Not a safe word

Last weekend a writer named Will Saletan stirred up a hornet’s nest when he tweeted some advice for parents:

Teach your daughters to say No firmly, and mean it. Men sense women’s willingness to yield. Make clear you mean business.

Like other feminists, I saw this as a classic case of a man weighing in with very little understanding of the issue at hand, and no appreciation of the reasons why it was bound to provoke a storm of criticism. But the exchanges the tweet prompted, on Twitter and elsewhere, reminded me of another, perhaps less popular opinion I hold: that discussions of sexual consent and refusal very often present the issue of ‘saying no’ as less complicated than it really is.

In 2015, just before I started this blog, I wrote something on this subject which I never found a home for. In fact, I’d more or less forgotten it, until Saletan’s tweet made me remember why I’d felt the need to write it.  So I pulled it out, reworked it slightly, and—for whatever it’s worth—here it is.


In BDSM subcultures, participants in sexual encounters may agree in advance on a ‘safe word’—a word which can be uttered at any time to communicate the message ‘stop this now’. In theory, any word will do: all that matters is that the parties know it and agree to respect its meaning. But there are some words that can’t be used, and one of those words is ‘no’.

The fantasies played out in BDSM involve a dominant partner imposing their will on a submissive one. But the pleasure of imposing your will can only be experienced fully if the other appears to be unwilling. A show of resistance is part of the fantasy, enhancing the erotic charge for both partners. ‘No’, the prototypical verbal token of refusal, is used (along with other prototype expressions like ‘don’t’ and ‘stop’) to enact this simulated resistance. Consequently it cannot be a safe word, the word you utter when your refusal is real.

This principle doesn’t just apply in dungeons. When I was at primary school in the late 1960s, the girls had a playground chant that went: ‘Oh, Sir Jasper, do not touch me’. The game was to repeat this chant, leaving out the last word each time, like this:

Oh, Sir Jasper, do not touch me!

Oh, Sir Jasper, do not touch!

Oh, Sir Jasper, do not!

Oh, Sir Jasper, do!

Oh, Sir Jasper!

Oh, Sir!


I now know that these are the words of a rugby club song which is thought to date from the early 1950s. In its original form it includes a chorus after each line: ‘she lay between the lily-white sheets with nothing on at all’. The scenario evoked in the song has echoes of the popular Victorian melodramas in which an innocent girl is ‘seduced’ (i.e., raped) by an aristocratic villain. In the song, though, she isn’t innocent. Her resistance is simulated, there to be overcome.

At the age of 9 or 10 ‘Oh, Sir Jasper’ was just a game: we didn’t know where the song came from or understand what it was really about. (Today a 10 year-old might have more idea, but we were still pretty ignorant.) With hindsight, though, chanting these words was part of our informal education in the patriarchal rules of heterosexual conduct. We were absorbing the idea that a good girl refuses a man’s sexual overtures (and certainly does not make overtures to him). But we were also learning that her refusal is not sincere: really, she wants what he wants, she just can’t admit it straight away. The man’s job is to wear her resistance down, to persist until ‘Oh, Sir Jasper, do not!’ turns into ‘Oh, Sir Jasper, do!’

A decade after leaving primary school I would find myself in another group of women chanting a different set of words: ‘However we dress/ Wherever we go/ Yes means yes/ No means no’. ‘Yes means yes, no means no’—these are statements of the obvious, self-evident truths, tautologies. But when you put them together with the cultural script I’d first encountered as a young girl, they do not look quite so obvious, nor quite so simple.

The script says that whatever they want, women should offer some token resistance. If they say ‘yes’ too easily they risk being branded as sluts. (In an age when Teen Vogue promotes the joys of anal sex you might be thinking this rule no longer applies, but there is plenty of evidence that the charge of being a ‘slut’ (or ‘slag’ or ‘skank’) has not lost its power: it is a basic and ubiquitous component of the sexual bullying endured by thousands of teenage girls). The script also presupposes that it will be the man who asks the question, while the woman’s role is just to answer ‘yes’ or ‘no’. If she does the asking she is not just a slut, but potentially a ball-breaking nymphomaniac. For as long as these conventions persist, the unpalatable truth is that some women, sometimes, will offer resistance which is not meant as an absolute refusal. And some men will get pleasure from overcoming that resistance.

I want it to be crystal clear what I’m not saying here. I am not saying that women are to blame for being raped because they don’t always say ‘no’ and mean it as an unequivocal refusal. Nor am I suggesting that men can be excused for ‘misunderstanding women’s signals’ and believing women consent when they do not. Those defences are both used in rape cases, and I reject them absolutely (I’ll say more about why later on). What I’m saying is that the context in which heterosexual encounters take place is (still) one in which men are defined as sexual subjects, while women are defined as sexual objects. That understanding of their respective roles affects what each participant is allowed to say, and how their words will be understood. In this sexual and linguistic economy there can be no guarantee that a woman’s ‘no’ is always and by definition an unequivocal refusal, nor that her ‘yes’ is always and by definition an active, uncoerced expression of desire.

Of course it is entirely possible for a woman to intend to refuse sex or to consent to sex, and to express that in terms that she herself considers unequivocal. But the thing about language is, you’re never a free agent—at least, not when you’re talking to someone other than yourself. Humans are not mind-readers: we do not have direct access to other people’s intentions, but only to the words they utter. And to understand what other people mean we have to do more than just decode their words. We also have to make inferences about how the words were intended. What is meant may be quite different from the literal meaning of what is said (as in irony or sarcasm); the key to what is meant may lie in what is conspicuously not said (as in hinting or sulking).

Figuring out the meaning behind the words is the hearer’s job, and to do it s/he uses both contextual information and common-sense assumptions about the world. That last part is where the problem arises. Will Saletan’s claim that ‘Men sense women’s willingness to yield’ implies that men just respond to cues they detect in women’s behaviour, but in reality, the men he’s talking about assume women’s willingness to yield. They’re working from the script in which ‘women say “no” when they don’t mean it’ is a common-sense assumption, a truism. This is not a problem with the way individual women express themselves. It is a problem with the world in which they do it.

Men who have been accused of rape will often point out that they can’t read women’s minds. ‘She didn’t say “no”, so what was I supposed to think?’ And in many cases it will be true that she didn’t utter the actual word ‘no’. But it doesn’t follow that she wasn’t refusing. English-speakers very rarely communicate refusals by saying ‘no’, firmly or otherwise.

Refusing is one possible move in response to a proposal or an invitation; the other is accepting. These two options form what conversation analysts call a ‘preference system’. One response, acceptance, is ‘preferred’, and you can express it very simply and briefly. If a colleague asks me to go for a drink after work and I want to accept the invitation, I can say something like ‘great, see you in the pub’. It isn’t a problem that this response is brief and bald, because I can assume it’s what my interlocutor wants to hear (a person who issues an invitation is usually hoping it will be accepted). But if I don’t want to go to the pub I will need to take a bit more care, because (as it says on the old notice about not asking for credit), ‘a refusal may offend’.

Detailed analysis of real-life refusals shows there’s a formula we use to mitigate the offence. It goes: hesitate + hedge + express regret + give a culturally acceptable reason. As in ‘um, well, I’d love to, but I promised I’d be home early tonight’. Or ‘[pause] I’m sorry, but I’ve got a report to finish’.

Imagine responding to someone who suggests going for a drink after work with a simple unvarnished ‘no’. Or ‘no, I can’t’, or ‘no, I don’t want to’. The person you said this to would think you’d been raised by wolves. It’s curt, it’s rude, and it will be heard as arrogant or aggressive. Why would we imagine that saying ‘no’, firmly, is a reasonable thing to tell a woman to do in a situation where she has reason to fear the consequences of giving offence? Why would we blame her for trying to refuse diplomatically, when we’d do the same ourselves in far less risky situations?  And why would we believe that ‘men don’t understand anything less direct than “no”’? The formula for (non-sexual) refusals is used and understood by speakers of both sexes. It’s absolutely normal. Saying ‘no’ is not.

Since it’s not considered ethical to record people’s sexual encounters, linguistic researchers have no direct evidence about sexual refusals in real-life situations. But one study, carried out by Celia Kitzinger and Hannah Frith, gathered some indirect evidence by asking women in focus groups to talk about their experiences of refusing sex. The strategies these women said they used followed the formula for other kinds of refusals: they involved hedging, apologizing and giving acceptable reasons. In this context, an ‘acceptable’ reason was one that did not imply any lack of desire for the man who was asking. It was better to emphasize circumstantial obstacles—headaches, periods, early starts—or your own emotional problems (‘well, I’m flattered, but I’m just not ready for another relationship’). Most women agreed that ‘just say no’ was bad advice, especially if the man was putting pressure on you, because of the risk that it would make him angry, and prompt a physically aggressive response.

So, the issue isn’t whether ‘no means no’, and whether men understand that. And it’s certainly not whether women say it ‘firmly’ enough to show they ‘mean business’. The issue is whether men are capable of interpreting—i.e. inferring the intentions behind—the verbal strategies which are normally used to indicate a lack of enthusiasm for something another person proposes. And the answer to that question must be yes, since in all other contexts men use those strategies themselves.

Since Kitzinger and Frith’s study, other research has provided evidence that men are able to interpret refusals which don’t contain ‘no’. This points to another unpalatable truth: in most cases where men have sex with women who don’t want to have sex with them, the men must know that’s what they are doing. Some may persuade themselves otherwise (using the time-honoured script in which resistance is token, there to be overcome). Others just don’t care. None of them should be able to get away with it.

In the last couple of years a new consensus has emerged about the importance of educating young people about consent. Many universities are doing this, and in future it will be a required element of the Relationships and Sex Education curriculum in schools. Like most feminists, I’m in favour of this, if only because so many people still have no understanding of their basic rights and responsibilities under the law. But on its own, I think it will only make things slightly better (which I acknowledge isn’t nothing, given how bad they are at the moment). It may help to make refusals more intelligible as refusals—that is, challenge the part of the script which assumes ‘women’s willingness to yield’—but it won’t solve the problem of women saying yes, under social and emotional pressure, to sex they don’t really want, or being prevented from pursuing their own desires by the fear of being branded sluts.

Ultimately what we need to do is rewrite the whole script, not just the ‘saying no’ part. In an ideal world, sex wouldn’t just be consented to (like medical treatment, or the terms and conditions offered by internet providers). It would be an actively and mutually desired exchange between free and equal human beings. We are still a long way from that world; but while of course feminists must go on fighting for what women need in the present, we must also go on trying to imagine a more radically different, and better, future.