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.

The return of ‘female email’

Do you remember your 2016 new year’s resolution? Was it to get more exercise, maybe? Give up the demon drink? Spend less time on Facebook and more with your real-life friends? Or was it, perhaps, to send ‘more effective email’, as recommended by the developers of an app called ‘Just Not Sorry’?

This app was intended to empower working women by encouraging them to delete ‘just’ and ‘sorry’ from their emails. If you hovered your mouse over one of the offending words you’d see a pop-up message from a communication ‘expert’, like “just” demeans what you have to say’, or ‘using “sorry” frequently undermines your gravitas’.

But even this ingenious invention seems not to have fixed women’s email problem. Last month the Telegraph ran a piece entitled ‘Sorry to bother you: how women can stop writing emails “like a girl” at work’. It begins with what the writer claims is a typically female email:

Hello! Hope you’re well and that you’re having a lovely week! So sorry to bother you, but I was just wondering if you could read the below article I’ve written? No worries at all if not – I know you must be super busy. Thank you so much for your time! Best wishes.

These 50-odd words are like a whistle-stop tour of women’s language stereotypes from the last half-century: they include a ‘just’, two ‘sos’, a ‘sorry’, a ‘lovely’, a superpolite indirect request (‘I was just wondering if you could…?’), and a veritable forest of exclamation marks. If the message had only ended with a smiley face emoji we could all have shouted ‘House!’

This much-maligned email style is generally assumed to be something women acquire in their teenage years, carry with them into the workplace, and need remedial instruction to get rid of. But last week a piece on Canada’s Global News website turned that assumption on its head. According to the reporter Meghan Collie, women in workplaces around North America are being told by their bosses, not to stop writing email ‘like a girl’, but on the contrary, to make their emails more girly.

Take Carlee Barackman, a former employee at a tech startup in Detroit who describes her email style as ‘short and to the point’:

Barackman thought she was emailing like everyone else — until her CEO pulled her aside to talk about her “harsh” language… While he didn’t explicitly ask her to soften her writing style, Barackman said it was implied, and she decided against it. “I had work to do and I didn’t want to spend extra time trying to convey my bubbly personality in an email,” she said.

Sometime later, Barackman replied to an email with “okay, thanks,” — no punctuation, no emojis — and her CEO called her out. Barackman agreed to try and “lighten it up,” but she didn’t really know what that meant. It was salt on the wound when Barackman saw an email thread between her male colleagues with writing nearly identical to the style that got her in trouble.

“I remember sitting down at my desk and having no idea who to ask about how to email like a woman. Is emailing like a woman even a thing?” she said. “I felt worried that, by adding extra fluff to an email, I would appear unprofessional, and also worried that, if I kept my replies short and direct, everyone would assume I was angry,” she said.

Carlee Barackman was only one of the many women who responded to the call Meghan Collie put out on Twitter: ‘Do you work in an office? Have you ever felt pressure to use emojis or exclamation points to soften your message?’ Affirmative answers flooded in, and they suggested that emojis and exclamation marks were only the tip of the iceberg.

I’ve been told numerous times to soften up my emails. I use smilies and ! In almost every email, and say please and thank you so much it would be weird if we were in person. I also throw in “just” a lot.

I have no idea what you’re talking about [followed by a screenshot of an email that reads “Awesome! I have been in and made the required edits! Thank you 😊]

I have been told to soften my tone, I notice that men and some women that they favor for whatever reason, are allowed to be rude, abusive and abrupt by email or message. The rest of us…get our tone policed. I have used emoji or “if that makes sense” a lot

I think it also comes down to what men can get away with in emails that women can’t — I once had a male manager write in all caps to get his point across.

I find men can get away with being short, rude and degrading but as soon as a woman does it, they get pulled in for it.

I hate exclamation points. Absolutely hate them. …But yes, I feel forced to use them to blend in & be polite! All the time! I’m so excited about absolutely nothing & here’s the punctuation to prove it!

I have consciously been removing exclamation points and emojis, apologies and just-a-quick-question from my emails for years. Why diminish yourself when you are simply communicating?

I read about how women apologize a lot in emails. Especially with saying the word “just”. I noticed how often I did it and it has been a LONG JOURNEY to remove those things from my email repertoire! No need to excuse myself for doing my job.

I confess I was taken aback by these vignettes.  Although I’ve spent a fair bit of my life observing the policing of language at work, the verbal hygiene practices described in this Twitter thread stand out for both their intrusiveness and their pettiness: managers scrutinizing internal emails in minute detail, and pulling individual employees aside (especially, it seems, if they’re female) to warn them about their tone. How is this a productive use of anyone’s working time?

The women who responded to Meghan Collie were also, for the most part, critical of the practices they described, often stating explicitly that the style they felt obliged to adopt did not reflect their own preferences. Some women clearly resented the tone-policing of their email, and a few reported actively resisting it. Many of these resisters invoked the competing, ‘Just Not Sorry’ genre of verbal hygiene to justify their rejection of ‘weak’ and ‘demeaning’ language. The irony of this–using one kind of sexist bullshit to fight another–isn’t lost on me, but I can’t really quarrel with the perception of ‘Just Not Sorry’ as the lesser of the two evils. ‘Empowerment’ may be a weasel word, but it’s surely preferable to self-abasement.

The ‘Just Not Sorry’ message has had a lot of media exposure because it resonates with the aspirational, ‘lean in’ ethos of the media outlets which commission pieces like the Telegraph’s. Precisely because it can’t so easily be spun as ’empowering’, the ‘Softly Softly’ approach hasn’t attracted the same attention. (I notice that no one has developed an app called ‘Soften Your Message’, or ‘Everything Is Awesome!’, with pop-up messages like ‘if you don’t add a smiley face people will think you’re angry’, or ‘do you love your job? Then say it with !!!’) But despite its low cultural profile. ‘Softly Softly’-style language policing is evidently a reality in many workplaces. What, we might wonder, is this about? Why are women–and, to some extent, men too–being instructed to ‘soften’, ‘lighten up’ or add ‘extra fluff’ to their emails?

On closer inspection, what Meghan Collie and her correspondents call ‘message softeners’–things like exclamation marks, emoji, hedges like ‘just’ and stock phrases like ‘sorry to bother you’ or ‘if that makes sense’–seem to serve two main purposes, which can in turn be related to two overarching norms of workplace communication.

First, there seems to be a clear norm prescribing the explicit expression of positive affect and high involvement. It’s not just that negative messages are frowned on: neutral, low-key formulations like Carlee Barackman’s ‘okay thanks’ are not acceptable either. This is what motivates the liberal use of exclamation marks and emoji (or more exactly, a subset of emoji–smileys and thumbs-up signs rather than, say, piles of poo). As conventional signifiers of excitement, enthusiasm, happiness or satisfaction, they inject a note of unambiguous positivity into even very short and banal communications. Accentuating the positive is also the function of phatic formulas like the Telegraph writer’s ‘hope you’re having a lovely week!’ and the hyperbole of responses like ‘Awesome!’ The message is something like, ‘I want you to know I’m thrilled to be at work, delighted to be communicating with you and eager to show I value your contribution’.

The second overarching norm complements the first: it could be glossed as ‘minimize the risk of conflict or offence by avoiding anything that could conceivably be read as angry, critical, overbearing or even just a bit inconsiderate’. This is the purpose served by formulas like ‘please’ and ‘thank you’, ‘sorry to bother you’ and ‘just a quick question’ (implying: ‘I know your time is precious’). It’s also the point of appending ‘if that makes sense’ to, for instance, a series of instructions or a piece of critical feedback. Here what’s being ‘softened’ is the presumptuousness of judging others or telling them what to do.

As some readers will doubtless have noticed, the two norms just outlined call for, respectively, the use of ‘positive’ and ‘negative’ politeness. (These terms are taken from the work of politeness theorists Penelope Brown and Stephen Levinson: in their model, positive politeness addresses the desire every person has to be approved of or cared about (prototypical positive politeness formulas include ‘have a nice day’, and ‘congratulations!’), while negative politeness addresses people’s desire not to be imposed on (prototypical formulas include ‘excuse me’ and ‘sorry’).)  As I’ve explained in previous posts, one of my main beefs with the ‘Just Not Sorry’ brigade is their insistence on treating politeness features as ‘fluff’ or ‘clutter’, things that detract from the message and so impede communication, when in fact they’re essential elements of any interaction between humans. Politeness per se is not a problem: taking out all the ‘justs’ and ‘sorries’ is only good advice if your ambition is to sound like a jerk. However, two things about the ‘Softly Softly’ approach do strike me as more problematic.

One problem is that the rules are so inflexible. In everyday life, the way we use linguistic markers of politeness reflects our assessment of how seriously what we’re saying might hurt, offend or impose on the other person. You wouldn’t hedge a request to pass the salt in the same way you’d hedge a request to lend you £100; you wouldn’t congratulate someone as enthusiastically on winning a pub raffle as you’d congratulate them on winning a Nobel Prize. In ‘Softly Softly’ world, however, everything gets the same ‘I’m so excited’ or ‘I’m so sorry’ treatment: as some of Meghan Collie’s correspondents observed, maintaining this high level of excitement or solicitude can be exhausting, and it can also come across as quite bizarre.

The other striking thing is the emphasis placed on expressing positive feelings, about everything and to everyone. In workplaces I do think that’s a novel development–particularly if we’re talking about internal back-office communications (accentuating the positive has a longer history in customer service). And what’s behind it, I would argue, is a combination of recent changes in workplace culture and innovations in digital communication.

Over the last 30 years, many workplaces have become less formal and overtly hierarchical, and more focused on collaborative teamwork. In the current era of precarity, companies also tend to have fewer permanent employees and more short-term contract staff. Arguably, these conditions provide fertile ground for things like the demand to accentuate the positive in dealing with co-workers (which displays your credentials as a ‘team player’) and the pressure to display enthusiasm for routine tasks (if you appear bored or disengaged you’re potentially giving your employer a reason not to renew your contract).

At the same time, more and more workplace interactions that would once have been conducted face-to-face have moved online. Email, though still available for the purposes it originally served in business contexts (sending the digital equivalent of letters and internal memos), has also become a medium for co-workers to ask each other quick questions, give brief reports and engage in rapid-fire problem-solving interactions. And what seems to have happened is that the workplace email has borrowed some of the strategies developed for text-based interaction outside work (e.g. on social media and via instant messaging apps), such as the repurposing of punctuation marks to signal affect. (As any teenager will tell you, not putting an emoji or a ! at the end of a text message risks coming across as angry; ending texts with a traditional full stop is rude because it signifies disapproval–though the students who made me aware of this say they try not to judge clueless old people like their mothers too harshly for this offence.)

Imported into the workplace, however, these strategies can create problems that don’t arise, or not so markedly, in other contexts. Some people find email messages larded with emoji and exclamation marks contextually inappropriate–too informal for professional settings, or too personal for interaction with non-intimates. Others find this mode of expression insincere—and not without reason, since at work you’re very likely to be communicating feelings or attitudes you don’t actually have, to people who also know you’re faking, because they’re doing the same thing themselves. (Has anyone ever read a message like ‘I’m so excited for this afternoon’s meeting!!!’ and taken it as a faithful reflection of the writer’s true feelings?)

In principle, the new workplace norms apply to everyone, men as well as women: one man told Meghan Collie that ‘In a previous role, I was told to be “20% friendlier” in my emails and to soften them with smileys’. In practice, however, many contributors to the thread believed that women’s language was more heavily policed than men’s. Whereas men’s failure or refusal to comply with the rules was frequently tolerated (even, reportedly, when this involved such gross breaches as ranting at length in all caps), women could rarely get away with even slight deviations from the prescribed style.

This double standard isn’t hard to explain. The new workplace verbal hygiene is about fostering co-operation and maintaining harmonious relationships by paying solicitous attention to people’s feelings–a responsibility that has been assigned to women since time immemorial. Women are thought to be ‘naturally’ caring, more emotionally expressive than men and more sensitive to others’ needs. We expect them to do more emotional caretaking, we hold them to higher standards, and we punish them more severely when they fall short.

But as depressing as all this is, the Twitter comments quoted earlier give me hope. They show women aren’t just sucking it up: they are critical of the linguistic demands made in their workplaces, and in some cases they are refusing to ‘soften their message’. This rejection of sexist bullshit has my full and unequivocal support. Rise up, sisters: you have nothing to lose but your !!! 😊😊–if that makes sense.

The year in language and feminism, Part II: selected reading

I created this blog primarily as a vehicle for my own thoughts and opinions, but what I write for it is always informed by other people’s research, and by ideas I’ve encountered in other people’s writing. So, to complement my recent review of the year, I’d like to share ten things I read in 2017 which I found interesting, informative and thought-provoking—and which aren’t too technical to be accessible to non-specialists.

Four books

Mary Beard, Women and Power: A Manifesto. A short book which takes the long view on the silencing of women in patriarchal societies.

Emma Jane, Misogyny Online: A Short (and Brutish) History. An Australian journalist turned academic researcher examines the development and impact of online misogyny, and its characteristic linguistic register ‘Rapeglish’, from 1998 to the present.

Angela Nagle, Kill All Normies: Online Culture Wars from 4Chan and tumblr to Trump and the alt-right. Before anyone was talking about the ‘alt-right’, Angela Nagle was investigating the online subcultures from which it emerged, tracking the people involved, the platforms they used, the political positions they espoused and—from a linguist’s perspective most interestingly—the evolution of their distinctive communication style. This isn’t as distinctive as we might think: it has much in common with earlier celebrations of transgression (‘kill all normies’ is reminiscent of Baudelaire’s ‘il faut épater les bourgeois’), and its emphasis on men rebelling against the domesticating influence of women recalls the leftist counter-culture of the 1960s (think Jack Nicholson in One Flew Over the Cuckoo’s Nest). What this shows, Nagle argues, is that we shouldn’t equate being transgressive with being politically progressive. She thinks opponents of the ‘alt-right’ need to take a critical look at their own style of discourse.

Jennifer Sclafani, Talking Donald Trump. Another short book in which an interactional sociolinguist analyses Donald Trump’s use of spoken language during the contest for the Republican nomination. Sclafani doesn’t say much about Trump’s performance of masculinity (which became more salient after he won the nomination and was pitted against a female opponent, Hillary Clinton), but what she does do, by concentrating on small but interactionally significant details, is get beyond the linguistically superficial received wisdom (‘he’s inarticulate/ can’t construct a proper sentence/ has a vocabulary as small as his hands’) to show what’s actually distinctive (and effective) about Trump’s style of public speaking.

Six shorter reads

Language, gender and politics

Unsurprisingly, 2017 produced many reflections on the outcome of the 2016 presidential election, and one issue some of these reflections addressed was the role played by gendered language in shaping responses to the candidates. Among the most intriguing approaches to the question was a dramatic experiment asking ‘What if Donald Trump and Hillary Clinton had swapped genders?

Speaking while female in the workplace

Though working women in 2017 continued to be lectured about their dysfunctional ‘verbal tics’, the idea that inequality in the workplace might not be the result of women’s own linguistic shortcomings appears to be gaining more traction. The research reported in ‘A study used sensors to show that men and women are treated differently at work’ led the researchers to conclude that the problem is ‘bias, not differences in behavior’.

Representing violence against women

Watching the TV adaptation of The Handmaid’s Tale, which was one of the feminist cultural events of the year, prompted Emma Nagouse, who researches Biblical and contemporary rape narratives, to write ‘Handmaids and Jezebels: anaesthetising the language of sexual violence’, about the way language is used to normalise sexual violence and exploitation in the fictional world of Gilead. Later in the year it would become apparent that language serves a not dissimilar purpose in our own world. In ‘The complicated, inadequate language of sexual violence’, Constance Grady reflected on the difficult linguistic choices writers face in reporting women’s experiences of sexual harassment.

Language, gender and artificial intelligence

There was a steady stream of commentary this year on the rise of intelligent machines and what it might mean for the future of humanity. A question of interest to feminists is whether the Brave New World of AI will look any less sexist than what preceded it. In her short but pithy ‘What is a female robot?’, Gia Milinovich asked what it means to treat a  machine as ‘female’. Another memorable piece about the way gender affects human-machine relationships was ‘Siri is dying. Long live Susan Bennett’. Susan Bennett is the woman whose recorded voice was used, without her knowledge, to create the first version of Apple’s virtual assistant Siri. There’s nothing feminist about the writer’s take on her story, but for a feminist reader it contains plenty of food for thought. You could think of it as a Pygmalion narrative for the 21st century, set in a technologically advanced world where women are still seen as raw material to be shaped and improved on by male ingenuity.

Bonus: something to listen to

One of my professional sheroes, the cognitive neuroscientist Sophie Scott, gave 2017’s Royal Institution Christmas lectures for young people. In the run-up to the lectures she made this podcast, which is interesting on a range of frequently asked questions about language, evolution and the brain, and includes some trenchant debunking of  myths about male-female differences.

As Sophie Scott observes, challenging popular beliefs about men and women is an uphill struggle. Though I’ve only mentioned a few by name in this post, I want to salute all those women (and men) who have, nevertheless, persisted.

 

 

 

On being explicit

Note: towards the end of this post there are some examples of sexually graphic and threatening language.

It’s almost exactly a year since I first read Emma Jane’s book Online Misogyny: A Short (and Brutish) History. It does exactly what it says on the tin: in just over 100 pages it tracks the development of online misogyny from the late 1990s to the present. And it doesn’t spare us the details. On the very first page we’re given several examples of the distinctive register Jane calls ‘Rapeglish’. To the question she knows some readers will be asking–‘why didn’t you give us a content warning?’–she replies that there was no warning for the women these messages were sent to.

Jane believes that if we’re serious about tackling online misogyny, we need to know what it looks like and what it feels like:

We must bring it into the daylight and look at it directly, no matter how unsettling or unpleasant the experience may be.

This point doesn’t just apply to online abuse. In recent weeks, sexual harassment has been high on the mainstream news agenda; and this has sparked debate on what kind of language to use in reporting it.

As regular readers may recall, in early November I published a post criticising the mainstream media for their endless repetition of the formulaic phrase ‘inappropriate behaviour’–a bland, all-purpose euphemism whose effect is to minimise the seriousness of the issue. I feel the same about another media favourite, ‘sexual misconduct’. This is slightly less mealy-mouthed than ‘inappropriate behaviour’ (since it doesn’t totally erase the sexual element), but it’s still an affectless, catch-all term which allows us not to look directly at what the perpetrator actually (or allegedly) did.

Not long after I wrote my post, Vox published a piece entitled ‘The complicated, inadequate language of sexual violence’, in which the journalist Constance Grady laid out the dilemma she faces when reporting on sexual harassment:

You can make your language clinical but vague, or you can make it graphic but specific. … I have found that the less specific my language is, the more invisible the violence becomes. But I also worry that the more specific I get, the more sensationalized my language feels.

There is no easy solution to this problem. Grady doesn’t want to downplay the violence, but being specific in this context means being sexually explicit, and that can cause problems of its own:

A survivor…could easily be triggered; even if you’re not a survivor, reading multiple graphic images…can be emotionally trying or even numbing. Such descriptions can also swing the other way, and become luridly fascinating in a way that feels exploitative, as if I am writing pornography rather than reporting on a sexual assault case.

‘Respectable’ mainstream news outlets do generally try to avoid sexually explicit language–not because they share Grady’s feminist concerns, but for more traditional reasons of ‘taste and decency’. Hence their fondness for such bland, generic formulas as ‘inappropriate behaviour’ and ‘sexual misconduct’.

This isn’t just a journalists’ dilemma, it’s also a longstanding problem for feminist campaigners on the issue of sexual violence. To make women’s experiences speakable you have to name them; but if you want them to be speakable in a court of law, or in the New York Times (whose masthead famously proclaims that it reports ‘all the news that’s fit to print’), the words you use have to be acceptable, not (porno)graphic or otherwise offensive. That, however, increases the risk that over time they will be depoliticised, used in such vague, euphemistic or trivialising ways that they no longer serve their original feminist purpose.

In October the New York Times published an op-ed piece which made exactly this argument about the current usage of ‘sexual harassment’. The author asserted that since it first acquired mainstream currency in the mid-1970s, this originally feminist coinage had been ‘co-opted, sanitized [and] stripped of its power to shock’. Corporations, she argued, had taken ‘a term that once spoke to women about revolution’ and made it into a piece of ‘corporate-friendly legalese’, the stuff of HR manuals and training courses designed less to advance the cause of workplace equality than to protect employers from lawsuits.

This criticism is all the more damning if we consider the identity of the critic. The words I’ve just quoted are the words of Lin Farley—the woman who literally wrote the book on sexual harassment at work, and who is credited with introducing the term into mainstream public discourse. Farley now wants feminists to reclaim and re-politicise it. How does she think we should do it?

By talking about the details — every time. By making the reality of what it looks like clear. …In this context, the most valuable part of the exposures of men like Harvey Weinstein and Roger Ailes may lie in the excruciating, unforgettable details. This is where the heart of understanding the truth of sexual harassment resides.

Emma Jane is also in favour of talking about the details. In Online Misogyny she reproduces not only ‘a multitude of examples, but….a multitude of unexpurgated examples’. Her insistence on quoting abusers’ own words reflects her belief that when academics or the media skate over the details–when they simply describe messages as ‘graphic’ or ‘threatening’, without repeating their actual content–they are unwittingly contributing to the problem. The refusal to be explicit tells women who are experiencing abuse that the details should not be aired in public; it also allows people who have not experienced abuse to go on believing that it’s really not that serious–that women who get upset are just ‘princesses’ who need to ‘toughen up’.

Women who have been targets of abuse have made similar points themselves. The classicist Mary Beard, for instance, who was viciously attacked after a TV appearance in 2013, told an interviewer:

You never know what it’s like, because no mainstream paper will print it, nobody on the radio will let you say it…

Though Beard had received numerous rape and death threats, along with other sexually graphic messages, what the media reports foregrounded was the abusers’ insulting comments on her appearance. Consequently, she said, her concerns were decried as trivial:

It came to look as if I was worried that they’d said I hadn’t done my hair.

A few months later there was a sustained attack on Caroline Criado-Perez, the feminist who had successfully campaigned for a woman to be represented on a Bank of England banknote. The abuse Criado-Perez experienced was so intense and so threatening, two of those responsible would eventually be sent to prison. But while it was actually happening, as she recalls in her book Do It Like A Woman, the news reports ‘spoke vaguely of online abuse’, and whenever she was interviewed she was warned to keep her language ‘polite’.  ‘I was forced’, she writes,

to shield members of the public from something from which no one had been able to shield me. And I have been labelled a ‘delicate flower’ by certain commentators as a result. They thought I was just complaining that someone had sworn at me.

She goes on to reproduce a few of the messages she received, and at this point I am going to do the same (I’ve avoided it so far, but complete avoidance is starting to feel hypocritical):

FIRST WE WILL MUTILATE YOUR GENITALS WITH SCISSORS, THEN SET YOUR HOUSE ON FIRE WHILE YOU BEG TO DIE TONIGHT. 23.00

I have a sniper rifle aimed directly at your head currently. Any last words you fugly piece of shit? Watch out bitch.

SHUT YOUR WHORE MOUTH…OR ILL SHUT IT FOR YOU AND CHOKE IT WITH MY DICK

How can you make people understand the effect of receiving thousands of messages like this in the space of one weekend if you cannot repeat them, or say any of the words they contain?

We are back to Constance Grady’s dilemma: repeated exposure to sexually graphic and violent language may cause readers and listeners distress, but shielding them from the reality of abuse by wrapping it up in linguistic cotton wool means that women’s experiences will be trivialised or denied.

It may also mean that perpetrators are given the benefit of the doubt. Vague language has been a gift to apologists like Matt Damon, who has talked about ‘a spectrum of behaviour’ (meaning, OK, there are extreme cases like Harvey Weinstein, but most men who’ve been accused of ‘misconduct’ have done nothing really wrong). By contrast, it would hardly be convincing to talk about ‘a spectrum of threatening to shoot a woman in the head’, or ‘a spectrum of whipping your penis out and forcing a woman to watch you masturbate’.

Violent men throughout history have not only relied on women’s fear to keep them compliant, they have also relied on women’s shame to keep them silent. In the last few weeks many women have broken their silence (in some cases a silence that had lasted years); but when their accounts are presented in a veiled, inexplicit language, that subtly reinforces the idea that their experiences are somehow shameful. We cannot put that shame where it belongs–with the perpetrators, not their victims–if we cannot describe the details of what was done and what was said. So, while I don’t dismiss the problems Constance Grady discusses, I am ultimately of the same opinion as Emma Jane, Lin Farley and Caroline Criado-Perez: it’s important to be explicit.

Are women over-emojinal?

To mark World Emoji Day on July 17, the Empire State Building was lit up in emoji yellow, and Leah Fessler wrote a piece for Quartz* about why feminists should stop using emoji. She’d realised, she explained, that larding your messages with smiley faces, love hearts and thumbs-up signs (as well as exclamation points, which she treats as honorary emoji) is yet another form of emotional labour the world demands from women.

From childhood, women are conditioned to smile and nod to ensure that others feel comfortable and confident. This dynamic translates in digital communication through emoji and exclamation points.

Fessler was determined to break her emoji habit, but initially she found it hard; she felt ‘rude and awkward’ replying to messages with a simple ‘OK’ or ‘sure’ rather than a cheery thumbs-up or an enthusiastic ‘absolutely!’ But a few days in, she found she was starting to reap the benefits. Ditching the emoji, she reports, ‘wasn’t just a relief—it was empowering’.

Ah, the E-word–so often the canary in the coalmine of language-policing bullshit. Despite its use of sociological terms like ‘conditioning’ and ’emotional labour’, this is still basically an example of the formula I’ve analysed in previous posts about ‘just’, ‘sorry’, uptalk et al. First, you identify something women do, or are believed to do, more than men; next, you explain why that’s a problem for women; finally, you exhort women to empower themselves by changing their behaviour. Embellish the basic argument with some personal anecdotes, finish with an Uplifting Thought (‘even the smallest changes can alter the way others view you, and more importantly, the way you view yourself’)–et voilà, job done.

But enough of the snark: let’s try to unpick the argument.

Since emoji came into widespread use, their merits or otherwise have been extensively discussed, and opinion has been divided. On one hand they’ve been lauded in pieces (misguidedly) proclaiming them a new universal tongue, ‘the world’s fastest growing language’; on the other they’ve been belittled in comments like this one from the New York Times:

Given their resemblance to the stickers that adorn the notebooks of schoolgirls, not to mention their widespread adoption as the lingua franca of tweens and teens everywhere, some people wonder whether grown men should be using [emoji] at all.

This has something in common with the popular response to other linguistic innovations, like uptalk and vocal fry, which are associated with young women. When a linguistic form is stereotyped as a ‘girl thing’, you can bet that people will disparage it–and also that they will project a meaning onto it which reflects their ideas, or prejudices, about girls. Uptalk, for instance, has persistently been interpreted as a sign that the speaker doesn’t know what she really thinks, and/or is desperate for others’ approval–a story we’d find less intuitively plausible if it were told about something middle-aged men did. (If you want to know why linguists who’ve studied uptalk don’t buy this interpretation, see here).

By likening them to the decorative stickers young girls put on their school notebooks, the writer quoted above implies that women’s enthusiasm for emoji is of a piece with their more general fondness for frivolous embellishments. In scholarly discussions you’re more likely to encounter a different stereotype: women use emoji more than men because they’re more ’emotionally expressive’. Apart from being suspiciously circular (is there any evidence that women are more emotionally expressive apart from the kinds of emotional expression which their emotional expressiveness is meant to explain?), this argument presupposes that expressing emotion must be the function of emoji. Which might seem to be self-evident (isn’t the clue in the name?), but is actually an oversimplification.

That emoji are neither purely decorative nor all about the expression of emotion becomes clearer if you know something about their history. The precursors of emoji, emoticons (the earliest of these were smiley and winky faces made by combining ASCII characters, and they were invented by a grown man, or so he claims) were taken up by participants in early online forums to address a problem in what was then a new communication medium. Written language offers fewer resources than speech for signalling how you intend your message to be taken. In speech you’ve got the pitch, tone, loudness and quality of your voice (plus in face-to-face contexts facial expressions and body language), but in text-based interaction you’ve got none of these. This was leading to conflict when messages that were meant to be ironic or humorous prompted angry reactions from others who read them ‘straight’. Emoticons were ‘meta’ devices which enabled writers to signal their intentions more explicitly.

Today’s emoji are more diverse than emoticons in both their forms and their functions. They do provide resources for emotional expression, but that isn’t the only thing they’re used for. They’ve retained their usefulness as indicators of ironic or humorous intent, and they can also serve as tools for managing the mechanics of text-based interaction. On Twitter, for instance, the heart is often used simply as an acknowledgment token, to let someone know you’ve seen a tweet rather than to express your feelings about its content. In that case its affective meaning (‘heart = love’) is irrelevant, and competent users understand that. The same applies to the office conventions Leah Fessler complains about, like acknowledging meeting reminders with a thumbs-up emoji. Though the thumbs-up gesture conventionally symbolises enthusiasm, in this context it’s no more likely to mean ‘I’m really excited about this meeting’, than the Twitter heart is to mean ‘I’m in love with this tweet’. Using it is less a form of emotional labour than a labour-saving device (it’s quicker and easier than composing a verbal acknowledgment).

But if emoji aren’t just about emoting, or decoration, how do we explain women using them more than men? Actually, let’s go back a step: do women use them more than men?

Some of the evidence presented to support this claim should be approached with caution, because it comes from studies which asked people to report on their emoji use rather than analysing their actual output. Self-reports vary in accuracy, and they’re liable to be influenced by the subjects’ beliefs about who uses what kind of language (if a form is associated with women, that in itself may lead men to under-report their use of it). However, the generalisation that women use emoji more than men does have some credible research evidence to back it up. One frequently-cited study was done in Texas in 2012: researchers analysed 124,000 real text messages produced by subjects who granted access to their phones without knowing what the research was about. Women in this sample were twice as likely as men to include emoticons in a text–though their overall frequency was low for both sexes (only 4% of the texts sampled contained any).

But research also shows that gender is not the only variable affecting how, and how much, emoji are used. You can’t easily generalise across genres, platforms and devices–text messages aren’t the same as emails, and what goes for Tinder doesn’t apply to Twitter. Nor can you sensibly talk about ‘women’ or ‘men’ as homogeneous categories, without reference to intersecting variables like age and social background. It can also make a difference who the messages are addressed to. One early study of emoticon use in newsgroups (which were important online forums before Web 2.0 and smartphones) found that women in all-female or female dominated groups used more emoticons than men in all-male or male dominated ones. In mixed newsgroups, however, there was no significant difference: women used emoticons at much the same rates as they did in all-female groups, but men used them much more frequently when they were participating in a mixed group. This is not an unusual observation. Masculinity and femininity are often performed differently in single-sex and mixed sex interaction, where people are responding to different kinds of peer pressure, and where they may also be having different kinds of conversations.

In sum: ‘women use emoji more than men’ is probably true as far as it goes; there’s also plenty of evidence showing that certain emoji are used more by women and others are used more by men. However, these descriptive generalisations (like all statements about ‘men’ and ‘women’) mask significant variation within each group. And no descriptive generalisation (if you’ll forgive me for stating the obvious) constitutes an explanation of the facts it describes.

As I’ve already said, there’s a tendency for researchers to go straight for the ‘women are more emotionally expressive’ explanation. But as I’ve also already said, this is not entirely satisfactory. And there are other possibilities which fit at least as well with things we know about gender and communication in other contexts. For instance:

The difference might be a matter of style (a word which shouldn’t, in the context of language, be taken to denote something trivial). Online and offline, language is one of the symbolic resources which people draw on to create a distinctive persona, and to mark themselves as members or non-members of particular social groups or subcultures. Many small differences in pronunciation do this job: different variants of the same sound may be used to differing degrees or in different ways by speakers of different ages, classes, ethnicities and genders. The same principle might explain why men and women use an overlapping but non-identical range of emoji, and why women use them more frequently overall. You could compare it to the way we style our clothes or our hair: with emoji as with clothes, it seems that mainstream masculinity is less flamboyant than the feminine equivalent. (Interestingly, a study that investigated the perceived gender of the most popular emoji found that the ones ranked most ‘feminine’ included ‘face throwing a kiss’ and ‘face with tears of joy’, while the most ‘masculine’ were the more prosaic ‘thumbs-up’ and ‘OK’ symbols).

The difference might also be a by-product of the fact that men and women, if we insist on considering them as aggregates, tend to be members of different social networks in which they have conversations about different things. This is a point demonstrated in many tedious Big Data studies of vocabulary: I wish I had £10 for every article I’ve read which announces that ‘men and women use different words’, when all the research really shows is that people generally use words which pertain to the subject under discussion. Amazingly enough, women posting family news on Facebook tend to use more words relating to kinship and family occasions (like ‘grandma’ and ‘wedding’) than men debating current affairs or the performance of Arsenal Football Club. Most emoji are less subject-specific in their application than most words, but it’s still reasonable to think that which and how many of them you use might have something to do with who you’re talking to and what about.

It’s also possible that women’s more frequent use of emoji exemplifies the common pattern where innovations which are destined to spread through the whole population become visible first among young women. Because of that, we start out assuming they’re a ‘girl thing’ and looking for explanations which connect them to femininity; but as they spread it becomes apparent that they weren’t so much a girl thing as a new thing that girls got onto first. This is what has happened with uptalk: once a Valley Girl signature, it’s now heard among young and even middle-aged speakers, of both sexes, in many parts of the English-speaking world. Maybe the emoji gap between men and women will also narrow over time, making the question of why women use more emoji redundant.

This is not a multiple choice test where there’s only one right answer: the social life of language is too complex for one-size-fits-all explanations. And that’s an especially important point to bear in mind when you’re trying to explain gendered behaviour, because the categories ‘men’ and ‘women’ are so internally diverse–different subsets of men and women may be using or not using emoji for different reasons. I know I keep on saying this, but it really can’t be over-emphasised: no single thing can explain the behaviour of every member of a group which comprises half of humankind.

Do I think Leah Fessler is describing a real phenomenon? Yes. I think the pressure to be relentlessly upbeat and positive is a feature of many workplace cultures, and I also think there are good reasons to be critical of it (if you’re interested, this is the subject of Barbara Ehrenreich’s book Smile or Die). But do I think it follows that women should stop using emoji? No. Emoji are not the problem here. Like other attempts to ’empower’ women by changing their behaviour rather than the conditions it’s a response to, the ‘feminist case against emoji’ is fundamentally a pile of poo.

*Thanks to Mercedes Durham for drawing my attention to the Quartz piece.

2016: the bad, the bad and the ugly

Once again tis the season to look back on the last twelve months, and since we’re talking about 2016, that may not make for uplifting reading (unless your heroes are the Four Horsemen of the Apocalypse and President-Elect Donald Trump). If the Words of the Year chosen by dictionaries are any guide, the mood among English-speakers is darker than it was a year ago. Whereas Oxford’s choice in 2015 was the ‘face with tears of joy’ emoji, in 2016 it has gone for ‘post-truth’; other dictionaries’ selections have included ‘paranoid’, ‘surreal’ and ‘xenophobia’.

The reasons why this year sucked were not primarily to do with language, but language played a part—in some cases quite a prominent part. So, this review will be more about the lowlights than the highlights. Here are six of the worst:

Bantering bigots. In my 2015 annual round-up I named ‘banter’ as the word I’d most like to ban (if banning words were either feasible or desirable, which IMHO it isn’t). But banter continued to be exchanged in 2016, and the word ‘banter’, and variations thereon, continued to be used to wave away accusations of misogyny and bigotry. Both these tendencies peaked in October with the release of a 2005 tape in which Republican presidential candidate Donald Trump engaged in what he and his defenders called ‘locker room talk’. He was elected just a few weeks later.

Relentlessly sexist commentary on female politicians, often focusing (most notably in the case of Hillary Clinton) on their voices and style of speaking. All the familiar word-weapons—‘shrill’, ‘harsh’, ‘grating’, ‘aggressive’—were deployed by all the usual suspects.

If you’re thinking, ‘but surely there was plenty of critical commentary on Donald Trump’s language too’, you’re not wrong, but the comparison is instructive. When negative judgments are made on the speech of a female politician, her alleged failings are typically presented as the failings of her sex in general. Trump’s failings, on the other hand, were presented as his alone. They were ‘Trumpisms’, not ‘man-isms’ (it was even argued that Trump talks like a woman). The one exception was the ‘locker room talk’, where the idea that this was typical male behaviour got wheeled out not to condemn Trump but to excuse him.

If a female politician is widely acknowledged as an excellent public speaker, you can always accuse her of talking too much. In April, Owen Smith MP (in case you’ve forgotten, he was the man who unsuccessfully challenged Jeremy Corbyn for the Labour Party leadership) tweeted about his visit to a café in Millport in Scotland. He included two photos, one showing him with his arms around two of the ‘ladies’ (his description) who worked there, and the other showing a jar of old-fashioned gobstoppers. The part of the tweet relating to this second image said: ‘they’ve got the perfect present for @NicolaSturgeon, too’. A gobstopper, geddit? Because Sturgeon, the First Minister of Scotland (and at the time—before Theresa May became PM—the most powerful female politician in the UK), talks entirely too much and needs a good shutting up.

The continuing war on the word ‘women’. Two of the most popular posts I published this year touched on the question of why ‘women’ now seems to be the hardest word. In April the women’s section of the UK Green Party set off a Twitterstorm with its use of the term ‘non-men’. Across the Atlantic in September we had Planned Parenthood talking about ‘people’ being ‘criminalised for their pregnancy outcomes’. And throughout the autumn there were regular sightings of a new addition to the lexicon of ‘women’-avoidance: ‘menstruators’.

Having rejected sex or gender-based labels as essentialist and exclusionary, promoters of this term apparently felt that bodily function-based labels were the way to go. I, by contrast, feel pretty sure they aren’t. If you don’t want to say ‘women’, OK, I get it, but why not try using your linguistic judgment to find a contextually appropriate alternative? In this case, where the news story was about the removal of sales tax on pads and tampons, ‘sanitary product buyers’ would have worked—or where the report had already made clear what products were being discussed, just ‘customers’. If you’d find it offensive, or just plain weird, to read statements like ‘the recent fall in the price of toilet paper has been welcomed by defecators across the country’, or ‘perspirers have questioned the classification of deodorant as a luxury’, then you shouldn’t be giving house-room to ‘menstruators’ either.

More terrible advice and stupid opinions about women’s speech. This year hasn’t (yet) brought us anything quite as ludicrous as the ‘Just Not Sorry’ app that appeared at the very end of 2015, but bullshit continued to be churned out by the bucketload. It remained a truth universally acknowledged that women apologise too much, and constant criticism of female ‘verbal tics’ was once again presented as empowering rather than underminingAn op-ed piece in the New York Times added ‘I feel like’ to the list of words and phrases women should avoid if they want anyone to take them seriously—while also managing to relate the rise of ‘feeling like’ to Everything That’s Wrong With Our Society Today. (If anyone from the Times is reading this, I’d be happy to advise on what linguistic opinions editors should avoid giving space to if they want anyone to take them seriously.)

Not all bad advice is addressed to women: some of it is advice for men on how to make women’s lives a misery. The example that got most attention this year advised on how to make a woman take off her headphones and PAY ATTENTION. Because it’s part of a woman’s job description to be available to random men who want to converse with her AT ALL TIMES.

Death. It’s become a truism (though maybe not an actual truth) that 2016 brought a bumper harvest for the Grim Reaper. Two posts on this blog reflected that: one was a response to the death of the architect Zaha Hadid and the other was prompted by the murder of Jo Cox MP.

Online misogyny. In 2016 the abuse directed at women online was widely acknowledged as a significant problem, and in Britain it was the subject of a high-profile cross-party campaign—which was launched with a report that managed to blame half of the problem on women. (If you want to read something more sensible on this subject, I can recommend Emma Jane’s new book Misogyny Online: A Short (and Brutish) History.)

There were a few small consolations:

Resolution 109. The American Bar Association made the use of patronising endearment terms to women lawyers a breach of professional standards. (Meanwhile in the UK, a female judge responded to a male defendant who called her a cunt by saying ‘you’re a bit of a cunt yourself’.)

Women political speakers kicking ass. In the wake of the referendum that brought us Brexit, Nicola Sturgeon showed once again that few politicians can touch her when it comes to rhetorical skill. The US presidential campaign brought another outstanding female political speaker to the world’s attention: Michelle Obama.

Arrival. Not the best thing I’ve ever seen, but hey, Hollywood made a film about a woman linguist who saves the world!

In real life, of course, linguists don’t save the world: the best someone like me can do is try to make a bit more sense of some of the things that are happening in the world. As ever, my efforts to do that this year have been indebted to the work of many other researchers and/or bloggers, and I’m grateful to everyone I’ve cited/linked to in my posts.

I’ll be back with more feminist guiding in 2017, but in the meantime I thank everyone who reads the stuff I put here (there are a lot more of you than I ever thought there would be when I started this blog in 2015), and I wish you as much peace, love and joy as you can find in these unsettled and discouraging times.

Misogyny by numbers

Last week saw the launch of Reclaim the Internet, a campaign against online misogyny. Both the campaign and the (copious) media reports of it leaned heavily on research conducted by the think-tank Demos,  which investigated the use of the words ‘slut’ and ‘whore’ in tweets sent from UK-based accounts over a period of about three weeks earlier this year. The study identified 10,000 ‘explicitly aggressive and misogynistic tweets’ containing the target words, sent to 6500 different Twitter-users. It also found that only half these tweets were sent by men—or, as numerous media sources put it, that ‘half of all online abusers were women’.

So frequently and insistently was this statistic repeated, the message of the day almost became, ‘look, women are just as bad as men!’ Women like the journalist and feminist campaigner Caroline Criado-Perez, who were sought out for comment because of their experience of online abuse, got drawn into lengthy discussions about the misogyny of other women.

Of course, it isn’t news that some women call other women ‘sluts’ and ‘whores’ (or that women may be involved in the most serious forms of online abuse: one of the people prosecuted for sending death-threats to Criado-Perez was a woman). But ‘who sends abusive messages?’ is only one of the questions that need to be addressed in a discussion of online abuse. It’s also important to ask who the messages are typically addressed to and what effect they have, not just on their immediate recipient but on other members of the group that’s being targeted. But those questions weren’t addressed in this particular piece of research, and it was difficult to raise them when all the interviewers wanted to talk about was that ‘half of all abusers are women’ statistic.

These discussions reminded me of the way anti-feminists derail discussions of domestic violence with statistics supposedly showing that women are as likely to assault men as vice-versa. Feminists have challenged this claim by looking at the finer details of the data the figures are based on. They’ve pointed out, for instance, that female perpetrators are most commonly implicated in single incidents, whereas men are more likely to commit repeated assaults, and to do so as part of a larger pattern of coercive control. It’s also men who are overwhelmingly responsible for the most serious physical assaults, and for the great majority of so-called ‘intimate partner killings’.

Once you focus on the detail, it’s clear domestic violence isn’t an equal opportunity activity. Online misogyny probably isn’t either (especially if you focus on the kind that really does deserve to be called ‘abuse’—stalking, repeated threats to rape and kill, etc). But the Demos study didn’t capture any of the detail that would allow us to see what’s behind the numbers.

In this it is fairly typical of the kind of research which funders, policymakers and the media increasingly treat as the ‘gold standard’, involving hi-tech statistical analysis of very large amounts of information—what is often referred to as ‘big data’, though that term has come to be used rather loosely. Strictly speaking, the ‘big data’ label wouldn’t apply to the Demos study, whose sample of 1.5 million tweets is very small beer by big data standards. At the same time, it’s too much data to be analysed in detail by humans: the researchers employed NLP (natural language processing), using algorithms to make sense of text, and their findings are essentially statistical—figures for the frequency of certain kinds of messages, along with the gender distribution of their senders.

You may be thinking: but doesn’t it make sense to assume that ‘bigger is better’—that the more data you crunch through, the more reliable and useful your results will be? I would say, it depends. I’m certainly not against quantitative analysis or large samples: if the aim of a study is to provide information about the overall prevalence of something (e.g., online misogyny on Twitter), then I agree it makes sense to go large. Actually, you could argue that Demos didn’t go large enough: not only was their sample restricted to tweets which contained the words ‘slut’ and ‘whore’, the time-period sampled was short enough to raise suspicions that the findings were disproportionately affected by a single event (the surprisingly high number of woman-on-woman ‘slut/whore’ tweets may reflect the massive volume of abuse directed at Azealia Banks by fans of Zayn Malik after she attacked him publicly).

What I am against, though, is the idea that the combination of huge samples and quantitative methods must always produce better (more objective, more reliable, more revealing) results than any other kind of analysis. Different methods are good for different things, and all of them have limitations.

The forensic corpus linguist Claire Hardaker knows a lot about what can and can’t be done with the tools currently available to researchers, and she has explained on her blog why she’s sceptical about the Demos study. Her very detailed comments confirm something a lot of people immediately suspected when they first encountered the claim about men and women producing equal numbers of abusive tweets. That claim presupposes a degree of certainty about the offline gender of Twitter-users which is not, in reality, achievable. (This isn’t just because people disguise their identities online, though obviously a proportion of them do; Hardaker explains why it’s a problem even when they don’t.)

Another thing Hardaker is sceptical about is the researchers’ claim to have trained a classifier (a machine learning tool that sorts things into categories) to distinguish between different uses of ‘slut’ and ‘whore’, so that genuine expressions of misogyny wouldn’t get mixed up with ironic self-descriptions or mock insults directed at friends. Her observations on that point deserve to be quoted at some length:

We can guarantee that the classifier will be unable to take into account important factors like the relationships between the people using [the] words, their intentions, sarcasm, mock rudeness, in-jokes, and so on. A computer doesn’t know that being tweeted with “I’m going to kill you!” is one thing when it comes from an anonymous stranger, and quite another when it comes from the sibling who has just realised that you ate their last Rolo. Grasping these distinctions requires humans and their clever, fickle, complicated brains.

When you depend on machines to make sense of linguistic data, you have to focus on things a machine can detect without the assistance of a complicated human brain. A computer can’t intuit whether the sender of a message harbours particular attitudes or feelings or intentions; what it can do, though, is identify (faster and often more accurately than a human) every instance of a specific word. So, what happens in quite a lot of studies is that the researchers designate selected words as proxies for the attitudes, feelings or intentions they’re interested in. In the Demos study, these proxy words were ‘slut’ and ‘whore’, and the presence of either in a tweet was treated as a potential indicator of misogyny.

One obvious problem with this is that it excludes any expression of misogyny that doesn’t happen to contain those particular words. The researchers themselves were well aware that tweets containing ‘slut’ and ‘whore’ would only make up a fraction of all misogynist tweets (one of them told the New Statesman they were ‘only scratching the surface’).  But that point got completely lost once their research became a media story. The media need short-cuts: they’ve got no time for the endless qualifications that litter academics’ prose. Consequently, the figures given in the report for the frequency of ‘slut’ and ‘whore’ soon began to be presented as if they were a definitive measure of the prevalence of online misogyny in general.

The researchers were also aware that in context, ‘slut’ and ‘whore’ aren’t always expressions of misogyny. They may be being used in an ironic or humorous way; they may turn up in feminist complaints about ‘slut-shaming’ or ‘whorephobia’. So after searching for every instance of each word, the researchers used a classifier to filter out irrelevant examples and sort the rest into various categories.

Since the full write-up of the 2016 Demos study doesn’t seem to be available yet, I’ll illustrate how this works in practice using the report of a study which the same research group carried out in 2014, apparently using much the same methodology. In this earlier research they investigated three words, ‘slut’, ‘whore’ and ‘rape’. When they analysed the ‘rape’ tweets, they started by getting rid of irrelevant references to, for instance, ‘rapeseed oil’. Then they used a classifier to distinguish among tweets which were discussing an actual rape case or a media report about rape (these made up 40% of the total), tweets which were jokes or casual references to rape (29%), tweets which were abusive and/or threats (12%), and tweets which didn’t fit any of those categories and so were classified as ‘other’ (it’s possibly not a great sign that 27% of the sample ended up in the ‘other’ category).

This looks like the kind of classification task that computers aren’t very good at for the reasons explained by Claire Hardaker (distinguishing abuse from humour, for instance, calls for human-like judgments of tone). But the limitations of current technology may not be the only problem. As a check on the classifier’s reliability, a few hundred tweets from the sample were classified by human analysts. Some of these manually-categorized examples are reproduced in the report to illustrate the different categories. To me, what these examples show is that once messages have been extracted from their context, there’s often enough ambiguity about their meaning to cause problems for a human, let alone a machine.

Here’s a straightforward case—a tweet the humans categorised as a joke.

@^^^^ that was my famous rape face 😉 LOL Joke.

This is unproblematic because the tweeter has taken a lot of trouble to make its status as a joke explicit (adding a winky face, LOL, and then the actual word ‘joke’). But how about this tweet, which comes from the 12% of rape references which the humans categorised as ‘abusive/ threats’?

@^^^^ can I rape you please, you’ll like it

If you think that’s a repugnant thing to tweet at someone, you’ll get no argument from me. But I don’t think it’s self-evident that this strangely polite request is intended as a serious threat rather than a ‘mock’ or ‘joke’ threat. The original recipient will have decided how to take it by using contextual information (e.g., whether the tweeter was a friend or a random stranger, what if anything the tweet was responding to, whether there was any history of similar messages, etc.) Without any of that context, the significance of a message like this one for its original sender and recipient is something an analyst can only guess at.

The example I’ve used here is a ‘threat’ that might conceivably have been intended and taken as a ‘joke’, but it’s likely there are also cases of the opposite, tweets the researchers classified as ‘jokes’ which were intended or taken as serious threats. So I’m not suggesting that the proportion of actual rape threats was lower than the reported 12%; I’m suggesting that the classification—even when done by humans—is not sufficiently reliable to base that kind of claim on. And that the main reason for this unreliability is the way large-scale quantitative studies of human communication detach individual communicative acts from the context which is needed to interpret their meaning fully.

Whether we’re academics, journalists or campaigners, we all like to fling numbers around. There’s nothing like a good statistic to draw attention to the scale of a problem, and so bolster the argument that something needs to be done about it. And I’m not denying that we need the kind of (large-scale and quantitative) research which gives us that statistical ammunition. But two caveats are in order.

First, large scale quantitative research is not the only kind we need. We also need research that illuminates the finer details of something like online misogyny by examining it on a smaller scale, but holistically, with full attention to the contextual details. There’s a lot we could learn about how online abuse works—and what strategies of resistance to it work—by using a microscope rather than a telescope.

Second, if we’re going to rely on numbers, those numbers need to be credible. In that connection, the Demos study hasn’t done us any favours: I’ve yet to come across any informed commentator who isn’t at least somewhat sceptical about its findings. While some of the problems people have commented on reflect the way the media reported the research—pouncing on the ‘women send half the tweets containing “slut” and “whore”’ claim and then reformulating that as ‘women are half of all online abusers’ (an assertion whose implications go well beyond what the evidence actually shows)—there are also problems with the researchers’ own claims.

‘My issue’, says Claire Hardaker, ‘is that serious research requires serious rigour’.  When research is done on something that’s a matter of concern to feminists, its quality and credibility should be an issue for us too.