'Speech which spreads hate should never be tolerated.' Discuss.

Essay

Although government ought to place some legal restrictions on hate speech because of the harms it causes, not all such hateful speech should be banned. Toleration of propositionally expressive hate speech, at both the legal and social level, is important for the ability of citizens to participate in democratic discussion, and ultimately for social welfare. In this essay, I first provide a working characterisation of what it means for speech to spread hate, and clarify what is meant by “toleration”. I then present the consequentialist case for restrictions on hateful speech, identifying the empirical premisses which must be met for the argument to hold. I then demonstrate that some restrictions on hateful speech should be endorsed even by those who believe that free speech is intrinsically valuable, or that it is a means to ends other than those of utility (for example, autonomy). Finally, I conclude that whilst the fact that some particular speech spreads hate is a pro tanto reason for its prohibition, this interest in direct harm prevention may sometimes be outweighed by the indirect harm done by intolerance of others’ speech.

Defining hate speech is an important first step in any discussion of what our policy towards it should be. As Anderson and Barnes (2022) note, whilst hate speech could be identified in terms of the effects it has on the targets, this results in a definition which is vague and under-specified. Moreover, using this as the basis to defend prohibitions on hate speech seems to put the cart before the horse, in that it assumes there unquestionably are great harms being done by this speech. Therefore, I follow the approach taken by Parekh (2012, 40), centred around the content of the speech in question: hate speech is that which “expresses, encourages, stirs up, or incites hatred

Parekh clarifies what he means: “Hatred is not the same as lack of respect or even positive disrespect, dislike, disapproval, or a demeaning view of others. It implies hostility, ill will, severe contempt, rejection, a wish to harm or destroy the target group, a silent or vocal and a passive or active declaration of war against it.”

against a group of individuals… based on an arbitrary and normatively irrelevant

In reality, I find this very unsatisfying as a definition. Parekh claims that “[t]o say that one hates murderers and wants them all locked up or executed is not a case of hate speech”, but presumably somewhere in substituting “murderers” for “mixed-race people” you cross over into the territory that he’s marked out as hate speech? Deciding what’s normatively relevant is a contentious matter – just look at all the arguments around whether biological sex matters. (cf Simpson and Srinivasan on no-platforming and “disciplinary expertise”). If we lived in a society where everybody could agree on what was of normative relevance, there’d be much less reason to worry about how to balance the harms of hate with the risks of regulation.

feature”. Merely articulating one’s mistrust or dislike of such a group would not qualify as hate speech.

Toleration, in this context, refers to the absence of legal interdictions and social penalties against certain speech. These forms of prohibition and intolerance are closely related, and it would be a mistake to draw a false dichotomy between the two: whilst the magnitude of costs facing an individual who decides to break the law may be significantly larger than those borne by one who decides to break with social convention, the qualitative effect is the same (Mill 1859, 63). Social tolerance does not require that we sit mute whilst others spout what we believe to be bigoted falsehoods, but it does entail

that we refrain from launching ad hominem attacks or seeking to make someone’s life more difficult simply because they aired certain views. (Ibid, 144). Of course, identifying whether tolerance is permissible (or required) in the case of hate speech is the goal of this essay.

For those with a utilitarian moral framework, free speech and tolerance are both only valuable insofar as they promote the general welfare. We can therefore construct a consequentialist case for prohibitions on hate speech as follows:

P1: Hate speech does harm to those it is directed at.

P2: Ceasing to tolerate hate speech would reduce its incidence.

P3: Any costs to aggregate welfare from ceasing to tolerate hate speech are outweighed by benefits from reducing the incidence of such speech.

C: Society should cease to tolerate hate speech.

Few would reject P1. As well as the direct psychological harms inflicted on the targets of hate, such speech also has pernicious longer-term effects, creating conditions in which groups of society are more likely to be ostracised, discriminated against, and subjected to violence (Parekh 2012, 44-45). P2 and P3 are more contentious, though ultimately empirical, premisses. Opponents of restrictions on hate speech marshal a variety of arguments against P2, though perhaps the two most common objections are (a) that intolerance of hate speech in the public sphere would only cause it to instead fester underground, and (b) that the most effective way to root out hate is through debate and discussion (Baker 2009, 155). The case for (a) looks especially weak: if hate speech alone occurs underground, then it inflicts no harm on targets; equally, if this speech leads to discrimination and violence, then the government should additionally act to limit and punish this further law-breaking (Anderson & Barnes 2022). As for (b), the following series of events illustrates one instance where further debate and discussion cannot be an effective way of reducing the incidence of hate:

Stephen launches a hateful series of attacks against Ruby, and everyone else with red skin. Ruby experiences psychological harm, and her attempts to calmly rebut the vilification simply lead Stephen to engage in more diatribes. As a result, Ruby and other reds withdraw from the public sphere, in the hope that Stephen is less likely to attack them.

Just as government intervention in some sections of the economy is necessary to prevent market failure and collapse

e.g., see the famous Market for Lemons paper.

, so too is regulation of the marketplace of ideas. We can formalise this by appealing to the concept of speech acts. In the context of pornography, the speech of pornographers limits women’s perlocutionary and illocutionary power, by making it more likely that men ignore them when they say “no”, and by circulating the notion that ‘“no” really means “yes”’, respectively (Langton 1993). More generally, hate speech inhibits others’ ability to enter, and be heard in, the public square. This is directly analogous to how, in our definition of tolerance, we recognised that overbearing social penalties reduce individuals’ ability for free speech.

Once the issue of hate speech is viewed from this angle, the outline of a rights-based justification for intolerance of hate emerges. For those who believe that free speech is not simply a means to welfare, the consequentialist argument will not hold – it is not enough to present on a balance-sheet the costs and benefits if free speech is viewed as something of special worth (Dworkin 2009, vii). Yet Langton’s argument cuts through this, by showing that hate speech poses a threat to speech more widely, and not just to the distinct goods of equality and dignity. Indeed, even if one believed that liberty of expression had lexical priority over any other intrinsic good, one would have to concede

I’m unsure whether this is really true if you have some kind of deontic morality. “Balancing” of interests doesn’t really work there? Maybe you think there’s a rule that the government isn’t allowed to interfere with any free speech, even if doing so would increase the level of free speech overall (i.e. you reject the idea of negative responsibility).

that there is at least a potential case for restrictions on speech that undermines others’ free speech.¹

Returning to the consequentialist argument for hate speech restrictions, what about the soundness of P3? A free speech absolutist may attempt to undermine the claim that intolerance of hate is beneficial overall by pointing to the risk of stumbling down a slippery slope of censorship. Sufficiently narrow framing of the scope of laws

And social rules, though this is much harder to enforce since there’s much, much less explicit coordination around it.

against hate speech might mitigate against this risk: we already have bans on libel and defamation, but not against criticism or petty insults (Anderson & Barnes 2022). However, that seems to suggest that perhaps some hate ought to be tolerated, as part of the price for avoiding an undesirable chilling effect on legitimate speech. Although it certainly seems naïve to think that more speech, no matter how hateful, is always ultimately better, that is a far lower bar than a proponent of the titular statement must clear – their task is to show that, if it includes any content spreading hate, tolerating

This word makes things clunkier, but I think it is (possibly) doing some important work. Someone might conceivably claim that there is some hateful speech that it would be better to have than not have, but also that it is better not to tolerate it. Put differently, maybe legal and social rules about what is allowed come apart from what is morally right, for instrumental reasons of maximising good across the board.

more speech is never better. Let us now examine more closely the plausibility of that claim.

There are some kinds of speech which spread hate by propagating negative claims about groups of individuals, but do not explicitly vilify the targets – for instance, statements like “All reds are dirty and uneducated criminals!”. Against the view that speech which spreads hate should never be tolerated, I suggest that we should, in fact, tolerate speech of this nature, because it inflicts only limited harms on the targets, and does not adversely affect their ability to exercise freedom of speech. The distinguishing feature of this permissible speech is that its content is primarily propositional. That is, it communicates objects of belief which are either true or false. Whilst the targets of such speech will no doubt still be harmed, the extent of this is greatly reduced by the fact that they are able to challenge and refute the propositions in question with rational argument. As Yong (2011) puts it, “[b]ecause assertions of fact and evaluative opinions have clear cognitive content, they can appropriately be answered through deliberative and articulate speech”.

This holds whether one adopts a consequentialist or rights-based position in assessing the value of free speech. From a rights-based perspective, it seems unlikely that this speech significantly impairs

I think this is actually dubious. If people believe that you’re a “dirty and uneducated criminal”, they’ll very probably stop listening what you’re saying.

the targets’ right to exercise their own speech, or that their interest in being treated with dignity as equal members of society is adversely affected, when considered relative to the claim that speakers have to speak. Equally, taking a consequentialist standpoint, the damage done by undermining the principle of free debate would be greater than that which would be avoided by banning this form of speech. Again, there is a clear parallel with our understanding of tolerance, grounded in Mill’s view of liberty: people must be able to robustly criticise their intellectual opponents, but may not use rhetoric to inflict harm on someone because they are an intellectual opponent; similarly, people must be able to robustly criticise other groups of citizens, but may not use rhetoric to inflict harm on someone because they are part of a certain group of citizens. Identifying the motivations behind an individual’s speech is impractical, but tolerating only hate-spreading speech whose content is primarily propositional is an eminently workable operationalisation of this principle.

So, to conclude, it is not true that speech which spreads hate should never be tolerated. The fact that some speech expresses or causes hate provides a *pro tanto*

On prima facie vs pro tanto reasons, Shelly Kagan says: “A pro tanto reason has genuine weight, but nonetheless may be outweighed by other considerations. Thus, calling a reason a pro tanto reason is to be distinguished from calling it a prima facie reason, which I take to involve an epistemological qualification: a prima facie reason appears to be a reason, but may actually not be a reason at all.”

– but not indefeasible – reason for it to be regulated. This is because although hate is harmful (through both its direct and indirect effects on victims, which include a reduction in their ability to exercise free speech), so is intolerance of hate speech. I have argued that, on consequentialist and rights-based justifications for free speech, there are strong reasons for hateful speech that is mostly non-propositional to be regulated, but for speech which indirectly spreads hate via the expression of propositional statements to be tolerated. Such an approach strikes the right balance between society’s interests in the prevention of harm and protection of liberty.

References

Anderson, L., & Barnes, M. (2023). Hate Speech. In E. N. Zalta & U. Nodelman (eds.), The Stanford Encyclopedia of Philosophy (Fall 2023 Edition). https://plato.stanford.edu/archives/fall2023/entries/hate-speech.

Baker, C. E. (2009). Autonomy and hate speech. In I. Hare & J. Weinstein (eds.), Extreme Speech and Democracy (pp. 139–157). https://doi.org/10.1093/acprof:oso/9780199548781.003.0009.

Dworkin, R. (2009). Foreword. In I. Hare & J. Weinstein (eds.), Extreme Speech and Democracy (pp. v–x). https://doi.org/10.1093/acprof:oso/9780199548781.002.0004.

Langton, R. (1993). Speech Acts and Unspeakable Acts. Philosophy & Public Affairs, 22(4), 293–330. http://www.jstor.org/stable/2265469.

Parekh, B. (2012). Is there a case for banning hate speech? In M. Herz and P. Molnar (eds.), The Content and Context of Hate Speech (pp. 37–56). https://doi.org/10.1017/cbo9781139042871.006.

Mill, J. S. (1859). On Liberty (ed. Himmelfarb, G., 1985). Penguin.

Yong, C. (2011). Does freedom of speech include hate speech? Res Publica, 17(4), 385–403. https://doi.org/10.1007/s11158-011-9158-y.

I am careful to state there is only a “potential” case. It could be true that restrictions would infringe upon the liberty interests of the speaker more than they would protect the liberty interests of the target – but this would an empirical matter to be resolved with a specific problem at hand. ↩︎

‘Speech which spreads hate should never be tolerated.’ Discuss.

Essay

References