Dissing the dictionary

This week the Guardian writer Emine Saner drew my attention to a petition on Change.org asking Oxford University Press, the publisher of the Oxford Dictionary, to change its entry for the word ‘woman’.

od def woman

The petition condemns this entry as ‘unacceptable in today’s society’, noting that many of the synonyms it gives for ‘woman’ are derogatory, and the illustrative examples are variations on old sexist themes:

A ‘woman’ is subordinate to men. Example: ‘male fisherfolk who take their catch home for the little woman to gut’, ‘one of his sophisticated London women’.

A ‘woman’ is a sex object. Many definitions are about sex. Example: ‘Ms September will embody the professional, intelligent yet sexy career woman’, ‘If that does not work, they can become women of the streets.’

‘Woman’ is not equal to ‘man’. The definition of ‘man’ is much more exhaustive than that of ‘woman’. Example: Oxford Dictionary’s definition for ‘man’ includes 25 ‘phrases’ (examples), ‘woman’ includes only 5 ‘phrases’ (examples).

The creator of the petition, Maria Beatrice Giovanardi, has found that Oxford’s entry is not an isolated case. In a piece entitled ‘Have you ever Googled “woman”?’ she examines the entries in several widely-used dictionaries and points out similar problems with all of them. Her petition targets Oxford, however, because as well as being, in her view, the worst offender, it’s also got a market advantage: it’s the dictionary you get with Apple’s products and the one that pops up first in searches on Google, Yahoo and Bing.  Her petition calls on the Press

  1. To eliminate all phrases and definitions that discriminate against and patronise women and/or connote men’s ownership of women;
  2. to enlarge the definition of ‘woman’ and equal it to the definition of ‘man’;
  3. to include examples representative of minorities, for example, a transgender woman, a lesbian woman, etc.

As I told Emine Saner (whose own piece you can read here), I’ve got mixed feelings about this campaign. I do think the entries Giovanardi reproduces are terrible, and I’ll come back to that later on. But the petition is based on ideas about dictionaries–how they’re made, what they aim to do, and how they’re used–which, to my mind, are also a problem. I’ve written about this before, but let’s just recap.

Modern dictionaries are descriptive: their purpose isn’t to tell people how words should be used, but rather to record how words actually are used by members of the relevant language community (or more exactly, in most cases, by ‘educated’ users of the standard written language). What’s in the Oxford Dictionary entry for ‘woman’ does not represent, as Giovanardi puts it, ‘what Oxford University Press thinks of women’. The dictionary is essentially a record of what the lexicographers have found out by analysing a large (nowadays, extremely large—we’re talking billions of words) corpus of authentic English texts, produced by many different writers over time.

I’m not trying to suggest that dictionaries compiled in this way are beyond criticism: they’re not, and that’s another point I’ll come back to. But a distinction needs to be made between the producers’ own biases and biases which are present in the source material they use as evidence. A lot of the sexism Giovanardi complains about is in the second category: it’s the result of recording patterns of usage that have evolved, and still persist, in a historically male-dominated and sexist culture.

For instance, one reason why the Oxford entry for ‘man’ is longer than the one for ‘woman’ is that men have been (and still are) treated as the human default. Men are both people and male people; women are only women. The historical reality of sexism also makes itself felt in the presence of so many degrading and dehumanizing terms on lists of synonyms for ‘woman’. These reflect the social fact that women have been sexually objectified in ways that men have not; they have also been treated, in life as well as language, as men’s appendages or possessions. It’s not even two centuries since that was their status in English law.

If the vocabulary of a language reflects its users’ cultural beliefs and preoccupations, then it’s no surprise that many English words and expressions denoting women highlight their dependence on and inferiority to men, their physical appearance and their sexual availability. (In an early feminist study of this phenomenon, the linguist Julia Penelope Stanley identified over 200 words for ‘woman’ that meant, or had come to mean, ‘prostitute’.)  It’s true dictionaries have a bias towards the usage of men from the most privileged social classes–the section of society that has historically had most access to the means of representation–but in this case it’s not obvious that a more balanced sample would paint a totally different picture. Sexism and misogyny have never been confined to a single class, or indeed a single sex.

The demand to eliminate ‘phrases and definitions that discriminate against or patronise women’ (or other subordinated groups: there have been similar campaigns in the past relating to entries where the issue was race, ethnicity or faith) implies that dictionaries should sanitise the reality of word-usage which is sexist, racist, anti-semitic or whatever, in the hope that this gesture will help to make a better world. Lexicographers tend to be wary of that suggestion, since it goes against their descriptivist principles. They also reject the popular belief that merely by including a certain word or definition they’re somehow endorsing it, giving it a degree of currency and acceptability which it would lack if it were not in the dictionary.

But while I agree this belief misunderstands the aims and methods of modern lexicography, I also think there’s something disingenuous about the standard ‘don’t shoot the messenger’ response. If the public at large treat dictionaries as arbiters of usage, then in practice, whether their producers like it or not, they do have authority, and therefore some responsibility to reflect on how it should be used.

In some areas it’s clear they have reflected, and concluded that some sanitising is justified. You may have noticed, for example, that Oxford’s entry for ‘woman’ doesn’t propose ‘cunt’ as a synonym: that can’t be because it isn’t used as one, so presumably it’s been excluded on the grounds of its offensiveness. In recent years there’s also been a move to eliminate the casual racism and homophobia which were features of some older dictionary entries (today you’ll find fewer references to ‘savages’ or ‘unnatural acts’). Casual sexism, by contrast, has mostly escaped the cull. I may not agree with Giovanardi’s proposed solution, but I think she’s right to raise this as a problem.

What solution would I be in favour of, then? Essentially, a more context-sensitive, ‘horses for courses’ approach. Different kinds of dictionaries serve different purposes and audiences: there are some cases where I think it would be wrong to sanitise the facts of usage, and others where nothing important would be lost by being a bit more selective.

The Oxford English Dictionary (OED)—the massive historical dictionary which is Oxford’s flagship product—contains an entry for ‘woman’ which is an absolute horror-show of sexism: the senses listed include things like

3a. In plural. Women considered collectively in respect of their sexuality, esp. as a means of sexual gratification.
4. Frequently with preceding possessive adjective. A female slave or servant; a maid; esp. a lady’s maid or personal attendant (now chiefly hist.). In later use more generally: a female employee; esp. a woman who is employed to do domestic work.

Further down there’s a long list of delightful idioms containing ‘woman’ (‘woman of the night’, ‘woman of the streets’, ‘woman of easy virtue’) and a section full of even more delightful proverbs and sayings (‘a woman, a dog and a walnut tree/ the more you beat them the better they be’). And many of the illustrative examples, even some of the most recent ones (this entry was last updated in 2011), are as terrible as you’d expect. Any feminist who reads the entry from beginning to end will want to go out and burn the world down. But in this case I believe that’s as it should be. The OED is designed primarily for scholars, and it’s an invaluable repository of cultural as well as linguistic history. Hideous though the ‘woman’ entry is, cleaning it up would do feminism a disservice: it would erase evidence that needs to be preserved for our own and future generations.

The Oxford Dictionary, however, as accessed via iPhones and search engines, is (or should be) a completely different animal. I said earlier that I thought the entry reproduced in Giovanardi’s petition was terrible, and what I meant by that was not only that it’s terribly sexist, but also that it seems poorly designed to meet the needs of those who use this type of dictionary. The basic definition of ‘woman’ is unremarkable (albeit redundant, since this is a high-frequency word that English-speakers generally acquire before they can read), but the thesaurus section is stuffed with archaic terms which would hardly be usable in any contemporary context. Who, in 2019, calls women ‘besoms’, ‘petticoats’ or ‘fillies’?  And if you did encounter one of these expressions in a novel or period drama, the context would make clear what it meant. So what is this list of synonyms good for? What pay-off would its makers cite to justify its rank misogyny?

But in any case, does anyone in real life use their phone, or Google, to look up common words like ‘woman’? After speaking to Emine Saner, I rather belatedly began to wonder what people do use dictionary apps for, and I put out a call on Twitter asking people to tell me if they used them, what they did with them, and what the last word they’d looked up was. Leaving aside the people who replied that they only ever used either bilingual dictionaries or monolingual dictionaries for languages they weren’t native speakers of, the responses clustered in three main categories. In order of frequency, these were

  1. Checking the meaning and/or spelling of low-frequency words (examples included ‘obviate’, ‘noctilucent’ and ‘eschatology’)
  2. Using the thesaurus element of an entry—or an actual thesaurus app—to find synonyms to use when writing
  3. Searching for words while engaged in word-based leisure activities like doing crosswords or playing Scrabble.

One or two people mentioned other uses, like checking the pronunciation of a word they’d only ever met in writing, or looking into the history of a word, or decoding the slang terms used by their children (for this purpose Urban Dictionary was the go-to source). But no one reported looking up basic vocabulary items whose meaning, spelling and pronunciation they already knew. Of course you can’t draw general conclusions from what you’re told by a self-selected sample of your Twitter followers, but it seems possible the Oxford Dictionary entry for ‘woman’ isn’t exerting a malign influence on our attitudes to women, simply because so few of us will ever look at it.

I’m not saying that makes it OK to leave the entry as it is: I do think the dictionary makers should revisit their synonyms and illustrative examples (they could start by getting rid of any expression that no one under the age of 85 has ever uttered). But if I had to make them a to-do list, revising their ‘woman’ entry wouldn’t be at the top. Dictionaries reinforce sexism and gender stereotyping in other ways, which are arguably more pernicious because they’re not so immediately obvious.

An example is the persistent use of sex-stereotyped illustrative examples in entries for words that have nothing to do with sex or gender. I discussed this form of banal sexism in a previous post, prompted by the row that broke out when the anthropologist Michael Oman-Reagan queried the use of ‘a rabid feminist’ in the Oxford Dictionary entry for ‘rabid’. We’ve also got dictionaries for learners of English in which men mop their brows while charwomen mop the floor, and men slip on their shoes while women slip off their dresses. How do these sexist clichés enhance anyone’s understanding of the words ‘slip’ and ‘mop’? Do the entry writers think learners are planning to practise their English on a coach tour of the 1950s?

As I’ve said before, though, my feelings about political activists lobbying dictionaries are mixed: I can see why you might want to put pressure on what are, whether they acknowledge it or not, influential cultural and linguistic institutions, but something that bothers me about this approach (that is, apart from the problems I’ve already mentioned) is its linguistic authoritarianism. It’s buying into the idea that dictionaries can and should lay down the law—that it’s fine for them to prescribe, so long as what they prescribe reflects our own preferences rather than those of our political opponents. I know I’ve used this example before, but let’s not forget that Christian conservatives lobbied dictionaries in an attempt to stop them changing their entries for ‘marriage’ to include the same-sex variety–and progressives applauded when the dictionaries took no notice. In that instance feminists were happy to endorse the principle that dictionaries just record the facts of usage. I don’t think we can reasonably demand that they should be descriptive when it suits us and prescriptive when it doesn’t.

But the deeper problem underlying these contradictory demands is the status we accord to ‘the dictionary’ as the ultimate authority on language. In the past that was something feminists questioned: they were less interested in harnessing the power of what Mary Daly called the ‘dick-tionary’, and more interested in challenging its patriarchal claims to ‘authoritative’ and ‘objective’ knowledge.

During the 1970s, 80s and 90s, a number of feminists produced alternative dictionaries that embodied this challenge in both their content and their form. The compilers of these texts didn’t deny that their selection of headwords, definitions and examples represented a non-neutral point of view; rather they maintained that this was covertly true of all dictionaries (Cheris Kramarae and Paula Treichler’s A Feminist Dictionary defined a dictionary as ‘a word book. Somebody’s words in somebody’s book’.) Their entries emphasised the variability of meaning—words can mean different things for different speakers—and the fact that it’s contested rather than consensual. Though admittedly they were not much use if what you needed was a definition of ‘noctilucent’.

In some ways, as Lindsay Rose Russell points out in her recent history of women’s contributions to dictionary-making, these late 20th century feminist dictionaries anticipated digital-era efforts to democratise and diversify knowledge, as exemplified by Wikipedia and Wiktionary, and (in a different way) Urban Dictionary. They were anarchic, utopian, and they often featured a multiplicity of voices which they made no attempt to homogenise. They used sources that conventional dictionaries neglected, and tried to amplify the voices those dictionaries had marginalised or muted.

In theory the digital revolution offered an opportunity to develop this tradition further, and perhaps to produce feminist alternatives to the mainstream online dictionaries which most people use today. But in practice that hasn’t happened. Digital knowledge-making communities have been dominated by men, and they can be hostile environments for women—witness the many female Wikipedians who say they’ve been bullied or sidelined by men who treat the site as their turf. Any attempt to create a feminist dictionary online would be a magnet for these mal(e)contents, who would either want to take it over or to take it down. Ironically, there may be less space for feminist lexicography in the limitless expanse of cyberspace than there was in the olden days when dictionaries were books.

But although I regret the fading of the more radical spirit that animated projects like A Feminist Dictionary, I’m not completely opposed to the reformist approach. As Lindsay Rose Russell told Emine Saner, ‘we ought to care what definitions are made most readily available and why’–and we have every right to bring our concerns to the attention of the people responsible for those definitions. Even if Oxford’s public response is defensive or dismissive, I suspect that Maria Beatrice Giovanardi’s petition might still prompt discussion behind the scenes. Its demands won’t be met in full: a descriptive dictionary can’t eliminate all sexism from the record of a language in which sexism is so pervasive. But if it makes the producers aware that a lot of people find their entries for ‘woman’ gratuitously offensive, outdated and useless, they may, eventually, make some changes. In this case I think they should, and I hope they will.


Dictionaries, dick-tionaries and dyketionaries

Exactly ten years ago, in June 2005, I was contacted by a man from the British Potato Board. He wanted me, in my capacity as a professor of the English language at Oxford University, to endorse the Board’s campaign to get the expression ‘couch potato’ removed from the Oxford English Dictionary. It gave potatoes a bad name, he explained, by suggesting they were unhealthy, when in fact they were virtually a superfood, packed with fibre and vitamin C. The Board wanted the OED to replace ‘couch potato’ with ‘couch slouch’, which would convey the same meaning without unfairly maligning potatoes.

Initially I suspected this was a wind-up; but then a group of people turned up, dressed in potato costumes, to protest outside the offices of the OED’s publisher, Oxford University Press. Basically it was a publicity stunt: I’ve never been sure how serious they were about getting the dictionary to alter its entry. But even if the aim was just to get media coverage for the health benefits of potatoes, the campaign still traded on the popular belief that dictionaries function as a kind of supreme authority on the existence, validity and meaning of words. As if removing ‘couch potato’ from the dictionary were equivalent to banishing it from the language.

This attitude was on display again last week when it was announced that the most recent additions to the OED included an entry for the word cisgender.  On one hand you had people triumphantly hailing this news as proof that the word, and by extension the concept it denotes, is both real and legitimate.  On the other hand you had people—many of them feminists who dispute the political analysis behind the word—saying that the OED should not be giving legitimacy to such a flawed and biased term.

When I hear this kind of argument I feel pulled in two directions. The linguist part of me wants to defend the OED, explaining–just as I explained to the man from the British Potato Board–that the function of modern dictionaries isn’t to authorize words or make judgments on their value, it’s to document their existence and the way they’re used in the language. The main criterion for adding new words is that they have entered mainstream usage. Cisgender, for instance, used to be a little-known piece of in-group jargon, but now it turns up regularly in publications like newspapers which are aimed at a general audience.

Since the OED is a historical dictionary, whose aim is to chart the development of English vocabulary over time, it’s full of words which are obsolete, arcane, useless, offensive or frankly silly. Like scrolloping, a word that appeared once in the work of Virginia Woolf. Or phlogiston, an 18th century name for a chemical element that never actually existed. Learned discussions of phlogiston completely misrepresented reality, but the word was once in regular use, so the OED records it. It’s now doing the same for cisgender: even if you think the concept is to 21st century gender theory what phlogiston was to 18th century chemistry, there’s no good argument for not including it in a historical dictionary of English.

But if that last paragraph sounds like a straightforward defence of modern dictionary-making as an objective, value-free, scientific enterprise, let me clarify that I don’t think it’s that straightforward. That’s why I said I feel pulled in two directions. I’m a feminist as well as a linguist: I may not agree with those sisters who have complained about the OED’s inclusion of cisgender, but I have no doubt that mainstream dictionaries, including the OED, belong to an androcentric (male-centred) and sexist tradition. Since all major dictionaries rely heavily on past scholarship, the traces of earlier male bias are still highly visible; and some forms of bias have persisted into the present.

Criticizing the male bias of mainstream lexicography was a serious concern for feminists in the 1970s and 1980s, and so was making alternatives to what Mary Daly dubbed the ‘dick-tionary’:

Dick-tionary, n: any patriarchal dictionary: a derivative, tamed and muted lexicon compiled by dicks.

The source I’m quoting here was a product of this alternative tradition: Daly and Caputi’s Websters’ First New Intergalactic Wickedary of the English Language, originally published in 1987. But while Daly’s reputation has kept the Wickedary visible to later generations of feminists, the fact that it was part of a flourishing enterprise of feminist dictionary-making is less well known, as are most of the other dictionaries second wave feminists produced (though many are still available, and I’ve added links where that’s the case).

In 2011, Lindsay Rose Russell published a scholarly article, ‘This is what a dictionary looks like’, which lists 18 examples of feminist dictionaries published between 1970 and 2006 (though the largest number appeared in the 1970s and 1980s). She was prompted to write by the appearance of a weighty tome about the history of lexicography which failed to mention a single feminist dictionary—though it did discuss specialist dictionaries of agriculture, botany, chemistry, law, medicine, seafaring and surnames. This struck Russell as a classic case of feminism being written out of history. Like her article, this post is an attempt to write it back in.

Serious feminist criticism of mainstream/malestream dictionaries began to appear in the early 1970s, when several kinds of systematic male bias were identified. These included

Selection of sources. The pre-existing source material from which lexicographers take their evidence about word-usage (which words are used, and with what range of meanings) has historically been heavily skewed towards texts written by male rather than female authors. While recently some attempt has been made to redress the balance, the overall bias remains an issue.

Androcentrism in selecting and defining words. As more dictionaries have moved online, the pressure to exclude words for reasons of space has been reduced, but in the past editors did have to make judgments on what was important or central and what was trivial or peripheral, and their decisions often reflected a male/masculist perspective. girls dictionarySo did some of their definitions: Alma Graham, who was involved in making a non-sexist school dictionary in the early 1970s, cites one dictionary that defined youth as ‘the part of life between childhood and manhood’.

One word the OED treated for decades as both too peripheral and, we might suspect, too distasteful to include, was lesbian. It has been used in print since the 18th century, so by the OED’s usual criteria it would certainly have merited an entry; but it was not only left out of the original edition (along with a number of other words the Victorian editors preferred to pass over in silence), it was also excluded from the supplement produced to update the dictionary in 1933, though the editor at that point did add the term homosexual. When lesbian finally got an entry, in a further supplement published in 1976, one of the quotations used to illustrate its use was this gem from the work of the writer Cecil Day Lewis:

I shall never write real poetry. Women never do, unless they’re invalids, or Lesbians, or something.

This brings us neatly to another kind of bias feminists criticized in the 1970s:

Sexism in illustrative quotations. The OED isn’t the only dictionary that gives examples (either real quotations or made-up sentences) to illustrate how words are used: this is also a feature of many pedagogic dictionaries, designed for schoolchildren or foreign language learners. And in some of the most popular learners’ dictionaries, reading the examples is like flicking through a catalogue of offensive gender stereotypes.

A student of mine who did a project on this topic found that one of the dictionaries she analysed had a pattern of alternating between male and female references in illustrative quotations: this may have been meant to ensure gender ‘balance’, but the effect was almost comically sexist, as though the examples had been chosen by Benny Hill. In its entry for the verb slip, for instance, the dictionary offered ‘he slipped on his shoes and went outside’ followed by ‘she slipped out of her dress’. Another classic, illustrating the verb mop, had ‘he took out a handkerchief and mopped his brow’ alongside ‘the charwoman had just mopped the linoleum that covered the stairs’. Though they compiled it in the 1990s, the makers of this dictionary apparently felt that what foreign learners needed was a good understanding of the gender roles that prevailed in Britain circa 1953 (also the class relations and the household technology—when did anyone last see a ‘charwoman’ wielding a mop?)

One feminist response to these shortcomings was to try to reform the dick-tionary from within (as with the nonsexist school dictionary discussed by Alma Graham), but there was also a flurry of explicitly oppositional feminist dictionary-making. Its products were varied in their form and political aims. Daly and Caputi’s Wickedary is part of a ‘utopian’ subgenre which also includes Monique Wittig and Sande Zeig’s earlier Lesbian Peoples: Material for A Dictionary (1979). Other feminists focused on recording and defining feminism’s own terminology (e.g. Joreen and Marleen Dixon’s Dictionary of Women’s Liberation (1970)), offering alternatives to sexist words (e.g. Rosalie Maggio’s The Nonsexist Wordfinder (1987)), or conversely listing and critically discussing them (e.g. Jane Mills’s Womanwords: A Vocabulary of Culture and Patriarchal Society (1989), which explores the history of misogyny by looking at the evolution of terms like frigid, hysteria and slut).

There were also examples like Cheris Kramarae and Paula Treichler’s A Feminist Dictionary (1985), which offered a direct challenge to the dick-tionary’s authority and its claims to objectivity. Kramarae and Treichler made no such claims. In some entries they defined a word (e.g. ‘powerlessness’) by simply juxtaposing quotations in which it had appeared, all taken from the speech and writing of feminist women. In other cases they offered a gloss which was designed to make visible the biases rendered invisible by the standard, supposedly neutral definition:

CUCKOLD. The husband of an unfaithful wife. The wife of an unfaithful husband is just called wife.

The first part of this does not differ greatly from what you might find in a non-feminist dictionary entry for cuckold, but the second part pointedly draws attention to the absence of any parallel term for a woman whose husband is unfaithful, and so makes a (non-neutral) comment on the underlying sex inequality.

In 1992 A Feminist Dictionary was reissued under the new title Amazons, Bluestockings and Crones. Lindsay Rose Russell calls this ‘a move from suffragist to supplement’: it makes the book sound less like a challenge to the conventions and the authority of the mainstream product, and more like ‘a harmless helpmate to the lexicographical tradition, a dictionary of boutique terms (Amazon, bluestocking, crone)’. Later examples continue the ‘harmless helpmate’ theme, bearing titles like From the Goddess to the Glass Ceiling (1996) and Wimmin, Wimps and Wallflowers (2001). There are also collections of words coined by and for a specific community or subculture, like the Chicago-based online Dyketionary. These are not aiming to talk back to the dick-tionary, to criticize its sexism or question its objectivity. They are, as Russell says, supplementary lists of ‘boutique terms’.

It could be argued that the function of the second-wave feminist dictionary has been superseded by more recent developments. On one hand, new technology has enabled mainstream dictionaries to expand their coverage and widen the range of sources they trawl through, thus reducing some of the old biases (it’s also worth pointing out that many mainstream lexicographers today are women: they may or may not be feminists, but they aren’t bearded Victorian patriarchs who find women alien and faintly repulsive). On the other hand, new media have democratised the process of recording and defining words, bringing us completely crowd-sourced examples of amateur/popular lexicography like Urban Dictionary, or Dyketionary.

But I wouldn’t make that argument myself, or at least not unreservedly. The response to Oxford’s announcement about cisgender shows that the authority of the established, mainstream dictionary has not been superseded. It also shows that there are still gaps in the public understanding of what lexicographers do, and how they go about it. If one consequence is that feminists sometimes make specific criticisms of dictionaries that (in my view) are misplaced, another is that they don’t make more general criticisms that are still highly pertinent.

That was one of the functions of feminist dictionaries. They were much more than boutique wordlists. Outdated though much of their content is, they offer, as Lindsay Rose Russell says, ‘a usable past’ for feminists reimagining the dictionary today.