What’s going on with ‘it’s okay to be white’?
How philosophical accounts of the difference between what a sentence means, and what someone means by saying it, illuminate the controversy.
A couple of weeks ago, the Australian senate voted on a motion for the Senate to acknowledge that “it is a motion calling OK to be white”. The motion was moved by the populist right wing One Nation party, but the controversial aspect is that it gained the support of the governing conservative Coalition. The motion failed, but the damage was done. While the leader of the government in the Senate, Matthias Cormann, was quick to explain that their apparent support for the motion was due to ‘administrative error’, considerable opprobrium was directed at the government over this vote. The furore was rekindled last week with the appearance of signs bearing the same phrase outside the electoral offices of MPs who were publicly critical of the original vote.
The initial controversy stems from the history of the phrase. As has been detailed by the Anti-Defamation League, the phrase emerged as a conscious attempt to ‘troll’ the anti-racism movement by those sympathetic to white supremacist movements. The idea was to use an apparently innocuous phrase which would nevertheless signal racist ideology to those who were attuned to its presence. When anti-racists objected to the slogan, the idea was, they would be revealing themselves as hypocritical racists themselves.
One Nation, no strangers to trolling the Australian Senate, clearly had this background in mind given the motion they proposed. The reaction was indeed as predicted: the left-leaning opposition parties (Labor and the Greens) lined up to condemn the motion, but many people found themselves confused at the apparent ‘reverse racism’ on display:
I think we have to acknowledge that there is some substance to these queries. Pointing out the problematic history of the phrase doesn’t help these Twitter users, and many people like them, to understand why this phrase signals racist ideology in the first place. Why is this phrase not innocent? How can someone reject the claim that it is okay to be white, without thereby committing themselves to the racist belief that it is not okay to be white? And why are other parallel phrases — such as ‘it’s okay to be different’ — not similarly controversial?
Some useful tools for explaining how this phrase communicates what it does have been developed by philosophers of language. Of particular use for us is the influential work of Paul Grice on the notion of conversational implicature. Grice was trying to understand the difference between what a sentence literally means (just given the meanings of the words in it, and how they are put together), and what someone who says it means by it. The contrast can be seen in many familiar examples. Suppose Andrew asks Chloe whether she is going to a party to which they’ve both been invited, and Chloe replies ‘I have to work’. What her sentence means is just what it overtly states: that she has to work. But what she means by it is that she is not going to the party, and that is what Andrew will take her to have communicated.
Grice was trying to understand two very fundamental issues about this sort of language.
- How does what Chloe literally says manage to communicate something quite different? and
- How it is that Andrew, Chloe’s hearer, was able to work out what she meant by what she said?
Grice was the first to offer a systematic theory of this phenomenon, answering both questions. His insights remain the foundation for almost all contemporary approaches.
Grice proposed that people involved in a conversation ought to obey a Cooperative Principle, which (roughly) says that when you speak, you should make a positive contribution to the aim of the conversation you are participating in. He argued that this means you ought to obey these subsidiary rules:
- Relevance: say things that are relevant to the topic of the conversation;
- Quality: say things that you believe to be true;
- Quantity: say things which are adequate to what the conversation needs, neither uninformative nor over-detailed; and
- Manner: express yourself clearly and concisely.
Because we ought to obey these rules, or maxims, in order for conversations to go well, Grice thought that most of time speakers do in fact obey these maxims. If mostly we were uncooperative, conversation wouldn’t be either fun or useful, and we wouldn’t do it so much. Crucially, not only are we typically cooperative, but we also know (or at least assume until we get evidence otherwise) that our conversational partners are being cooperative too.
Grice showed how, by relying on the assumption of joint cooperation in a conversation, we can work out what someone else might mean by her speech. In our earlier case of what Chloe said, a broadly Gricean account of how Andrew implicitly reasons might be as follows:
What Chloe literally said was that she would be at work. She is being cooperative, so this is true (Quality). But this is only relevant if she believes this information has some connection to the party, and the only plausible connection is that her work conflicts with attending the party (Relevance). So Chloe thinks that she will not attend the party. It would be pointlessly detailed to say this explicitly, because I can work it out, so that is why she didn’t say it directly (Quantity). So that is what she communicates.
This is all much more laborious than anything we’d actually do in conversation. But most philosophers and linguists agree that we learn how to do something like this automatically and tacitly as we use language socially. There is some developmental evidence for this hypothesis, as many children and neuro-atypical adults have trouble taking things too literally — that is, trouble working out what, apart from the literal content, is intended to be communicated. That is some evidence that non-literal aspects of communication are separable from the literal content of words and sentences.
How does all this apply to ‘It’s okay to be white’? One case commonly touted as success story for Grice’s framework is how it handles what are known scalar implicatures (also known as limit implicatures). These appear in cases like this: suppose Evelyn, a teacher, says of her class ‘Someone got an A’, we take her to be implying that only some students got an A. This is despite what she literally said being compatible with all students getting an A. (While it may be false, there is nothing incoherent about uttering the sentence ‘someone got an A — indeed, everyone did’.) The idea here is that if Evelyn relevantly says ‘someone got an A’ while she secretly knows that everyone did, then she would not be being cooperative — she would be saying less than is required by the conversation, and thus violating Quantity. So she is only being cooperative if she cannot truthfully say something stronger than what she actually said. So what she said implies ‘someone got an A but not everyone did’, which amounts to ‘only some students got an A’.
This account of scalar implicatures help us understand what is problematic about saying ‘It’s okay to be white’. But we need to do some other reasoning as well, to explain why there is implicature here at all. So there are two parts to the reasoning we do (implicitly, to ourselves). The first part goes something like this:
Someone who says that it is okay to be white must be saying something they believe to be true. But it is so obviously true—especially in our society—that being white is acceptable that it is trivial: it is neither surprising nor informative. So if this is all they were saying, they would not be being cooperative, as they would be violating Quantity. So they must be trying to communicate something else — maybe even something they can’t say explicitly in polite society.
This part of the reasoning is what makes us think there is implication here at all, because it would be pointless for someone to say this and mean by it only what it literally says. But what could it be that they are trying to communicate. The second part of the reasoning follows the line of thought about scalar implicature just above:
A claim that is logically stronger than ‘It’s okay to be white’ is ‘It’s okay to be any colour at all’. (If the latter is true, the former must be too.) Someone who says ‘It’s okay to be white’, if they are being cooperative, must not think they can truthfully utter this stronger claim. So if they are being cooperative, they must not accept the stronger claim. So they are committed to accepting that it is okay to be white, but not that it is okay to be any colour at all. So they must accept that it is only okay to be white. And so that is what they communicate by their statement.
This helps us address the concerns of our Twitter users. When One Nation senators endorse the claim ‘It’s okay to be white’, they are saying one thing, and communicating something else, by implication. They are implying that it is only okay to be white. This is undoubtedly racist. What the dissenting MPs objected to was not the literal meaning, but this additional implied message.
However — and here’s partly where the confusion comes in — it is rather tricky in English to object only to the implications of a statement. When someone assert something which normally would implicate something, but which does not, we say that the implication has been cancelled. We saw an example earlier, where ‘Some students got an A — in fact, they all did!’ involves saying something which doesn’t implicate that only some students got an A, even though the first part of the utterance would typically have that implication. And if Evelyn says ‘some students got an A’, I can accept her claim while targeting the implication if I follow up by saying ‘In fact they all did’ — in which case my utterance cancels the implicature.
But what is clear is that I cannot cancel the implicature simply by saying ‘No!’ or ‘That’s not true’. If I followed up Evelyn’s speech with either of those statements, I would be taken (rightly) to be denying that any student got an A, rejecting the original statement and its implications. (Some cases are even harder to cancel, such as presuppositions. Presuppositions of a sentence are those implications which are retained when the sentence is embedded in some larger construction, like denial. So denying the original sentence in fact leads to further affirmation of the presupposition. Both ‘Jill has stopped running’ and ‘Jill hasn’t stopped running’ imply ‘Jill once ran’. If I think Jill is an inveterate couch potato, who never ran, and I wish to deny the presupposition, I need to engage in some more sophisticated linguistic manoeuvre than merely denying the original claim that she has stopped running. I need to say something like ‘Jill hasn’t stopped running — she never started!’, using stress to indicate that I’m targeting the presupposition-trigger ‘stop’.)
Since it is trivial and everyone agrees that white is an okay colour to be, no one can target the implicature by just saying ‘It’s not okay to be white’. This explains why those who objected to the motion have resorted to more roundabout means of denial. A good example is furnished by Labor MP Anne Aly, who responded to the posting of the slogan at her electorate office by saying ‘It’s okay to be you, whoever you are; but this racial intimidation and harassment is not okay’:
Words matter. When our leaders use white supremacist and neo-Nazi slogans in Parliament, and when the Government votes…
Ordinary English doesn’t permit her to deny the implication in any more straightforward way. She has to first affirm that it is okay to be anything you like (including white), and then register her objection to racial intimidation, part of which is of course the racist implicature of the sign she is commenting on.
I raised earlier the question of why parallel sentences are not similarly controversial. If the Gricean account is correct, then why doesn’t saying ‘it’s okay to be different’ also have a problematic implication—in this case, that it is only okay to be different, that only non-conformity is okay?
The answer, sadly, lies in the fact that it is not a merely uninformative triviality to be told that it is okay not to conform to societal expectations. Remember that our search for what is communicated by ‘it’s okay to be white’ had two parts: the first part which prompted us to search for something non-literal that is meant by the sentence, the second suggesting what that something might be.
But in any society, difference is by definition not normal or typical. There is only a very small gap, in ordinary use, between the descriptive and pejorative senses of ‘abnormal’. Looking around at how difficult it is to be noticeably different from the typical case, one can reasonably conclude that there is a structural bias in favour of standard and normal situations and ways of living. Enough people think that we all should conform to social norms that it is an informative contribution to a conversation to try and remind everyone that it is after all acceptable to diverge from those norms. Given that, it cannot be a mere trivial platitude to assert ‘it’s okay to be different’ — and hence it is entirely plausible that what someone might mean by saying ‘it’s okay to be different’ is just its literal meaning. The same cannot be said for ‘it’s okay to be normal’. This is a platitude in its literal content, and anyone saying it will prompt their hearers to search for what non-literal thing they might have meant in saying it.
A similar explanation is involved in arguing why saying ‘Black lives matter’ doesn’t imply ‘Only black lives matter’, but rather simply entails the addition of its literal content to what’s accepted in the conversation already. In effect, it simply reminds us that black lives matter too.