The Harm in Hate Speech 
by Jeremy Waldron.
Harvard, 292 pp., £19.95, June 2012, 978 0 674 06589 5
Show More
Show More

‘You are a totalitarian asshole.’ It’s probably not the sort of email that often drops into an All Souls professor’s inbox but, as Jeremy Waldron tells us, some people take the doctrine of free speech literally, and cut up rough when they think it has been slighted. All the same, one assumes that the sender of the email would defend to the death Waldron’s right to question the doctrine, or even (though Waldron goes nothing like this far) to reject it. Free speech is nothing if not a platform for totalitarian assholes.

The sender’s rage was sparked by Waldron’s questioning the US consensus on free speech, whose core text is the clause of the First Amendment that enjoins Congress not to make any law abridging the freedom of speech or the press. Waldron wants to exempt hate speech from its scope, though it’s important to distinguish between prohibiting a form of speech and merely failing to grant it constitutional protection. Waldron’s arguments are about withdrawing protection, not prohibition.

As usually interpreted, the amendment applies broadly, particularly as regards what counts as ‘speech’. The Supreme Court’s Citizens United ruling of 2010 lifted a previous ban on the corporate funding of political advertising – for example, in support of candidates in election campaigns. In Ashcroft v. Free Speech Coalition (2002), the Ninth Circuit held that free-speech protection should extend to child pornography created either by virtual means or by using adult actors presented as children. In Hudnut (1985) it was held on appeal that an Indianapolis statute which aimed at outlawing material depicting ‘the graphic sexually explicit subordination of women’ violated the amendment.

However, some ‘fighting words’ have been judged not to benefit from constitutional protection, as in the Supreme Court verdict in Chaplinsky (1942). This upheld a Jehovah’s Witness’s conviction for describing a New Hampshire marshal, who had tried to stop him preaching, as a ‘racketeer’ and ‘damned fascist’. The court defined ‘fighting words’ as those ‘that by their very utterance inflict injury or tend to incite an immediate breach of the peace’. Subsequent rulings by the court have diluted this judgment, as justices seek to pick out a line between allowing injurious speech to flourish unchecked and repressing the robust exchange of opinions. In Snyder v. Phelps (2011), it ruled that the bereaved family of a marine killed in Iraq, whose funeral was picketed by bigots from Westboro Baptist Church chanting ‘fag troops’ and ‘thank God for dead soldiers’, could not recover damages for tortious infliction of emotional distress. The chanters regarded the death of the marine, Matthew Snyder, as God’s punishment for the state’s tolerance of homosexuality (Snyder himself was not gay); the court found that their chants were entitled to special protection. One judge, Samuel Alito, dissented, denying that the First Amendment includes ‘a licence for the vicious verbal assault that occurred in this case’.

So we arrive at the seemingly simple question that Waldron’s book addresses: should a commitment to protecting speech include hate speech? The competing positions are neatly summarised by the 1966 International Covenant on Civil and Political Rights. Article 19 states that ‘everyone shall have the right to freedom of expression,’ while Article 20 provides that ‘any propaganda for war shall be prohibited by law’ (which you might think applies to, say, the Blair government’s case for war in 2002-3), along with ‘any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence’. The tension between the two articles exposes the oddity of the liberal consensus on free speech, which tries to merge two hard-to-merge thoughts: first, that because access to speech matters, people should be free to say what they want, and stifling speech inflicts harm; second, that words can be freely aired because they are far less harmful in their effects than other kinds of action (this might be called the ‘only words’ doctrine). But why should banning speech cause harm, whereas verbalising doesn’t, even though an obvious reason for wanting to gag people is the fear that not doing so may cause harm? Clearly, the prospect of such harm can be exploited to squelch opinions one doesn’t care to hear. But the fact that some people cite harm disingenuously doesn’t show that the preponderance of harm arises from a regime of gagging rather than talking (proponents of free speech can be disingenuous too, as with the UK newspaper industry’s pieties during the Leveson inquiry).

The idea that the only bad thing that can come of speech is its suppression sits best with a view of language-use as blurting. Talk matters mainly, on this view, not to get something across, but to get it out – a sort of psychic expectoration. This is no accident, since early writers about free religious expression were principally concerned with ensuring that dispatches from the individual conscience could get aired. In the nascent free-speech doctrine, the interactive aspect of communication, other than with the Almighty, played a secondary role. The nub of this doctrine is ‘To thine own self be true,’ where integrity – saying what one conscientiously thinks, or at least not saying what one doesn’t think – is central, just as it is in John Rawls’s theory of justice of 1971.

The claim that access to speech matters doesn’t flatly contradict the ‘only words’ doctrine (both are endorsed in Mill’s On Liberty). But a supporting narrative needs to be added if they are to cohabit peaceably, one showing that only stifling speech, rather than facilitating it, can cause harm, or at least that the harm of free speech is of a lower order than that of stifling it. (Mill exempted speech serving as incitement.) It’s hard to see why speech can matter only in a good way: it can do so only if the one relevant aspect of speech is the act itself, rather than its effects – which, like those of any other act, may be good or bad. The freedom of ‘free’ speech then rests on the claim that it is costless, in that (whether or not it is commodified, as with internet paywalls) its expression is harmless. But speech often causes harm, as with the reputational damage done to Western clothes firms by the exposure of sweatshop labour practices in their supply chains. These are surely justifiable harms caused by speech.

There is, however, no obvious reason why all harms caused by speech must be justifiable. Waldron duly rejects the only words doctrine, on the plausible ground that some uses of speech – ‘hate speech’ – do indeed inflict serious harm without countervailing justification. His concern is with hate speech as currently defined in law, and not with all speech – such as the email with which I began – that might be thought to express hatred. As he argues, speech can systematically degrade people. If it can matter in a bad way too, then the only words doctrine is less than self-evident, and the justification for treating speech differently from other acts looks shaky: we’re then into a standard cost-benefit analysis, with harms and gains on both sides. It is here that the liberal consensus is vulnerable to political pressure.

Few people – Waldron distinguishes US from European attitudes – really think that the state should simply let it all rip. One of the useful things the state can do is to stop people from shutting others up by killing or threatening to kill those who speak out. Any state-free or state-lite vision has to say something credible about what happens in such cases; here, as elsewhere, libertarians are apt to take the cyclopean view that the only threats to liberty come from the state. In fact, some threats to liberty come from speech itself, as with incitement or the panoply of libel threats used by press barons to silence speech they don’t care for. (Robert Maxwell hounded Private Eye for years in the libel courts, as did Tiny Rowland and, more recently, the Barclay Brothers, owners of the Telegraph.) Sometimes, too, self-conscious adherence to principle leads people to act as provocateurs, as in the controversy over the publication by the Danish newspaper Jyllands-Posten of cartoons of Muhammad – a brouhaha, as Waldron says, seemingly cooked up by the editor for the very purpose of testing Muslims’ acceptance of free speech.

It seems that more people believe they believe in an unqualified norm of free speech than really do so. This calls for not one explanation but two: why people think they believe in it, and why when it comes down to it, they usually don’t. On the first point, free speech has sacred-cow status, and it doesn’t do to look – or to be seen to look – a sacred cow in the udder. Waldron mentions the hoary pseudo-Voltairean line about disagreeing with what a man says but defending to the death his right to say it – which, as Waldron says, was in fact a gloss by the American writer Beatrice Hall on Voltaire’s attitude to book-burning, rather than a defence by Voltaire of hate speech. Waldron is mainly concerned with access to speech, rather than with its content, but even the content of speech is routinely restricted – the second reason fewer people believe in it than think they do. You’re unlikely to read an article in this journal or anywhere in the mainstream press arguing that the Holocaust never happened, or that paedophilia is sadly underrated. But such views are freely traded on the internet, so perhaps the question should be not what is merely permissible, but what is given prominence. Some platforms carry much more power than others. One would have to search hard to find critical coverage of Israel in the US mainstream media; not many op-ed columns in the UK national press advocate socialist revolution. Often it’s those who most vigorously defend liberal positions elsewhere who most actively police acceptable speech at home – especially when it comes to perceived political bias in the media.

Free speech isn’t just a matter of content but also of context. Speech that is innocuous in one setting may be incendiary in another. It can also make a difference what speech gets aired when, by whom, and who else will hear it. Certain kinds of misrepresentation are actionable, as with perjury, making false statements on tax returns and libel. Or speech may be obliged to respect confidentiality clauses or the demands of ‘official secrecy’. A lot of the time, opinion is silenced on the grounds of irrelevance, or how it is expressed, that it is out of place or comes from the wrong person. Waldron cites Wendell Holmes’s example of someone who shouts ‘Fire’ in a crowded theatre; but more striking is the ban on shouting anything at all in theatres, at least during a performance. Maybe I’d be tolerated if I stood at the cheese counter in Tesco loudly reciting Beowulf, but I wouldn’t if I did it from the public gallery at a trial, or in a church or synagogue, during a seminar or the performance of a play. A woman who sang on an American Airlines plane was ejected, though not during the flight.

Waldron admirably tries to keep in mind how speech regimes work in practice. It’s not just a matter of philosophical principle, of what constitutional texts and case law say, but of looking at real speech acts under a given juridical regime. A conspicuous virtue of Waldron’s work generally has been his ability to imagine – as in his classic paper on homelessness, where he attends to the way the day-to-day lives of homeless people preclude settled routines for eating, excreting, rest and sleep – the relation of lived experience to questions of theory. Here, in an arresting chapter, he imagines a world in which hate speech against a certain minority is commonplace on billboards and on the airwaves. There have been times and places where this hasn’t been far from the truth, as in the Nazi demonisation of the Jews, or the genocidal incitement against the Tutsi population of Rwanda by Radio Télévision Libre des Mille Collines before the massacres of 1994. Hate speech, practised on this scale, reduces its victims to the condition of outcasts and a state of perpetual fear. Even when it does not result in murder, its effect may be to silence certain views or certain voices. Something like the targeted silencing of the population has happened to immigrants in recent years in the UK, where the denigration comes mainly from newspapers: to downgrade people’s status is also to deprive them of a voice. Rae Langton has made a similar point about the effect of pornography in limiting the speech acts that women can perform.

It is sometimes said that the free market in ideas, of which free speech is an instance, will weed out at least the grosser forms of hate speech. There is no guarantee of this. Markets work as devices for co-ordinating information about commodities, particularly their price. Speech is in some ways a peculiar commodity, in that – unlike a Scotch egg, say – the fact that one person has consumed it doesn’t mean that another can’t. In this sense speech could be thought of as ‘free’. But that doesn’t mean its production lacks opportunity costs, or that it can’t be commodified by such devices as paywalls: it can, as the enduring domination of academic publishing by large companies like Elsevier and Springer shows. Here as elsewhere, market power can act as cause and effect of cartelisation and other restrictive practices. In the US, TV programmes critical of Israel have been kiboshed by advertisers’ threats to pull funding. What outputs markets produce, in speech as in other commodities, depends on decisions by those with market power. The ‘free market in ideas’ reduces to a platitude: speech will approximate to a perfectly competitive market only if the state or other agencies succeed in regulating it so as to produce perfect competition. The notion that it will do so of its own accord is a fantasy entertained, sometimes sincerely, by the powerful.

Waldron’s argument that hate speech waged via billboards and TV ads can inflict serious harm on its targets is hard to dispute. Such invective, he says, is an assault on the victims’ dignity. But the notion of dignity is in some ways problematic. Suppose one asks if it has empirical content. If it does, then on any plausible view dignity is a quality, like natural talent, which can be distributed unequally between persons. I assume Waldron does not want his notion of dignity to have this feature. Dignity may also lead to puzzles: Waldron mentions the ban by the mayor of a French town of the sport of dwarf-tossing, which was upheld by the Conseil d’Etat on the grounds that the sport affronted human dignity, but challenged by a professional tossee on the grounds that the ban violated his dignity. There’s also the danger that vilification can prove self-validating: people can lose dignity, for example, through persecution devised for that purpose. If, on the other hand, claims about dignity are really normative claims about the equal treatment of persons, it seems that the idea of equality is doing the real work. Waldron, indeed, couples ‘dignity’ with ‘equality’ at several points, and rests his case on the ‘high status’, whether we call it ‘dignity’ or not, that ‘we accord every person in social and legal interactions’: he talks of ‘the dignity of equal citizenship’. He also uses a distinction drawn by Stephen Darwall between ‘appraisal’ respect, which can legitimately vary from person to person, and ‘recognition’ respect, which is paid at a flat rate to all. Recognition respect is paid even to those, say, who are suspected of having committed atrocities, in that we recognise their right to due process. Waldron says that recognition respect is ‘invariant’, which I assume means not only that it can’t vary, for one person, from one time to another, but also that its level cannot vary from one person to another. If that’s so, it’s a form of equality.

If, as Waldron argues, First Amendment protection should be denied to hate speech, how far should the denial be extended? The recent campaign on Twitter against Caroline Criado-Perez, bombarded with rape threats for campaigning for women’s images on banknotes, seems a prime example of hate speech. While Waldron’s main concern is with speech that expresses or incites racial hatred, he does discuss the case of pornography. He notes Catharine MacKinnon’s argument that pornography in its dominant form – namely, the depiction of women to gratify heterosexual men – constitutes a form of hate speech directed at women, and comments: ‘It seems to me that women are entitled to ask whether official legal tolerance of pornography and of its pervasive public display is consistent with our commitment to the dignity and equality of women.’ He doesn’t go so far as to ask whether porn should get the same treatment as hate speech and, if it should, what that would mean for top-shelf and billboard regulation. But if anything, MacKinnon’s case gains in plausibility once the visual ambience created by pornography is taken into account. After all, while we do not live in a world in which black people could be routinely traduced on hoardings, the same hardly goes for women, especially if one takes into account the pervasive use of porno-lite images to sell things in magazines, on TV, the internet and billboards. It doesn’t matter much if the advertisers or their clients aren’t trying to degrade women. It still comes out that way.

Either pornography is speech, or it isn’t. Suppose it is, as pornographers believe. They have long made the self-serving claim that commercial porn should benefit from First Amendment protection, on the grounds that it counts as speech, or at least as a form of expression – for instance, because it purports to be ‘art’. Some people, the late Ronald Dworkin included, have taken this claim far more seriously than it deserves; Dworkin even argued that there is a constitutional right to publish pornography. Before photo-reprographic technology developed, most porn was textual, but since the modern porn industry deals overwhelmingly in still and moving images, it is not clear, to say the least, that speech is involved. If porn is speech, however, an obvious question arises: why can’t it be conveyed in verbal rather than pictorial form, thereby meeting pornographers’ hunger for self-expression? The usual move here is to say that the form of expression is somehow integral to the content. For instance, when Paul Robert Cohen was arraigned for wearing a jacket embroidered with the slogan ‘Fuck the draft’ during the Vietnam War, it was the words used that were deemed objectionable; Cohen would have had no trouble if the jacket had said ‘I find conscription morally repugnant.’ If someone then says that this is a sorry travesty of the original, the same will go with added force for attempts to reduce pictorial content to text, and by those lights the identification of porn with speech, and pornographers’ claim to the special protection supposed due to it, become yet more doubtful.

But if, despite all this, porn is speech, then a good case can be made for regarding a lot of it – material that portrays women as legitimate objects of violence or degradation – as hate speech. Again, Waldron does not pursue this argument. I’ve focused on the case of porn because in some ways it seems no less eligible than racial hate speech for withdrawal of protection – at least when one pays attention, as he does, to real discursive conditions, where what matters isn’t just what can be said, but what view is dominant. So if porn is speech, it’s liable to the exemptions Waldron envisages for hate speech, and if it isn’t, it is ineligible for the protections pornographers claim under the First Amendment.

But the censorship question goes further than this. Appealing to equality as the grounds for not protecting hate speech threatens to open up a crevasse. People talk abusively about differences in race, religion, language, sexual orientation, gender, body size, age and so on. Sometimes they express hatred based on these traits and the groups that they define. Waldron doesn’t mention many of these categories, but he does consider religious hate speech, and how it might be limited in the interests of tolerance. It’s sometimes said that tolerating intolerance such as hate speech is incoherent. It isn’t, but hate speech does prompt the question whether it better serves the ends of tolerance to let the haters get on with their intolerance, or to be intolerant towards them. Since these options amount to permitting intolerance or being intolerant, abstract appeal to tolerance leads to an impasse. Equality thus threatens (to use a metaphor common in First Amendment case law) to have a ‘chilling’ effect on free speech.

Here paths divide, depending on how directive Waldron wants his argument to be. Perhaps he means it to work at the level of principle, not legal text, and aims simply to make the case for exempting hate speech from constitutional protection without trying to finesse hard cases. Otherwise one might try – as some do with ‘laundry list’ approaches to obscenity – to enumerate the intent, content and context that constitute hate speech. The problems with this approach are well known, but are not so different in kind from well-established legal concepts, like mens rea and reasonable belief, which rely on ascriptive judgments.

It’s often said that the problem is one of line-drawing. The no holds barred take on speech is nothing if not dazzlingly clear. By contrast, advocates of speech restrictions are continually accused of arbitrariness and of creating slippery slopes. But it isn’t clear that hate speech presents more of a problem than other issues – the law draws arbitrary lines all the time. Another objection is that qualifying First Amendment protection will yield ever greater encroachments on free speech. More subtly, it may be said that free speech norms are only arbitrarily related to the legal regime that claims to express them or, at least, that there are always bits that don’t fit. For instance, some pragmatists foreground truth-seeking as the aim of speech, and argue that speech must be free to get to the truth. But a lot of speech doesn’t even aim at the truth – take bullshitting, as defined by Harry Frankfurt, where the bullshitter tries not to say what’s true, but only to fill in discursive space. Is there a constitutionally protected right to bullshit? Speech regimes habitually display two failures of fit: between principles and legal (e.g. constitutional) norms; and between those norms and what speech is, in fact, protected. Examples of failures of fit are usually dealt with by saying they ‘don’t count’. Another instance is advertising standards: firms can’t say that users of their suppositories will gain instant sex appeal, or (in the UK) use subliminal advertising. Principles suggest a direction of travel, but there are always cases bearing out the timeless truth that speech is free, except when it isn’t.

In general, Waldron takes a kid-gloves approach to the First Amendment, perhaps cowed by the vehemence of its more dogmatic proponents. He takes care to distinguish hate speech from the mere causing of offence which, even if intended, concerns only ‘subjective aspects of feeling’ rather than ‘a person’s standing in society’. His main concern is with speech that aims to provoke hatred rather than to express it (of course, it may aim to do both), as this is what damages dignity. A familiar fork opens up: either there is an empirical fact about the damage done, where culpability will vary sharply with context (the damage will usually be greater if many people are saying similar things); or one has to gauge intent, with wide room for disagreement as to what the speaker had in mind. Some flexibility is also needed in handling different acts of hate: compare routine racial bigotry with the hatred felt by victims of sexual abuse for their abusers. This raises a question as to whether the law could use the uncomfortable idea of justifiable hatred in deciding what speech should and should not be protected. Which manifestations of hatred will be considered off-limits?

Certainly one can distinguish appraisal respect from recognition respect, to contrast derogatory but acceptable speech with invective that denies its victims’ right to recognition. But that distinction is available to hate-mongers as well as to bien-pensants: they can say that they don’t hate this person, they just hate his blackness, her gender etc. Such claims are flimsy and usually disingenuous, but how do they differ from abominating someone, say Anders Breivik, as a mass murderer, while accepting that he has inalienable rights even after conviction? One answer is to make the identity move: blackness, femaleness and so on are constitutive of what makes somebody this person, so that the target of hate speech can’t simply be peeled apart from their identity. But Waldron expresses unease about identity claims, since they lead people to claim more protection ‘for their interests and opinions than they are entitled to’. What merits protection instead is an abstract ideal of dignity to which everyone is equally entitled.

This throws up a puzzle. For Waldron, racist hate speech attacks human dignity, and that’s why it’s not entitled to legal protection. He’s well aware of the danger to free speech from removing protection from talk that people may find offensive (but which fails to affront their dignity). The puzzle isn’t how one tells affronts to dignity from mere offensiveness, though that poses obvious problems. It’s that the blackness or Jewishness of hate speech’s victims has at once to be incidental to their civic identity, since respect is paid at a flat rate to all, and Waldron wants to avoid a censors’ auction where victims bid in the currency of offence; but it has also to be sufficiently part of their identity for the speech to undermine the basis of respect for them. It isn’t clear that anything answers to this dual specification. Somebody might insist that if legal equality really is there, that’s all that’s needed. That of course sounds glib, but if the grounds for that judgment aren’t based on the subjective offence speech can cause, they seem to demand a more far-reaching critique of how legal equality coexists with the various disabilities, such as poverty, that undermine people’s civic understanding.

The strain shown by Waldron’s argument here stems from his concern to safeguard protection for speech, and the strong presumption that it be free. But that presumption is not indefeasible for all and every form of speech. It is strongest with decisions whose political legitimacy is in doubt – for instance, attempts to stifle speech itself. But even this has its limits: Waldron convincingly dispatches Dworkin’s strange view that any limit on hate speech must delegitimate other political decisions, so that (say) racially aggravated assault cannot legitimately be criminalised if the polity that outlaws it has already banned hate speech. A more sensible view is that sometimes speech has to be curbed so that everyone can have a proper say.

There’s a passing resemblance between the otherwise largely distinct US lobbies behind the First Amendment and the Second (enshrining the right to bear arms). First, both amendments are treated as articles of faith, indeed dogma, by many of their proponents, as a way of sidelining objections and, in fact, politics itself. Second, both lobbies voice a blunt libertarianism which takes the state as the only obstacle to freedom, ignoring both its potential role as a facilitator of freedom and the role of many private agencies, such as corporations, as freedom’s opponent. Third, both lobbies often respond to their opponents’ objections by arguing that all that’s needed is more of the same: more speech to counter the misuse of speech, more guns to counter the misuse of guns. In all this the political power of principle – even as the lobbies claim to be sidelining politics – is plain, and often not pretty.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 36 No. 1 · 9 January 2014

Glen Newey’s injection of a negative reference to ‘press barons’ into his review of Jeremy Waldron’s book on hate speech seems puzzling (LRB, 5 December 2013). The barons he cites as hounders of Private Eye – Maxwell, Rowland and the Barclays – were notoriously litigious long before they acquired national newspapers. Actually, as far as Private Eye is concerned, Rowland was best known for offering to bankroll the Eye’s efforts to fend off James Goldsmith, whose name is missing from Newey’s list despite his being the Eye’s most vindictive adversary. Of course, over the decades, the Eye’s preferred baronial target has been Rupert Murdoch, who is famous for never suing, however inaccurate Eye stories might be.

David Elstein
London SW15

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences