Politics, by definition

That troublesome word ‘woman’ has been causing controversy again.

Last week, a Twitter user who goes by @ShoelessJoe1910 shared two responses from the makers of Collins Dictionaries to people who’d contacted them about the dictionary entry for ‘woman’. One correspondent had received a reply that looked like a standard piece of boilerplate:

As lexicographers, our duty is to report the language as it is used… Whilst we do welcome all feedback received from our users, any changes we make to our definitions are the result of a detailed review process and evidence-based linguistic research.

Another correspondent who raised the same subject got a different response:

Thanks again for contacting us about the definition of ‘woman’. …We are currently reviewing all our gender-related vocabulary to make sure that we accurately reflect the evolution in the vocabulary of gender and sexuality. This review will be completed in the coming months, and your comments will most certainly be taken into account. We always welcome feedback from our users, so do not hesitate to contact us if you notice any other inaccuracies and omissions.

The subject of both communications was whether a dictionary entry for ‘woman’ should define the word as meaning ‘an adult female human being’ (as Collins currently does), or whether it should (also) inform users that ‘woman’ denotes a person who identifies as a woman. The first correspondent wanted the lexicographers to maintain the traditional definition; the second wanted them to change it.

What initially bothered @ShoelessJoe1910 was the contrast between Collins’s dismissive treatment of the first correspondent, a woman, and the deferential manner in which they addressed the second, presumed to be a man (it was later clarified that this correspondent was actually a trans woman). But what drew people into the thread was the question about how ‘woman’ should be defined. Most comments endorsed the traditional definition, and criticised the dictionary for considering any other. Some thought this was an Orwellian plot to cut the cord which tethers language to reality. One was sufficiently incensed to call for a boycott of HarperCollins’s products.

And what, I hear you ask, does this blog think? I think I’m about to piss off both sides in this argument by explaining why I believe it’s pointless to pursue your political objectives by lobbying lexicographers about dictionary definitions.

If you’ve read any of my previous posts about dictionaries, you’ll know that I don’t regard them as just objective and apolitical works of reference. They have historically exhibited all kinds of biases, including androcentrism and casual sexism, and there are some traces of that history which I think it’s reasonable to ask them to get rid of—especially their unreflective use of sex-stereotyped examples illustrating the current usage of words, which is neither necessary nor helpful to their users.

Other kinds of sexism are more difficult for dictionaries to eliminate while still fulfilling their core functions. For instance, if you read Collins’s current online entry for ‘woman’  you’ll see not only some thoroughly sexist example-sentences in the illustration section, but also some secondary senses of ‘woman’ (e.g. ‘domestic servant’; ‘wife, mistress or girlfriend’) and some idioms containing the word (e.g. ‘little woman’, ‘woman of the streets’) which feminists might well find objectionable. But their inclusion is not a mark of the lexicographers’ own sexism, it’s a reflection of the sexism of the community whose usage they’re describing. We might query the range of idioms selected—they’re a pretty dated-looking set—but even if some of them are no longer in common use, they still appear in sources (like Victorian novels) which 21^st century language-users encounter fairly frequently. Dictionaries have quite exacting criteria for declaring a usage obsolete, and one consequence is that they are rich sources of evidence about the prejudices of the past.

But whatever you think about the retention of old usages which offend modern sensibilities, one thing it’s not reasonable to ask lexicographers to do is ignore the development of new usages which express more contemporary attitudes. I’ve given this example before, but it bears repeating: what would we think of an entry for ‘marriage’ that defined it, in 2017, as ‘the union of a man and woman’ or ‘the relationship between a husband and a wife’? That’s what it used to mean, and it’s also what quite a lot of people think it should still mean. But theirs is no longer the majority view: in many parts of the English-speaking world the law has changed to permit same-sex marriage, and the usage of ‘marriage’ reflects that. Dictionaries have therefore felt the need to update their entries for the word. Collins’s, for instance, though it makes no explicit reference to same-sex marriage, is written in pointedly gender-neutral language.

Similarly, the gender-identity-based definition of ‘woman’ now reflects the usage of at least some people in at least some contexts. Whether that usage merits recording in a general-purpose dictionary will depend on the criteria the dictionary uses to decide if something has entered ‘general’ or ‘common’ usage: I assume that’s what the Collins lexicographers will be looking at in their review of gender-related vocabulary. I also assume that if they do decide to record the identity-based sense of ‘woman’, what they’ll do is add this definition to their revised entry, not substitute it for the current one. I’m confident the evidence is not going to show that English-speakers have stopped using ‘woman’ to mean ‘adult female human being’.

In my view, what Collins told the first correspondent was right: ‘thanks for your input, this is a question that’s on our radar, but our decision will be based on analysing a large sample of relevant linguistic data, not on random emails from a few individuals who feel strongly enough to lobby us about it’. That’s also what they should have told the second correspondent. If your policy is to base definitions on corpus evidence about word-usage (and if it isn’t you’re basically just Urban Dictionary) then you should spell that out to everyone who contacts you—ideally without implying that you regard them as either out-of-touch, prescriptive bigots or oracles of wisdom. (Of course, that means that when you say ‘we welcome all feedback from our users’ you’ll be lying about 99% of the time, but such is life for lexicographers. Some of the feedback they get makes the comments in the Daily Mail look sensible.)

If I were in charge of all things linguistic, what I’d want to change with a wave of my magic wand would not be the principles of descriptive lexicography (even if some of its practices could be improved), but the popular attitude which makes dictionaries perennial targets for political lobbying. By treating lexicographers as linguistic quality controllers—if a word or sense makes it into the dictionary that’s taken as a stamp of approval, a vote of confidence, a Papal Bull proclaiming that we should all be using/understanding the word that way—we give them and their products more authority than they deserve.

The view that dictionaries are or should be arbiters rather than just recorders of usage has a long history (interestingly discussed in Anne Curzan’s book Fixing English), and you can still see it reflected in things like Merriam-Webster’s periodic reports on its most popular online ‘look-ups’. The words M-W’s users look up tend to reflect what’s currently in the news: this summer, for instance, the solar eclipse prompted a spike in look-ups for eclipse-related terms like ‘penumbra’, while the ongoing drama of the Trump presidency had people searching out words like ‘impanel’ and ‘recuse’. In these cases, involving technical terms drawn from the registers of science and law, we can imagine people who were previously unfamiliar with a word going to the dictionary’s website to find out what it meant, or maybe how it was pronounced or spelled. But in other cases that’s an unlikely scenario. It’s hardly plausible that all the people who looked up ‘science’ during the row about Trump’s policy on climate change, or those who looked up ‘fact’ after Kellyanne Conway’s infamous reference to ‘alternative facts’, were just trying to remedy their ignorance about the meaning, spelling or pronunciation of these common words. More likely they were engaged in some kind of argument about what ‘science’ did or didn’t cover, or whether ‘alternative facts’ was a contradiction in terms, and had turned to the dictionary for an authoritative ruling.

I’m sure we’ve all at some point been involved in a political argument which someone has proposed to settle by looking a word up in a dictionary. But this will never definitively settle it, because the meanings of words (or at least, the sorts of words that provoke arguments) are always variable and contested; and anyway what you’re arguing about isn’t ultimately the words themselves, it’s the differing ideologies which lie behind the competing senses. Lobbying lexicographers on behalf of your preferred definition is fighting a political battle by proxy. What you need to do to win the battle is change the real-world usage of the word in question (something that will usually go along with  other, nonlinguistic social changes). If the dictionary definition is the only thing that shifts, your victory will be purely symbolic.

You might be thinking: but if people with a political agenda manage to change the definition given in dictionaries, won’t that in itself have an impact on real-world usage? In some cases the answer may be ‘yes’, but only if we’re talking about the sort of obscure word which is typically acquired through instruction rather than through the experience of hearing words used in context. ‘Woman’ is not that kind of word. It’s a basic item of English vocabulary, one of the thousand most common words listed in Collins’s dictionary.

If every dictionary in the world changed its definition of ‘woman’ tomorrow, that still wouldn’t stop future generations from understanding and using it to mean ‘adult female human’. That meaning, still the dominant one, will survive because it will continue to be acquired by children in the course of their everyday interactions. Whether they will also acquire the identity-based meaning is another question, and the answer to it doesn’t depend on the dictionary definition of ‘woman’ either: they’re more likely to be taught it in school, or to encounter it in the media, than to learn it by looking up ‘woman’ in a dictionary.  And if kids are learning the new sense from other sources, keeping it out of the dictionary will do nothing to halt its spread.

I’m not suggesting that all arguments about word-meaning are pointless (if I thought that I’d be in the wrong line of work); what I’m questioning is the equation of a word’s meaning with its dictionary definition, and the associated belief that if you can persuade a dictionary to change (or not change) a definition, you have thereby changed (or safeguarded) the language itself. This attitude to dictionaries is another interesting example of how conservative, when it comes to language, political radicals can be. It’s no good petitioning the King (especially as he abdicated long ago). The struggle for meaning is a grassroots campaign.

Advertisement