. The effect is cumulative, and arguably all the more insidious because we’re unlikely to be conscious of the pattern that produces it. This point rather got lost in the debate on ‘rabid feminist’. Oxford was held to account for that particular example, but not for the more systematic bias that Oman-Reagan had detected.
He isn’t by any means the first to have detected it. Feminists who study dictionaries have been complaining about the sexist example problem for decades. I discussed it myself in an earlier post, taking examples from a foreign language learners’ dictionary where the entry for ‘slip’ was illustrated with ‘he slipped on his shoes’ and ‘she slipped off her dress’, while ‘mop’ had men mopping their brows and women mopping floors. Once you’ve become aware of this pattern, you soon start to notice how pervasive it is. It’s not just a problem in one publisher’s products or one type of dictionary.
But whenever it’s pointed out, the dictionary-makers have a tendency to respond in the way Oxford responded to Oman-Reagan. Their examples, they say, are authentic: every phrase or sentence used to illustrate every entry was actually written by a real person in a real context. Dictionaries just describe usage, they don’t judge it, and they certainly don’t censor it. So, don’t shoot the messenger: don’t accuse lexicographers of sexism when they’re only documenting the sexism that exists in the wider world.
Fair point, or lame excuse? I’d say, a bit of both, but more the latter than the former. As Tom Freeman remarked on his Stroppy Editor blog, ‘even if a sentence isn’t theirs, they’ve still made the decision to use it’. And they can’t really argue that they didn’t have other options. The illustrative examples used in contemporary dictionaries come from very large collections of texts—Oxford’s corpus contains over two and a half billion words—so there isn’t a shortage of authentic examples to choose from. In some cases the argument might be made that a sexist example captures something significant about the usage of a word. We might suspect that ‘shrill’ and ‘strident’, for instance, are more often used about women than men. But in most cases it’s neither necessary nor illuminating to reproduce sexist stereotypes.
There’s also something a bit disingenuous about the protestations of dictionary makers that their products simply reflect the world around them. For the average user looking something up on their i-Phone, the dictionary isn’t seen as a neutral document, but as an authority on the existence, meaning, spelling and use of words—a view its publishers are happy to exploit when they use words like ‘authoritative’ in their advertising. It follows, as Tom Freeman observed, that
Dictionaries do help to set the cultural tone, whether they intend to or not. Their job is to describe the language neutrally but beyond that they should also be aware of how they come across. For example, I have a battered Oxford dictionary from 1969 on my shelf. It defines “jazz” as “syncopated music, & dance, of U.S. Negro origin”. Today, the Oxford website says jazz is “of black American origin”.
As this example suggests, there are areas of usage where the editors of dictionaries are anxious not to come across as culturally insensitive. The first edition of the Oxford English Dictionary (begun in the late 19th century and completed in the early 20th), contained numerous definitions and examples which would now be considered horrendously racist—and not only because they used words like ‘Negro’ (which, though offensive today, was regarded as a polite term in 1969). A famous example is the original entry for ‘canoe’, which distinguished between the type used by white people for sport and leisure, and the more ‘primitive’ type used by ‘savages’ as a mode of transport. This kind of thing has been weeded out during the ongoing process of revising the OED. But the sexism displayed in entries like the ones Michael Oman-Reagan reproduced does not seem to have been targeted in the same way.
Why is that? Partly, it may be because sexist examples are distributed in a different way from racist ones. Whereas racism tends to be concentrated in entries for words that relate directly to particular groups and cultures (like ‘jazz’ or ‘canoe’), sexism is an incidental feature of a much wider range of entries. To deal with it systematically, you’d not only have to get rid of the obvious stereotypes, you’d also have to look at the overall balance of your examples—for instance, check that you had roughly equal numbers with ‘she’ and ‘he’, distributed in a non-stereotypical way. Precisely because it’s so pervasive, eliminating sexism would be a major undertaking.
But I can’t help wondering if there’s a more basic problem here: most people just aren’t that offended by sexism—or at least, by the low-level sexism of clichés like ‘I will never really fathom the female psyche’. It’s a bit like the treatment of sexist verbal abuse in schools or football grounds, which is often talked about as if it were a different thing from the racist or homophobic equivalent: it gets put under the heading of ‘banter’, and women who complain are seen as humourless and over-sensitive.
The dismissive tweets for which Oxford later apologized were very much in that tradition. Their tone suggested that whoever wrote them did not feel obliged to take complaints of sexism seriously, and did not expect that stance to attract criticism. On this occasion the negative reaction prompted a climbdown–an apology for flippancy and a promise to look again at the example people had objected to. But if we want to see the problem of sexism addressed in a less piecemeal way, we’re going to have to keep sending the message that we don’t think it’s trivial or a joke. Become, in short, a bit less tolerant and a bit more rabid.