What are words worth? Thoughts on the pardoning of witches

Last month the Scottish government gave its support to a proposal to grant a posthumous pardon to people who were executed as witches. The campaign group Witches of Scotland estimates that between the passing of the Witchcraft Act in 1563 and its repeal in 1736, almost 4000 people were accused; around two thirds of them, more than 2500 people, were subsequently convicted and executed. As well as a pardon, the campaigners want an official apology and a public memorial to those who died.

In Scotland as elsewhere, a large majority of the victims of witch-hunting—around 84 percent of them—were women, and the campaign has been seen as a feminist issue. For some supporters its significance goes beyond the purely historical: it’s been suggested that the righting of this centuries-old wrong will also, in the words of Scottish Parliament member Natalie Don, ‘have an impact in challenging gendered and patriarchal attitudes in [present-day] society’. All of which raises some interesting questions about history, politics and (for reasons I’ll come to shortly) language.  

There are other cases where a pardon has been granted to a group of people who are considered, in retrospect, to have been criminalized unjustly. In 2016, for instance, the UK Parliament passed legislation pardoning anyone who had been convicted under the various laws that once prohibited consensual sex between men. Lord Sharkey, who proposed the relevant amendment, said that ‘a pardon is probably the best way of acknowledging the real harm done by the unjust and cruel homophobic laws, which thankfully we’ve now repealed’.

Not everyone agreed. George Montague, a gay man who was convicted of gross indecency in 1974, told the BBC he wanted an apology, not a pardon. ‘To accept a pardon means you accept that you were guilty’, he said. ‘I was not guilty of anything. I was only guilty of being in the wrong place at the wrong time’.

The Scots lawyer Andrew Stevenson has made the same point about the pardoning of witches:  

One pardons a wrongdoer, not the party wronged. Yet by means of a pardon conferred by statute the state is granting, not seeking, forgiveness. A pardon (of witches or anyone else) does not quash a conviction. It actually reaffirms its existence.

Pardoning is an example of the type of speech-act the philosopher J. L. Austin called a ‘performative’, meaning that the utterance of certain words actually performs, as opposed to just reporting, a specific action. Whereas statements like ‘it’s raining’ describe a state of affairs that exists independently of the speaker’s utterance (or doesn’t: I can check by looking out of the window), performative utterances like ‘I bet you £5 it rains today’ or ‘I swear to tell the truth, the whole truth and nothing but the truth’ are, in themselves, enactments of the bet or the oath. They bring a new reality into being—or at least, they do if they’re performed properly. Performatives don’t have truth conditions, but rather ‘happiness’ or ‘felicity’ conditions which must be met if the performance is to have the intended effect. In the case of pardoning those conditions include the prior existence of a crime or a wrong for which the person being pardoned was responsible. Consequently, Stevenson argues, pardoning the witches cannot achieve the intended effect: it will not bring into being a new reality in which they were never guilty of anything.

But while this argument may be legally correct, for most people in modern Scotland it is surely beside the point, since they already take the witches’ innocence for granted. Witches are a different case from gay men: a pardon is not being proposed because of a change in society’s attitudes to what they do (people used to think witchcraft was wrong, but that has now been recognized as an unjust prejudice), but rather because we now reject the idea that witchcraft is, or ever was, a real phenomenon. To us it is self-evident that the accusations made against witches—for instance that they had killed their neighbours’ cows with curses, transformed themselves into owls or cavorted on beaches with Satan—were false: no one could have been guilty of such absurd and impossible crimes. In that sense you could argue that the wrong has already been righted, to the extent it ever can be. But in that case, what is a pardon meant to accomplish? 

A cynical answer might be that it’s ‘performative’ not (just) in Austin’s sense, but in the now-popular pejorative sense: an ostentatious but superficial display of concern intended mainly to boost the performer’s claim to the moral high-ground. That criticism has sometimes been made about the formal apologies other governments have offered for more recent wrongs like the removal of indigenous children from their families in Australia, or the abuse of women in Ireland’s Magdalene Laundries. The problem isn’t necessarily that they’re insincere (I’m sure the Scottish government does genuinely believe it was wrong to execute people for witchcraft). But if the regret politicians express for the way certain people were treated in the past is not accompanied by any concern about the injustices those people still suffer in the present, their performances may be dismissed as just self-serving, empty words.

What about Natalie Don’s assertion that revisiting the history of witch-hunting in Scotland will serve the purpose not only of atoning for past sins, but also of ‘challenging gendered and patriarchal attitudes’ in the present? This argument is often made about the commemoration of atrocities: we should remember the Holocaust or the Atlantic slave trade not only as a mark of respect to the victims, but also as a way of educating ourselves, and so preventing comparable horrors in future. Painful though it may be, we can only learn from history if we face up to what really happened and why.  

The witch-hunts of the early modern period have not generally been commemorated in this way: they’ve been mythologized and trivialized by the entertainment and tourism industries. In the English town of Lancaster, the site of the 1612 Pendle witch-trials (in which ten women and two men were convicted and executed), the historian Rachel Hasted reported in 1984 that

The local tourist bureau has just launched an advertising campaign headlined The Magick of Lancaster, with a 17^th century woodcut of several women being hanged…Tourist shops all over the county sell little black-hatted figures on pipe-cleaner broomsticks and guide-books to ‘the witch country’ with lurid accounts of their doings.

A quick online search confirmed that some tourist attractions in the area are still presenting Lancashire’s witch-hunts (aka the torture and killing of human beings) as if they were quaint local traditions on a par with cheese-rolling or dancing around the maypole.

As Silvia Federici argues in her 2018 book Witches, Witch Hunting and Women, the lurid tales and tacky souvenirs both sensationalize and sanitize history: they recycle an image of the witch that was originally constructed by her persecutors, while glossing over the reality of her persecution. Federici would presumably applaud Scotland for facing up to that reality. But exactly how to present ‘what really happened and why’ remains a complicated question. Historians, including feminists, hold different and sometimes conflicting views.       

Back in the 1980s, Rachel Hasted took issue not only with the crass pop-history in tourist guides, but also with what she dubbed a ‘new myth of the Witch’ popularized by feminists. This account posited that the witches were ‘wise women’ and healers, dispensing plant-based natural remedies to the people of their peasant communities, and (in some versions of the story) holding on to ancient pagan beliefs. Witch-hunts were said to have occurred when these long-established activities, and the traditional knowledge that underpinned them, began to be seen as a threat to the authority of the church and the interests of the men who controlled the emerging medical profession.

But in Hasted’s view the Lancaster records did not support this interpretation. The Pendle women were pious Christians who neither laid claim to nor were accused of using any knowledge of medicinal herbs. In Scotland, too, it seems that few women tried for witchcraft were known in their communities as healers. These cases might fit better with an alternative account in which women were victimized not because the authorities felt threatened by their power, but on the contrary, because they were so powerless.

Some research suggests that accusations of witchcraft were disproportionately made against individuals who were already marginalized because they were old, disabled, or without regular employment, and consequently so poor that they would sometimes beg or steal from their neighbours. It wasn’t a coincidence that many of them were women: the exclusion of women from many kinds of work made them vulnerable to poverty, especially if they were single. But in this account what made them targets was not their sex in itself, but the perception of them as troublesome and undesirable. The implication is that accusers were motivated less by fear of witchcraft than by a desire to see people they disliked, disapproved of, or had some kind of quarrel with, punished by the authorities.

That desire has existed in every age, and been exploited by authoritarian regimes of all kinds. We now know, for instance, that in both Nazi Germany and Communist East Germany, many people who informed on their neighbours, workmates or fellow-students did so for personal rather than ideological reasons, to settle scores with their enemies or gain an advantage over their rivals. Recently there has been some discussion of this in Scotland, in relation to a controversial new hate-crime law which was finally passed last March. One concern expressed by critics of the legislation was that it would encourage zealots and grudge-bearers to drag the state into their personal or political feuds.   

The Scottish witch-pardon might also invite questions about the complacency of a society that condemns past abuses of women while tolerating comparable abuses in the present. When I say ‘comparable’, I obviously don’t mean that women in Scotland are still being executed for witchcraft. But many of the same things that were said about witches are still regularly said about women in modern courtrooms—for instance that they are liars, manipulative, vengeful and sexually predatory.

The actual language of witch-hunting is not dead either. In 2020, after the trial of former SNP leader Alex Salmond on multiple sexual assault charges ended in his acquittal, commentators in the Scottish media used it to attack some of the women journalists who had covered the case. A programme fronted by Kirsty Wark was said to have featured a ‘coven’ of women who were likened to the three witches in Macbeth; one of them, Dani Garavelli, was also described as ‘the Rapefinder-General’. The sexism of this rhetoric, which recasts women, the original witch-hunt victims, as persecutors of innocent men, is not, of course, unique to Scotland. But nor does Scotland have any special claim to have moved beyond it.

I’m not suggesting that feminists should oppose the pardoning of witches (or formal apologies or memorials to them, which IMHO might be apter choices), but I do think we should consider what we want these performances to accomplish. If all they accomplish is to distance the living from the superstitious beliefs of their long-dead ancestors, that’s fine as far as it goes, but there’s nothing especially feminist about it. For the gesture to ‘have an impact in challenging gendered and patriarchal attitudes’, it would need to go beyond saying ‘look, we’re not like our ancestors, we find what they did abhorrent’, and address the ways in which—regrettably—we are still like them.

What’s in a frame? Misogyny/hate

‘Women’, wrote Germaine Greer in 1970, ‘have very little idea how much men hate them’. Fifty years later, it seems we have woken up. The problem of woman-hatred is now widely acknowledged and discussed; in Britain there’s growing pressure for misogyny to be legally recognised as a form of hate. Campaigners have presented this as a question of parity, saying that the law should ‘treat misogyny like racism or homophobia‘ (which are already covered, along with religious hatred, transphobia and hostility to people with disabilities). It’s an argument that has resonated with many feminists, and it’s now under serious consideration. Though the Scottish Parliament recently rejected a proposal to include women in new hate crime legislation, a working party has been set up to examine the issue further. Meanwhile in England and Wales, the Law Commission issued a consultation paper last year which did recommend that the law should be extended. Since the outcry that followed the murder of Sarah Everard this proposal has attracted more mainstream political support.

So, it looks as if change is coming; but will that be a step forward for women? On reflection I have my doubts, and in this post I’m going to try to explain them.

In England and Wales currently there isn’t a specific hate crime law, but rather a patchwork of provisions threaded through other laws. One key provision is in the Criminal Justice Act 2003, which says that if someone who committed a criminal offence ‘demonstrated, or was motivated by, hostility on the grounds of race, religion, sexual orientation, disability or transgender identity’, the court should treat that as an aggravating factor and consider whether to impose a harsher penalty. This also indirectly brings what is popularly known as ‘hate speech’ into the picture (though the term itself has no status in English law), in that the language someone used may be treated as evidence of hostile motivation. Other legal provisions target verbal behaviour more directly. The Public Order Act 1986 includes an offence of ‘stirring up hatred’, which will often be done by way of language (one recent case involved a series of anti-Muslim posts on Gab), and also one of using ‘threatening words and behaviour with intent to cause harassment and distress’.

The Law Commission has recommended that these provisions should be extended to cover hostility on the grounds of sex, or hostility to women (which of these options to prefer is one of the questions posed in the consultation). To reach that conclusion, it explains that it applied three tests:

1. Demonstrable need: whether there is evidence that crimes against women are (a) prevalent and (b) linked to hostility and prejudice;
2. Additional harm: whether women victims are more severely impacted by crimes which are motivated by hostility/prejudice, and whether these also cause harm to other members of the target group (‘secondary victims’);
3. Suitability: whether an extension of the hate crime framework to crimes against women would be workable in practice and compatible with the rights of other groups.

The Commission concluded that the first two tests were met. Crimes which disproportionately target women (e.g. rape and sexual assault, domestic violence, forced marriage, FGM, street harassment, online abuse) are prevalent, rooted in prejudice, and have an impact on women in general. But some questions remain unresolved. One is the practical feasibility of extending the law, given the high number of crimes against women and the fact that the justice system is already overstretched. Another concerns the status of domestic violence/abuse, which some argue should be excluded because it isn’t motivated by hostility to women as a group; rather it arises within specific intimate relationships, which could be same-sex partnerships, or heterosexual ones where the abusive partner is the woman. The consultation paper does suggest that sex (more specifically, femaleness) should become ‘a protected characteristic for the purposes of hate crime law’, but it asks if there should be a ‘carve out’ for domestic violence.

This is one reason why some feminists are concerned about the Commission’s proposals. They fear the effect will be to create a new hierarchy of crimes against women, taking us back to the days when attacks carried out by strangers were seen as ‘worse’ than violence perpetrated by someone the victim knew. Feminists have also drawn attention to an even more basic problem, namely the failure of the criminal justice system to enforce the laws we already have. What good, they ask, is creating new offences, or giving the courts power to impose harsher penalties, when most of the crimes women currently report do not lead to a prosecution, let alone a conviction? And that’s not only because the system is under-resourced. Women are also denied justice because of longstanding biases, both in the system and in the surrounding culture. How can we trust institutions which are themselves riddled with misogyny to enforce new anti-misogyny laws effectively and fairly?

Campaigners for new legislation often argue that it will help to drive institutional and cultural change, by sending the message that ‘this is serious and will no longer be tolerated’. But in the case of crimes against women, this message often turns out to be no match for the prejudice it was meant to shift. For instance, this month the media reported on a school in Liverpool where girls had been told to wear shorts under their uniform skirts after several of them were ‘upskirted’ (i.e., boys took pictures of their underwear) on a transparent staircase in the sixth-form building. This story caught my eye because upskirting was recently the subject of a successful campaign to make it a criminal offence (it became one in 2019). The Liverpool boys, who were over 16, could in theory have been reported to the police. I’m not saying that would necessarily have been the right thing to do. I’m sympathetic to the argument that where possible we should try to educate young people rather than criminalising them. But it’s telling that this school did neither. Instead it chose to punish the girls, by imposing a dress-rule that would make them feel uncomfortable, undignified and as if they were the ones at fault.

Even if I had more faith in legislation as a remedy for social ills, I would still want to ask whether extending hate crime laws sends the right message about misogyny. My doubts on that score reflect my interest in language–in words and meanings and what might be called ‘discursive framing’. Treating misogyny ‘like racism and homophobia’ means slotting women into a pre-existing frame which was not originally designed for them. And that raises the question of how well the frame fits.

Categories have their prototypical members, the examples that spring to mind first when we encounter their generic label. Our prototype for the category ‘bird’, for instance, the kind of bird we’ll draw if we’re instructed simply to ‘draw a bird’, is something that looks like a robin or a sparrow, not an ostrich or a penguin. In the case of hate crime/hate speech the prototype is hatred of a racial or ethnic Other. This is where it began in the UK, with the outlawing of ‘incitement to racial hatred’ in the 1960s. Later religious hatred was added, and this was not a big stretch because it’s close to the prototype: often it’s as much about race/ethnicity as religious belief per se. The other types of hatred now covered by the law—homophobia, transphobia, hostility to disabled people—share some features with the prototype, in that they target minorities who are perceived as ‘different’, as outsiders. And there’s another thing these target groups have in common. Hatred of them is linked, historically and in our minds, to right-wing extremism. The prototypical (western) right-wing extremists, the Nazis, regarded Jews, homosexuals and disabled people as inferior and impure, and they did their best to exterminate them.

But this prototypical form of hate, the kind that motivates genocides and pogroms, that calls for the ‘repatriation’ of Black British people to ancestral homelands they have never set foot in or advocates the involuntary sterilisation of the ‘unfit’, is not what (most) misogyny is about. Though misogynists do see women as Other and lesser beings, who exist only in relation to men and for men’s benefit, few of them wish for a world in which women are not available to meet their emotional, domestic, sexual and reproductive needs. What they want is not to eliminate women, or to live entirely apart from them, but to exploit, dominate and control them. Misogyny, in short (as the philosopher Kate Manne has argued), is not a generalised hatred of women, but rather the punishment of women who refuse to stay in their subordinate place or to meet what men regard as their obligations. The extreme right has no monopoly on that kind of punishment, nor on the belief system which justifies it. Some forms of misogyny are so common and unremarkable, it hardly makes sense to label them ‘extreme’.

Because misogyny is so different from the prototype which hate crime laws were designed for, it’s difficult to just ‘add women and stir’. The Law Commission’s question about whether there should be a ‘carve out’ for domestic violence is one illustration of this difficulty: violence against an intimate partner is commonly understood as the consequence not of hate, but of its opposite, love, ‘gone wrong’. Murderers and family annihilators are said to have killed their ex-partners and sometimes their children because they couldn’t bear the pain of separation, rejection or ‘betrayal’.

I would have no hesitation in calling this behaviour misogyny, but I think what’s behind it is less a hatred of women than a sense of entitlement in relation to women. I would apply the same reasoning to, for instance, child abuse and elder abuse: what motivates these forms of violence is surely not a generalised hatred of children or old people, but rather a feeling of entitlement to use and abuse them, to exploit their relative powerlessness for your own gratification, or to punish them for making what you see as unreasonable demands. We should be able to recognise the seriousness of these forms of abuse, and to punish them as they deserve, without having to put them into a frame that doesn’t fit.

The notion of misogynist hate speech raises similar questions. According to the philosopher Alexander Brown, a typical legal definition of hate speech looks something like this:

(1) Speech [or other expressive conduct] (2) concerning one or more members of a protected group or class (3) that involves [expresses, incites, justifies] feelings of hatred toward group members.

Brown argues that this is too narrow, and that a better definition would reflect the way the term ‘hate speech’ is used/interpreted in ordinary language—which, as he points out, does not always treat ‘feelings of hatred’ as central. He goes on to offer a list of the types of speech (or writing) which in his view would ‘intuitively fall under the ordinary concept [of] hate speech’:

1. Slurs, epithets or insults vilifying members of historically victimized groups (e.g. the N-word, ‘dirty Jew/faggot’)
2. Forms of speech that assert or imply a group’s inferior or sub-human status (e.g. ‘these people [asylum seekers] are cockroaches’)
3. Group defamation or negative stereotyping: the false/overgeneralized attribution of qualities/behaviour to a group (e.g. the blood libel; ‘homosexuals abuse children’)
4. Incitement: advocating, justifying or glorifying hatred, violence or discrimination against a group (e.g. ‘kill all Xs’; symbols used to intimidate, e.g. burning crosses/nooses/swastikas)

Although this list makes no explicit reference to women–all the examples relate to race/ethnicity, religion and sexual orientation–it’s not hard to see how the framework might be applied to them. Clearly, there are slurs vilifying women (‘bitch’, ‘cunt’, ‘whore’); assertions of female inferiority and subhumanity are staples of online discussion among incels, MGTOWs et al.; negative stereotyping of women is commonplace; and under the heading of incitement/intimidation we could include the threats with which women are bombarded online, often expressed in the linguistic register to which Emma Jane has given the label ‘rapeglish’. Maybe we could even consider flashing, or sending unsolicited dick pics, as the misogynist analogue of the noose and the swastika. The problem with Brown’s taxonomy, then, isn’t that women can’t be slotted in at all. The problem is how much that leaves out.

One thing it leaves out is a feature of many kinds of misogynist discourse: the use of, specifically, sexualised speech to enact power and domination over women. A great deal of what women experience as intrusive, degrading or intimidating male behaviour is couched not in the language of hate, but ostensibly in the language of desire or sexual interest. Everyday street remarks like ‘nice tits’, or ‘give us a smile’, certainly don’t ‘intuitively fall under the ordinary concept of hate speech’: on the surface they seem appreciative rather than hostile, and men are quick to exploit that if women object (‘what’s the matter, can’t you take a compliment?’) But these comments are not innocent or harmless. As well as underlining women’s status as sexual objects, they are pointed reminders that women in public space are under constant male surveillance and must conduct themselves accordingly.

Other kinds of misogynist speech, like ‘rapeglish’, are closer to the ‘ordinary concept of hate speech’ because they’re explicitly violent and threatening. But even rapeglish tends not to be put in the same conceptual box as, say, racist or anti-semitic rhetoric, because its graphic sexual content prompts people to read it as a display of individual pathology rather than the expression of a hateful ideology. The same is true of indecent exposure, which is viewed more as a compulsion afflicting some (inadequate or disturbed) men than as an intentional form of expressive behaviour which is meant to humiliate and intimidate. Once again, the sexualised nature of the behaviour obscures the political purpose it serves. The philosopher Rae Langton has made a similar point about pornography, arguing that its sexual content tends to disguise its ‘status…as propaganda’. ‘For racial hate speech’, she writes, ‘hierarchy and subordination look like what they are… For pornography [they] look like what they are not–namely, the natural sex difference’.

Our belief in ‘the natural sex difference’ also makes it possible for certain non-pornographic messages that might otherwise be judged as hate speech to escape that categorisation. Consider the greeting card below, which was photographed in a bookshop: the fact that it was openly on display suggests that most people wouldn’t consider it hateful, even if some might find it tasteless.

Why not, though? Because it’s saying you can’t ‘shoot [women] and bury them in the garden’ rather than advocating that course of action? Because it’s clearly meant to be a joke? Maybe; but if the word on the card were not ‘women’ but, say, ‘Jews’ or ‘gays’, neither of those considerations would make it acceptable. Animosity between men and women (aka the eternal ‘battle of the sexes’) is understood to lie beyond the realm of politics and even culture: it’s seen as natural, universal and—crucially—reciprocal (just like the desire which draws the warring parties together). That’s why the one word you could replace ‘women’ with and still have an acceptable product is ‘men’—though you’d be glossing over the fact that in reality women very rarely kill men, whereas (in Britain) men kill women at a rate of 2-3 a week.

I’m not using these examples to argue that more kinds of speech should be legally defined as hate speech. I’m suggesting that ‘hate’ may not be the right frame for understanding or addressing the issue of misogyny. Feminists who favour that frame argue that equality requires inclusion: the exclusion of woman-hatred from existing provisions sends the message that women are less important than other groups, and that misogyny is less serious than other hatreds. But while I agree that misogyny is a real and serious problem, I don’t think that means it is, or should be treated, exactly like racism or homophobia. To me, taking it seriously means considering it on its own terms. Women need to be able to frame a response that begins from our experiences, our needs, and our ideas about what would truly make a difference.    

What makes a word a slur?

Content note: this post contains examples of offensive slur-terms. 

Last week, the British edition of Glamour magazine published a column in which Juno Dawson used the term ‘TERF’ to describe feminists (the example she named was Germaine Greer) who ‘steadfastly believe that me—and other trans women—are not women’.  When some readers complained about the use of derogatory language, a spokeswoman for the magazine replied on Twitter that TERF is not derogatory:

Trans-exclusionary radical feminist is a description, and not a misogynistic slur.

Arguments about whether TERF is a neutral descriptive term or a derogatory slur have been rumbling on ever since. They raise a question which linguists and philosophers have found quite tricky to answer (and which they haven’t reached a consensus on): what makes a word a slur?

Before I consider that general question, let’s take a closer look at the meaning and history of TERF. As the Glamour spokeswoman said, it’s an abbreviated form of the phrase ‘Trans-Exclusionary Radical Feminist’; more specifically it’s an acronym, constructed from the initial letters of the words that make up the phrase. Some people have suggested this means it can’t be a slur. I find that argument puzzling, since numerous terms which everyone agrees are slurs are abbreviated forms (examples include ‘Paki’, ‘Jap’, ‘paedo’ and ‘tranny’). But in any case, there’s a question about the status of TERF as an acronym. Clearly it started out as one, but is it still behaving like one now?

To see what I’m getting at, consider an acronym from the 1940s: ‘radar’. Do you know what all the letters stand for? I do, but only because I’ve just looked it up. I’ve been using the word for 50-odd years without realising it meant ‘RAdio Detection And Ranging’—a feat made possible by the fact that ‘radio detection and ranging’ isn’t really what it means any longer. Over time it’s become just an ordinary word, which is used without reference to its origins as an acronym. No one mentally expands the letters R-A-D-A-R into words; no one imagines that ‘gaydar’ must be short for ‘gay detection and ranging’. Also (a trivial but telling sign) no one now writes ‘radar’ in all caps.

I’ve been writing TERF in all caps, but these days you also see it written ‘Terf’ or ‘terf’. That’s one sign it’s going the same way as ‘radar’, becoming a word which can be used without knowing what the letters of the original acronym stand for. Another sign is the way it’s now used to describe people (e.g., men) who don’t fit the original specification, in that they aren’t radical feminists. It looks as if at least some users of the term don’t define it strictly as meaning ‘trans-exclusionary radical feminist’, but use it with a more generic meaning like ‘transphobic person’.

This kind of change is common in the history of words. Word-meaning is inherently unstable, liable to vary among different groups of users and to change over time, because we don’t learn the meanings of most words by looking them up in some authoritative reference book, we figure them out from our experience of hearing or seeing words used in context.

It’s easy to see how that might shift the meaning of TERF in the way I’ve just suggested. Imagine you hear two of your friends discussing a mutual acquaintance who they refer to as a TERF. You’ve never encountered the term before and you have no way of knowing it’s a short form of a longer phrase (because it’s a true acronym, pronounced not as a series of letters but as a single syllable that rhymes with ‘smurf’).  So you listen to what’s being said about the TERF in question and make the simplest inference compatible with what you’re hearing: that TERF means a transphobic person.

If TERF’s meaning has started to shift that’s actually a sign of its success (words evolve as they spread to new users and contexts). But it makes the argument that TERF is just a neutral descriptive label for a specific group of people less convincing. That argument either takes no account of the way usage has changed over time, or else it’s a version of the etymological fallacy (‘however people actually use a word, its original meaning is the true meaning’).

But the fact that a word isn’t a neutral description doesn’t necessarily mean it’s a slur. We’re back to the question philosophers and semanticists have found so tricky: on what basis can we say that a word is a slur?

As I’ve already mentioned, the people who’ve written on this subject don’t agree on what the answer is. And after reading their various accounts, I’m not sure I believe there’s a single right answer. Rather, I think there are a number of criteria which need to be considered. If we’re in doubt about a word’s status as a slur, we can try asking the following questions, and then looking at the overall balance of the answers.

My first two questions are based on what the philosopher Jennifer Hornsby proposes as the two fundamental features of a derogatory term or slur.

Is the word commonly understood to convey hatred or contempt?

Does the word have a neutral counterpart which denotes the same group without conveying hatred/contempt?

This definition seems to have been constructed using racial/ethnic slurs as a prototype. In these cases it’s generally understood that the slur term, used in preference to a neutral term which denotes the same group of people, communicates hatred/contempt as part of its meaning (that’s the difference between, say, ‘Jew’ and ‘kike’). This doesn’t help us much with terms like TERF whose status as slurs is disputed. TERF is certainly understood by some people to convey hatred and contempt, but others deny it conveys those things.

It’s also unclear whether there’s a neutral term which TERF contrasts with. TERF doesn’t so much refer to a pre-existing group as bring a new category into existence (there was a pre-existing group of radical feminists, but they weren’t defined as a category by the belief that trans women are not women, and in fact they still aren’t, since not all radical feminists hold that belief). So, to decide whether TERF is a slur we need to ask some other questions.

Do the people the word is applied to either use it to describe themselves or accept it when others use it to describe them?

Both parts of this question are important. If a group of people voluntarily use a word to describe themselves, then—on the assumption that people don’t generally slur their own group—you might conclude the word isn’t a slur. However, this does not allow for the possibility that a term might be a marker of identity and solidarity when used within the group, while remaining a slur if it’s used to/about the group by outsiders. (The classic example is the solidary use of the N-word among (some) Black people: it doesn’t make it OK for white people to use it. ‘Dyke’ for ‘lesbian’ is another example: fine if you are a lesbian, suspect if you aren’t. ) There are also jocular, ironic and self-mocking uses which don’t undermine the status of a word as a slur (women friends might refer to themselves in private as ‘sluts’ or ‘bitches’, but they wouldn’t accept being described in those terms in public or by non-intimates).

With TERF, I’d say the answer to both parts of the question is ‘no’. There may be people who use TERF ironically/self-mockingly in private, but I’m not aware of any who publicly define themselves as TERFs, and it’s common for those who are called TERFs by others to reject the label. Note that these observations concern attitudes to the word: there are certainly some feminists who publicly affirm the belief mentioned by Juno Dawson, that trans women are not women, but they may still deny being TERFs. This suggests they see TERF in the same way members of a certain ethnic group might see an ethnic slur: ‘yes, I am a member of the group you mean, but no, I do not accept the implications of the name you’re calling it by’.  Which brings me to the next question:

Do the people the word is applied to regard it as a slur (e.g. do they describe it explicitly as a slur, protest against its use, display offence/distress when it is used)?

For some writers, a ‘yes’ to this question is enough on its own to make a word a slur.  Luvell Anderson and Ernie Lepore argue that

…no matter what its history, no matter what it means or communicates, no matter who introduces it, regardless of its past associations, once relevant individuals declare a word a slur, it becomes one [emphasis in original]

What these writers are trying to account for is the fact that labels which were previously considered acceptable, or even polite, can get redefined as slurs (examples include ‘Negro’ and ‘coloured’), and the reverse may also happen (‘Black’ was not always acceptable, and ‘queer’ used to be unambiguously a slur). This isn’t a matter of what the term means (the literal meaning of ‘Black’ and ‘Negro’ is the same), but rather depends on the perceptions of ‘relevant individuals’ (members of the target group) at a particular point in time. If they declare a term offensive, then it’s offensive: it’s idle for non-members of the group to tell them they have no business taking offence.

On this criterion, TERF is indisputably a slur.  Many individuals who have been described as TERFs have called it a slur, protested against its use (witness the complaints about Juno Dawson’s column) and explicitly said that it offends them. But I’m reluctant to make that the sole criterion. I agree that for something to be a slur it’s necessary for members of the target group to regard it as offensive, but I’m not sure that’s a sufficient condition (and what do you do about cases where the target group is split? ‘Queer’, for instance, divides opinion in the LGBT community).

As a sociolinguist (unlike the writers I’ve been referencing), I’m also dissatisfied with the implication that members of a group just arbitrarily and randomly decide that, for instance, ‘queer’ has ceased to be a slur or ‘Negro’ has now become one. I think these developments can be related to the changing social and political contexts in which words are used (for instance, the context for the ‘unslurring’ of ‘queer’ was the surge of radical activism prompted by the HIV-AIDS epidemic). Perceptions of words have to be seen in relation to what the words are being used to do, either by the group itself or by its opponents. So another question I would want to ask is,

What speech acts is the word used to perform?

If a word is just a neutral description, you might expect it to be used mainly for the purpose of describing or making claims about states of affairs. If it’s a slur, you’d also expect it to be used for those purposes, but in addition you might expect to see it being used in speech acts expressing hatred and contempt, such as insults, threats and incitements to violence. (By ‘insults’ here, incidentally, I don’t mean statements which are insulting simply because they use the word in question, but statements which say something insulting about the group, e.g. ‘they’re all dirty thieves’.)  

There’s evidence that TERF does appear in insults, threats and incitements. You can read a selection of examples (mostly taken from Twitter, so these were public communications) on this website, which was set up to document the phenomenon. Here are a small number of items from the site to give you a sense of what this discourse looks like:

you vile dirty terf cunts must be fuming you have no power to mess with transfolk any more!

I smell a TERF and they fucking stink

if i ever find out you are a TERF i will fucking kill you every single TERF out there needs to die

why are terfs even allowed to exist round up every terf and all their friends for good measure and slit their throats one by one

if you encounter a terf in the wild deposit them in the nearest dumpster. Remember: Keeping our streets clean is everyone’s responsibility

Precisely because it was set up to document uses of TERF as a slur, this site does not offer a representative sample of all uses of the term, so it can’t tell us whether insulting/threatening/inciting are its dominant functions. It does, however, show that they are among its current functions.  It also points to another relevant question:

What other words does the word tend to co-occur with? 

It’s noticeable that on the website I’ve linked to, TERF quite often shows up in the same tweet as other words whose status as slurs is not disputed, like ‘bitch’ and ‘cunt’. Other words that occur more than once or twice in these tweets include ‘disgusting’, ‘ugly’, ‘scum’ and a cluster of words implying uncleanness (‘smell’, ‘stink’, ‘garbage’, ‘filth’)—which is also a well-worn theme in racist and anti-Semitic discourse.

One of the clues we use to infer an unfamiliar word’s meaning in context is our understanding of the adjacent, familar words; the result is that over time, recurring patterns of collocation (i.e. the tendency for certain words to appear in proximity to one another) have an influence on the way the word’s meaning evolves. The examples on the website are too small and unrepresentative a sample to generalise from, but if the collocations we see there are common in current uses of TERF, that would not only support the contention that it’s a slur, it might also suggest that the word could become increasingly pejorative.

In summary: TERF does not meet all the criteria that have been proposed for defining a word as a slur, but it does meet most of them at least partially. My personal judgment on the slur question has been particularly influenced by the evidence that TERF is now being used in a kind of discourse which has clear similarities with hate-speech directed at other groups (it makes threats of violence, it includes other slur-terms, it uses metaphors of pollution). Granted, this isn’t the only kind of discourse TERF is used in, and it may not be the main kind. But if a term features in that kind of discourse at all, it seems to me impossible to maintain that it is ‘just a neutral description’.

I believe in open debate on politically controversial issues, so I’m not suggesting the views of either side should be either censored or protected from criticism. My point is that when one of the key terms used in the argument has become a slur, it is no longer fit for any other purpose, and the time has come to look for a replacement.