Banal sexism

Last month I wrote about David Bonderman, the billionaire businessman who resigned as a director of Uber after suggesting that appointing more women to the board would mean ‘more talking’. Allegedly he meant this comment as a joke; but even if no one present had been offended, you have to wonder who would have found such a hoary old cliché amusing. An enormous amount of sexism is like this: thoughtless, repetitive, trite and formulaic. What—as bad stand-up comedians say—is that about?

Back in 1995, Michael Billig wrote a book about a phenomenon he called ‘banal nationalism’. The term ‘nationalism’ is most commonly used to denote what Billig refers to as ‘hot’ nationalism—a political ideology driven by strong emotions, which is often associated with conflict and violence. But his point was that there’s a less overt, lower-level form of nationalism which we don’t generally call by that name. Unlike the ‘hot’ variety, its main function is not to foment conflict or hatred of the Other. It’s to maintain our awareness of ourselves as national subjects—keep ‘the nation’ as a concept ticking over at the back of our minds. In Billig’s words:

National identity is remembered in established nations because it is embedded in routines of life that constantly remind, or ‘flag’ nationhood. However, these reminders or ‘flaggings’ are so numerous, and they are so much a part of the social environment, that they operate mindlessly, rather than mindfully.

The word ‘flag’ in this quote is a pun: one obvious daily reminder of nationhood is the national flag, flying (or as Billig puts it, ‘hanging limply’) on hundreds of public buildings. But banal nationalism takes subtler forms too, and many of them have to do with language.  For instance, the use of first person ‘we/us’ to mean ‘the people of this nation’, whereas the people of other nations are referred to with the third person ‘they/them’. The presence on every high street of businesses with names like the ‘Nationwide Building Society’ and—until recently—‘British Home Stores’. TV programmes hailing viewers with ‘Good Morning Britain’. Formulaic phrases that reference people’s shared membership of a nation, whether explicitly (‘best of British luck’) or implicitly (‘it’s a free country’).

The same idea can be applied to sexism.  Sexism also has ‘hot’ forms, and those are the ones mainstream discourse finds it easiest to recognise and condemn. The western media have no difficulty in recognising the sexism of the Taliban and Boko Haram; the more liberal parts of the western media have no difficulty in recognising the sexism of Gamergaters and Donald Trump.  But what you might call ‘banal sexism’—ordinary, unremarkable, embedded in the routines and the language of everyday life—is a different story. It does often go unnoticed, and when feminists draw attention to it they’re accused of taking offence where none was intended or embracing ‘victim culture’. These knee-jerk defences are often delivered with an air of surprise—as if the people responsible hadn’t realised until that moment that anyone could possibly dissent.

The idea that women talk incessantly is a classic example of banal sexism—it’s something people trot out on autopilot, as if they were commenting on the weather.  Most remarks about the weather fall into the category of small talk, or what the anthropologist Bronisław Malinowski called ‘phatic communion’: their function is not to exchange information, but just to establish common ground and reassure others of our good intentions. That’s why statements like ‘lovely day today’ are almost invariably met with agreement: ‘Yes, beautiful!’ It would be odd to respond with something like ‘well actually it’s two degrees below the mean temperature for mid-July’. That might be an impressive demonstration of your meteorological knowledge, but it would also reveal your social incompetence, since you’d have missed the whole point of a phatic exchange. It’s the same with banal sexism: challenging the proposition (‘well, actually studies show that men talk more than women in most situations’) will be seen as a peculiar and hostile act. It’s especially hard to challenge a joke, because no one wants to be accused of lacking a sense of humour.

In my youth I didn’t understand this. I remember the first time I ever heard Chas & Dave’s pop classic ‘Rabbit’, a jolly cockney moan about women who give their husbands earache. It was 1980, and—at the age of 21—I had recently discovered my inner Radical Feminist. I thought, ‘you may sell that record today, but it won’t be long before you’re history’.  I was wrong: nearly 40 years later, the myth of the Woman Who Never Shuts Up remains ubiquitous in popular culture. Consider, for instance, this advertisementIMG_7139 for cruising holidays, which was recently photographed by a Swiss follower of this blog*:

Translated into English, this says: ‘Peace/quiet on holiday? Make your wife simply speechless’.  It’s a banal sexism double whammy, combining the old ‘rabbit, rabbit’ cliché with the idea that you can always shut a woman up by spending your hard-earned wages on something she wants. The ad’s presuppositions are both insulting and false (women don’t talk more than men, and according to one 2013 industry survey they make about 80% of household travel plans), but whoever came up with it seems not to have been concerned about offending potential customers.

Nor do I suspect its creator of deliberately courting controversy, though that’s certainly a strategy some advertisers have used. Banal sexism doesn’t provoke outrage. It occupies the part of the spectrum that runs from ‘seen but unnoticed’ (like the ‘default male’ convention which I discussed in an earlier post) through to ‘annoying but not worth getting all fired up about’. You might shake your head, roll your eyes, post a photo with a scathing comment on Facebook, but most people wouldn’t bother to make a formal complaint.

But sometimes the zeitgeist changes, and a form of sexism which has previously been tolerated gets moved from the ‘banal’ into the ‘hot’ category. Last year, for instance, a friend of mine spotted this greeting card, womenpart of a range addressed to men, in a university bookshop. Greeting cards in general are like a bottomless well of banal sexism, and ‘humorous’ cards like this have been around forever: though feminists have long found their message objectionable, most people have treated it in the same way as the ‘make your wife simply speechless’ ad, as an essentially harmless (if perhaps tasteless) joke based on the banal trope of ‘the eternal battle of the sexes’.

But recently more people have become aware (thanks in part to the work of feminists like Karen Ingala Smith and her Counting Dead Women project) that in the UK a man actually does kill a woman, most commonly a current or former partner, about every 2-3 days. If you’ve thought about that statistic, you’re less likely to let a joke about ‘shooting women and burying them in the garden’ pass without protest. I wasn’t surprised to hear that my (feminist) friend had complained, but I was pleasantly surprised to learn that the bookshop manager had agreed with her–and had promptly withdrawn the card from sale.

But the issue here is not just about the (un)acceptability of joking about male violence. Banal sexism is also exemplified by the formulas used in serious news stories about the killing of women by men. In France (where the statistics are similar to the UK’s), the journalist Sophie Gourion has set up a tumblr called Les Mots Tuent (‘words kill’) to document and criticise the linguistic ‘banalisation’ (‘normalisation/trivialising’) of violence against women and girls. She is exasperated by the constant repetition of phrases like crime passionel (‘crime of passion’, a category that does not exist in current French law), drame familial (‘family drama’, typically referring to ‘family annihilation’ cases where a man murders his partner and their children before killing himself) and pétage de plomb (‘blowing a fuse’, ‘flipping/freaking out’, ‘having a meltdown’). As she notes, these terms imply that the perpetrator was overcome by a sudden, uncontrollable impulse—whereas in fact many of these killings turn out to have been premeditated, not uncommonly by men who have long histories of domestic violence.

Similar formulas are well-established in the English-speaking media. In 1992, Kate Clark published an analysis of the Sun’s reporting of violence against women and girls, and found a pattern in the language used to label perpetrators and victims. In cases where ‘innocent’ women (in the Sun’s worldview that meant young girls or dutiful wives and mothers) were killed or assaulted by strangers, the perpetrators were given dehumanising labels like ‘beast’, ‘fiend’, ‘maniac’ or ‘monster’.  By contrast, reports of domestic violence, including homicide, tended to label men in ways that both humanised them and emphasised their own status as victims. One man who killed his wife and then himself was referred to as a ‘tormented’, ‘debt-ridden Dad’ (the word ‘tormented’ recurred in the reporting of so-called ‘family tragedies’); another who shot his wife and her mother dead was described as a ‘spurned husband’. Even the affectionate diminutive ‘hubby’ appeared in one report about a man whose 12-year history of domestic violence was revealed in court after he almost killed his wife.

Kate Clark’s data were taken from reports that had appeared in the late 1980s, but much of her analysis remains pertinent today. In Ireland last year, for instance, when a man named Alan Hawe stabbed his wife Clodagh to death, strangled their three sons and then hanged himself, the case was reported in both the Irish and British media as a ‘family tragedy’. The Mirror printed a photo which showed the family (in the words of the caption) ‘smiling together before all five lost their lives’.  ‘Lost their lives’ suggests an accident rather than the intentional killing which actually took place, but in the ‘family tragedy’ frame, as Clark’s earlier study found, the killer is usually portrayed as another victim, and often as the primary victim. In the Hawe case, again typically, much of the media’s attention focused on the mental ‘torment’ that must have driven Alan Hawe (described in numerous sources as a ‘real gentleman’ and a pillar of the community) to such extremes. Some commentators even portrayed him as a victim of sexism—the sexism of a culture which does not permit men to show weakness or express emotion.

This representation only began to be questioned after a blog post entitled ‘Rest in peace, invisible woman’, by the Dublin-based feminist writer Linnea Dunne, was picked up by the mainstream media. Dunne remarked on the way media reporting centred on the killer and his imagined state of mind (there was no actual evidence that Alan Hawe had any history of mental illness), while those he killed were treated as minor characters, or erased from the story entirely. Even the discovery of the family’s dead bodies was couched in terms that adopted the killer’s perspective: they were said to have been discovered by ‘his mother-in-law’ (aka Clodagh Hawe’s mother and the children’s grandmother).

By contrast with the keen interest they took in his mental state, reporters did not ask if Alan Hawe had a history of domestic violence. It would later turn out that he did: in the words of one family friend, ‘he controlled everything around him, he controlled how his family lived, he controlled how they died’. It would also emerge that Clodagh Hawe’s family, initially portrayed as grief-stricken but forgiving, had fought an eight-month battle to have the killer’s body removed from the grave in which he had originally been buried alongside his victims.

As time went on it became clearer and clearer that the framing of this story by most of the press had persistently obscured the material facts. And this is far from being an isolated example. This month, the UK press has been reporting on the case of Francis Matthew, a Briton living in Dubai, who killed his wife Jane with what the Emirati authorities described as ‘a strong blow on the head with a solid object’. Initially Matthew claimed that the attack had been perpetrated by burglars who broke into their home. Later, when it was clear this story would not stand up, he admitted that he had thrown a hammer at his wife during ‘a row’, but he continued to insist that her death was an accident. This example differs from the Hawe case in that there was only one victim: no children were involved and the perpetrator is still alive. But reports on it (like this one in the Telegraph) have used many of the same generic and linguistic conventions. For instance:

  1. The repetition of the words ‘tragedy’ and ‘tragic’. If the crime really had been committed by intruders, the reports would have used words suggesting anger and condemnation, but when murder is ‘all in the family’, the emotions we are directed to feel are sadness and pity for both/all parties.
  2. The centring of the (male) killer and the near-total erasure of his victims. Dead or alive, he is the main protagonist of the ‘tragedy’, while the victims exist only in relation to him. In the Telegraph’s report, for instance, we are told a fair amount about Francis Matthew’s life history, and we also learn that ‘the couple…were a fixture of Dubai’s social scene’, but nothing is said about Jane Matthew’s history, activities, interests or personality. Like Clodagh Hawe, she is rendered invisible.
  3. The presentation of the killing as a sudden, inexplicable eruption of violence into a previously happy relationship. In this case (as in the Hawe case before it, at least immediately after the murder), the message that Matthew’s act was ‘out of character’ is conveyed by reporting the reactions of others: ‘Friends and associates of Mr Matthew said they were astounded to hear that the genteel editor was under arrest. “He is the biggest teddy bear I know,” said one family friend’. Another acquaintance is quoted describing him as ‘relaxed, calm and laid back’. Though the Telegraph does mention that he has been charged with ‘premeditated murder’, it does not probe the apparent contradiction between this charge and Matthew’s own  claim to have killed his wife accidentally in the heat of the moment.
  4. The inclusion of multiple details which portray the killer as a man of good character and reputation. The Telegraph‘s report is headed by a photo of Francis Matthew shaking hands with the Emir of the UAE; it goes on to extol his educational and professional achievements, and makes several references to his standing in the expatriate community. This, we infer, is what makes the case so ‘tragic’. Not that a woman died following a brutal assault (and who knows how much other abuse in the months and years preceding it), but that a successful man’s life has been ruined by a momentary loss of control.

If I’m putting this kind of reporting in the category of banal sexism, it’s not because I think it’s trivial, but because I think it operates, as Billig says about banal nationalism, more mindlessly than mindfully. I don’t think there’s some media conspiracy to defend homicidal men: it’s more a case of reaching for the familiar formulas (the ‘family tragedy’ frame and the associated clichés—‘out of character’, ‘pillar of the community’, ‘lost their lives’) without ever thinking to interrogate the assumptions that lie behind them. It’s the news-story equivalent of the political discourse which Orwell, in 1946, compared to a ‘prefabricated henhouse’—assembled rapidly and unreflectively from a pile of standard, mass-produced components.

Let me hasten to make clear, though, that this analysis is not meant as an excuse for the journalists who produce these stories. On the contrary, I think this mindless recycling of familiar banalities about domestic violence is an absolute dereliction of their professional duty. Professionals who like to think of themselves as fearless seekers after truth should not be taking the conventional ‘family tragedy’ story at face value, particularly when—thanks to several decades of feminist activism and research—the facts which contradict it are readily accessible. There is ample evidence, for instance, that intimate partner killings like the murder of Jane Matthew are rarely ‘isolated incidents’, and that many men who are violent in private appear ‘calm and laid back’ in public.

Journalists are also professional language-users, and as such should be expected to make considered linguistic choices. Would anyone in any other context talk about ‘spurned husbands’ and ‘tormented dads’? It’s 2017, FFS: why are news reports still full of these archaic, tone-deaf clichés? If you call yourself a writer, you should try engaging your brain and actually thinking about the words you use.

Words may not literally kill, but that doesn’t mean they don’t have consequences. The banal sexism we see in the reporting of domestic homicide cases echoes, and so contributes to perpetuating, some of the same attitudes which are held more actively by men like Alan Hawe—like the idea that women are appendages rather than people who matter in their own right, and the view that violence is an understandable response to the pressures society puts on men. (‘Women. Can’t live with them, can’t let them live if they don’t want to live with you’.) I’m glad that this traditional formula is now attracting more outspoken criticism, and not only from the usual feminist suspects. It’s lazy, it’s sexist and no self-respecting news outlet should give it house-room.

*thanks to Martina Zimmermann

Advertisements

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.

Passive aggressive

In 2014, someone set up a Twitter account called ‘Name the Agent’ as part of a feminist campaign challenging the way the media reported violence against women. Specifically, the campaign criticized the use of the passive voice in news headlines like ‘Woman raped while walking her dog’. This headline fails to mention that a man committed the crime. It presents rape either as something that ‘just happens’ to women, or as something for which women are indirectly responsible–as if the woman was raped because she was walking her dog, and not because a man decided to rape her. The campaign called on the media to abandon the passive in favour of active-voice headlines like ‘Man rapes woman dog-walker’.

Complaints about the passive have a long history. Advice to avoid it has been around for the best part of a century: I imagine many people reading this were taught at school that it was ‘bad style’. Originally the reasons for this judgment had nothing to do with politics: commentators in the 1930s said that active sentences were ‘strong’ while passive sentences were ‘weak’. The connection with politics was made by George Orwell, whose 1946 essay ‘Politics and the English Language‘ included ‘never use the passive where you can use the active’ on a list of rules for combatting the politically-motivated abuse of language. This helped to popularize the now-common idea that the passive isn’t just bad style, it’s a tool used by the powerful to conceal unpalatable truths and manipulate public opinion.

The feminist argument that passives are used to conceal men’s responsibility for violence against women belongs to this post-Orwellian tradition. But in this post I’m going to try to explain why I don’t think the argument is convincing–why it’s really not as simple as ‘active good, passive bad’.

Before I go on, let’s just run through some grammatical basics.

Below is a simple active sentence. It puts the agent—the doer of an action—in the grammatical subject position, which in English normally means before the verb.

A man attacked a woman

And here’s the passive voice equivalent:

A woman was attacked by a man

In the passive version the subject is ‘a woman’, the person affected by the action, while the agent, ‘a man’, has been relegated to a ‘by’ phrase after the verb. This ‘by’ phrase is optional. You can remove it and still end up with a grammatical sentence, like this:

A woman was attacked

This is a passive sentence with agent deletion: the attacker has disappeared, leaving the sentence to focus entirely on the woman and what happened to her. Agentless passives are common in news reports and headlines: ‘Woman raped while walking her dog’ is an example.

Agentless passives are also common in legal proceedings, and in that context the feminist argument has some force. Research has shown that men who are accused of sexual violence, and the lawyers who represent those men, very often make strategic use of what the linguist Susan Ehrlich calls ‘the grammar of non-agency’, including agentless passives. In her book Representing Rape,  Ehrlich analyses a sexual assault trial in which the defence lawyer asks his client questions like

‘I take it the sweater was removed?’

It’s not hard to see what the lawyer hopes to achieve by choosing an agentless construction that doesn’t specify who removed the sweater. If the court thinks the complainant took off her own clothes, that will support–or at least, not contradict–the defence’s argument that she consented to sex.

As Ehrlich says, it’s only to be expected that defendants and their lawyers will use this strategy. It’s more surprising, and perhaps more worrying, that the same tendency to downplay men’s agency has been observed in the language used by judges. When the researcher Linda Coates and her colleagues analysed the language used in judgments on sexual assault cases in Western Canada, they found many examples of judges using agentless passives like this:

There was advantage taken of a situation that presented itself.

This statement was made in the judgment on a case where a ten year-old girl had been sexually assaulted by a stranger in her home. The ‘situation’, in other words, was the presence of a child in her own bedroom, and it did not magically ‘present itself’, it was engineered by the defendant. A jury had found the defendant guilty, but the judge chose to minimize the seriousness of his offence by describing it in a way that implied he had no agency at all–as if he merely reacted, as anyone might, to the circumstances in which he (inexplicably) found himself.

The judge’s statement is an egregious example of ‘the grammar of non-agency’. But is the use of the passive the main problem here? I think we can see it isn’t if we recast the sentence in the active voice:

The defendant/Mr X took advantage of a situation that presented itself.

This reformulation names the agent, but it doesn’t solve the problem. The vague wording still glosses over what the defendant actually did, and the sentence still presents him as simply reacting to a situation that was not of his own making.

Naming the agent is not the same thing as holding him responsible for his actions. Conversely, not naming the agent doesn’t have to mean concealing or denying his actions.

We can see this if we go back to the newspaper headline ‘Woman raped while walking her dog’, which was criticized for failing to mention the key fact that the crime was committed by a man. It’s true that the headline doesn’t explicitly describe the perpetrator as a man. But it’s not true that the effect is to obscure his maleness from the reader. The word raped, which does appear in the headline, cues the reader to activate what psychologists call a ‘schema’—a sort of mental template for the kind of event the word is applied to. Part of that schema is the information that rapists are prototypically male. For many English-speakers rapists are male by definition, because the meaning of the word rape in their mental dictionary includes the idea of penetration with a penis. But even if they define the word more broadly, their schema will still incorporate the knowledge that rapists are almost always men. If the suspect in a rape case were female, you can be sure the report would say so, precisely because it would be so unusual.

In practice, therefore, the agent-naming headline ‘man rapes woman dog-walker’ communicates no more information than ‘woman raped while walking her dog’. The difference is only that the first version mentions the attacker’s sex explicitly while the second relies on the reader to infer it.

But if the two versions communicate the same information, why do headline writers so often favour the passive? If that’s not about excusing men and/or blaming women, what is it about?

The answer is, it’s about focus. When you choose between the active and the passive, you’re also choosing what to put in the grammatical subject position. In crude terms, you’re deciding what the sentence is about. And you don’t always want it to be about the agent. For instance, if a high-profile public figure is assassinated, the breaking news headline is more likely to be ‘President shot’ than ‘Gunman shoots president’. The story isn’t about the shooter: what makes it news is the identity of the victim.

In stories like ‘Woman raped while walking her dog’, the main news is simply that a rape has been committed. The report can’t say much about either the attacker or the victim: his identity is not yet known, while hers is legally protected. (That’s probably why the writer added the dog-walking detail—not to imply that the victim put herself at risk, but to enable readers to relate to her as an ordinary person engaged in an everyday activity.) In some circumstances the headline-writer might choose to focus on the attacker–for instance, if he’d been caught and arrested, or if the report concerned the latest attack by a serial offender. But if the attacker is just an unidentified, generic ‘man’, there’s no compelling reason to focus on him. It isn’t news to anyone that rape is committed by men.

So, I don’t think there’s a media conspiracy to deny men’s responsibility for violence by using passive-voice headlines. But as I’ve already pointed out, what actually gets communicated doesn’t depend exclusively on the intentions of the speaker or writer. It also depends on the inferences made by hearers or readers. In theory, a writer’s linguistic choices could affect readers’ interpretations even if that wasn’t the writer’s intention. Recognizing that possibility, a number of researchers have run experiments to investigate whether the grammatical framing of a report makes any difference to readers’ judgments of the case.

The basic procedure involves dividing a sample of research subjects into two groups, presenting one group with an account of sexual violence framed in the active and the other with a matched account in the passive, and then asking subjects to rate (a) the perpetrator’s degree of responsibility, (b) the victim’s degree of responsibility and (c) the degree of harm to the victim. Subjects may also be asked to complete a questionnaire about their attitudes to sexual violence, so researchers can see how their judgments relate to their pre-existing beliefs.

I’ll start with what you might call the good news. These studies suggest that we’re not dealing with a form of Orwellian thought control: readers who don’t already subscribe to rape myths are not susceptible to the influence of language. Their judgments are the same regardless of which report they’ve read. The grammar of a report only makes a difference to the judgments of people who have high RMA scores (RMA stands for ‘rape myth acceptance’. And before you ask, yes, gender does play a role here: men on average have higher RMA scores than women, so it’s mostly men who are susceptible.)

The next question is how grammar affects the perceptions of those subjects who are influenced by it. The answer isn’t clear cut: different studies have found different effects. The first group of researchers to do the experiment found what they’d predicted: subjects who read a passive-voice report judged perpetrators less responsible than those who read the active-voice version. But later studies found the opposite: subjects who were influenced by grammar judged perpetrators more responsible after reading a report in the passive.

This second pattern doesn’t fit with the theory that the passive downplays men’s agency and shields them from blame. To explain why it’s been found in some studies, we need to consider what else you can do with passive sentences.

One researcher who has thought about this is Tamar Holoshitz. She conducted one of the experiments which found that passive reports prompted higher ratings of perpetrator responsibility; she also analysed the language used by prosecutors in domestic violence cases, where she noticed that they often referred to the same act or event using both active and passive sentences. For instance:

The defendant gave her a single blow to the left eye

She was admitted [to hospital] after being hit in the eye, suffering from trauma and an orbital fracture

These two sentences are designed to do different things. The first directs attention to the perpetrator and describes what he did. The second directs attention to the victim and describes the consequences of the assault for her. The active sentence names the agent; the passive sentence names the harm.

Holoshitz argues that prosecutors use both these strategies to maximize their chances of getting a conviction. The first is necessary (to convict a defendant you have to show that he committed the crime he is on trial for), but prosecutors know that on its own it may not be sufficient. On any jury there are likely to be people who think violence against women is acceptable under some conditions (if it was ‘just a slap’, if she was ‘asking for it’, if he just lost control and lashed out without really meaning to, etc.). If you want jurors who think like this to return a guilty verdict, you need to address their belief that some degree of force is acceptable. Naming the agent doesn’t do that (they’re not disputing the claim that he punched her), but naming the harm–presenting an account that emphasizes the effect of his violence on the victim–gives you some chance of blocking the standard excuses (‘this wasn’t just a slap. He put her in hospital. You don’t break someone’s bones without meaning to hurt them’). Holoshitz thinks it’s this emphasis on harm that her experimental subjects were responding to when they attributed more responsibility to perpetrators after reading reports in the passive.

What all this boils down to is that passives can serve more than one purpose. The prosecutors in Holoshitz’s study used the passive strategically to highlight the effects of domestic violence on women; the defence lawyers in Susan Ehrlich’s research used it strategically to downplay the agency of their clients. They used the same grammatical construction, but in different ways to suit their different aims.

What matters for feminist purposes is the aims: we can criticize particular uses of the passive without suggesting it should never be used at all. If we do that, we won’t just catch the cases where it works against the interests of women, we’ll also catch the cases where it can work in women’s favour. Language is a resource; let’s not make it into a straitjacket.

Thanks to Tamar Holoshitz for allowing me to make use of her unpublished thesis ‘More than Words: Passive Voice Use in Courtroom Depictions of Violence Against Women’ (Harvard University, 2010).