The essays to which we are responding take the long and rather well-worn debate about hate speech in new directions. Weinstein’s central claim, which Waldron rejects, is that hate speech laws can undermine the legitimacy of the legal system as a whole and of particular “downstream” laws, such as laws prohibiting racial discrimination.
I found three features of this debate especially eye-catching. The first is Weinstein’s reliance on the concept of viewpoint discrimination. As a comparativist of freedom of speech, it is always striking to see an aspect of First Amendment law invoked as a core principle of freedom of speech, given that in its aversion to viewpoint discrimination (like much else) First Amendment law is highly unusual. In other countries – like Canada and Australia – viewpoint discrimination is relevant to the determining the law’s validity but does not carry anything like the weight it does in First Amendment law.
Second, it is worth noting the apparently confined nature of the disagreement. Weinstein accepts Waldron’s position that much of what qualifies as hate speech is of no value to public debate because in stable and mature democracies the commitment to elemental matters of racial and sexual equality and religious tolerance is not the subject of “serious or considerable contestation.”
Third, the debate, apparently at least, puts aside two often dominant questions: whether hate speech causes harm (the harm question) and whether the power to impose or enforce hate speech law will inevitably be used to protect government interests, favour the powerful and disadvantage the vulnerable (the abuse of power question). Weinstein accepts that hate speech causes certain harms and aims his arguments, for the most part at least, at the narrow kind of law that Waldron wishes to defend. (Laws, like section 18(1) of the Public Order Act (UK) that are directed at highly vituperative kinds of bigoted speech).
Leaving these questions to one side sharpens this debate by avoiding the need to engage with the resolve the messy questions of fact. It also raises the stakes: Weinstein defence purports to apply even though the speech protected is worthless and even if a law is well targeted and competently administered law. It apparently yields a very strong defence of freedom of speech.
However, I am not convinced that the question of legitimacy of hate speech laws can be resolved in this way. Viewpoint discrimination is not so obviously inimical to equal political participation as Weinstein suggests. On the contrary, a failure to enact hate speech laws may, consistently with Weinstein’s account, undermine the legitimacy of the legal system generally and the application of individual laws to the victims of hate speech. Thus the argument from political legitimacy illustrates how the values underlying speech can be wielded both for and against the protection of speech, a phenomenon I have elsewhere described as the “double-sided” nature of freedom of speech.
Precisely for these reasons, as I have also argued elsewhere, the messy realities to which the harm question and the abuse of power question direct us cannot be easily avoided. Nor can Weinstein so neatly avoid questions as to the nature and worth of hateful speech. On this point, Weinstein’s position is more complex than it first appears. Although he apparently accepts that much hate speech is of no value in public debate, he also believes that some of the expression caught by hate speech laws – notably speech motivated by religious belief – is worthy of protection precisely because it debates matters that are the subject of reasonable disagreement. If true, hate speech is a more variable and complex phenomenon than the argument otherwise admits. Although for the most part it appears that Weinstein defends hate speech even when—as in the case of speech denying basic racial and gender equality—there can be no reasonable case for its worth. However, there are points—in relation to religiously motived speech—at which the argument appears to rely on reasonable disagreement about the worth of speech to make the case for its protection.
- Viewpoint Discrimination and Political Legitimacy: The Argument
Weinstein’s reliance on the concept of viewpoint discrimination arises in response to Waldron’s point that hate speech laws that are targeted only at the most vituperative kind of hate speech have limited effect. They limit speech that is widely understood to be worthless and leave open other ways for the “propositional content” of this hateful speech to be expressed.
Weinstein’s reply is that because hate speech laws are viewpoint-based they “uniquely implicate the fundamental interest in governing as a political equal of those whose speech is suppressed by the restriction.” At another point, he expands:
[V]iewpoint-based speech restrictions on public discourse infringe the fundamental interest in equal political participation of those who want to express these forbidden views. To the extent that such censorship prevents people from expressing what they believe is best for society, it is insulting; in so far the speech restriction impairs their ability to promote or protect their own self-interest, it is also fundamentally unfair.
In giving this justification of the rule against viewpoint discrimination Weinstein takes a somewhat novel turn. More typically, the rule against viewpoint discrimination has been justified on grounds that viewpoint discrimination distorts public debate; interferes with citizen’s capacity property to inform themselves; and, because state regulation on grounds of viewpoint poses the greatest risk that it will attempt to reinforce its own interests or to reflect the lawmaker’s mistaken view of the truth or value of a message.
- Viewpoint Discrimination and Political Legitimacy: The Reply
Although novel, this justification for the rule against viewpoint discrimination runs into several objections. The most fundamental is that hate speech laws further, rather than undermine, political legitimacy. Alexander Brown, who deals with this argument in most detail, argues that hate speech laws ensure that people are accorded dignity in public debate, or at least that the state uses measures it has available to it to provide them with an assurance that their dignity will be respected. Consequently, a failure to enact hate speech laws may itself impose a cost on political legitimacy.
Weinstein’s response to this relies in part on the difference between specific and systemic legitimacy. The diminution in legitimacy that Brown identifies is “systemic.” It is the legitimacy of “the legal system” that has been diminished, not the obligation to obey or the morality of enforcement of a particular law. It is this latter kind of illegitimacy with which Weinstein is principally concerned. He objects to the state forcing “dissenters” to comply with a law with which they disagree, in circumstances where their capacity to express their dissenting ideas is circumscribed by hate speech laws. Weinstein concludes that this form of “specific” illegitimacy is not commensurate with, and therefore not offset by, the systemic illegitimacy of which Brown writes.
Waldron and Brown have both offered responses to this step in the argument. Waldron doubts that illegitimacy can, as a general matter, be attributed to specific “downstream” laws as Weinstein supposes. Brown argues that fundamental principles of justice justify the application of “downstream” laws to individuals who transgress them, even if those individuals have been prevented from articulating hateful and discriminatory views.
Without doubting the first two responses, a third may be available. If the supposed diminution in legitimacy by hate speech laws can attach to specific laws in their application to individuals, why could the reverse not be true? That is, why could not the failure the enact hate speech laws also undermine the legitimacy of specific “downstream” laws and thus undermine the justification for applying those laws to specific individuals?
To make this point, we need to revisit the question of the harm hate speech causes and to consider directly the question of whether hate speech “silences.” If we take seriously that claim, it might interfere with the capacity of its victims to be equal citizens in much the same way as Weinstein claims that hate speech laws operate. If hate speech victims are silenced, it seems that hate speech undermines the capacity of its victims to oppose laws that have specific application to them. Indeed, Weinstein acknowledges this much positing the instance of an indigenous population unable, because of their vilification, effectively to argue in public debate for accommodations of their religious traditions. Katharine Gelber provides another illustration arguing that same-sex attracted people may be unable fully to participate in public debate in opposition to laws prohibiting same sex marriage, yet those laws apply to deprive them of the right to marry. The cases reveal an equally strong case argument that it is unfair and insulting to subject the victims of hate speech to laws on which they have not had a fair opportunity to express their views.
- Revisiting the Harm Question and the Abuse Question
Among the most obvious responses to the arguments I have just made are as follows: First it might be doubted that hate speech “silences” in the way I have suggested. Second, a distinction might be drawn between silencing by the state’s act and silencing where the state fails to address the cumulative effect of the speech of others.
A full treatment of these objections requires more by way of analysis than I can give them here. Moreover, not all the questions they raise are susceptible of empirical investigation. The question of whether hate speech can “silence” is in part a question about the nature of speech and its relationship to action. The question of whether a state failure to prevent silencing is equivalent to state action raises difficult questions about the nature of rights. (Freedom of speech is usually – but not universally – conceived of as a negative and vertical right, which entails that it is exercisable only against the state and acts as a shield against interference rather than as an entitlement to some form of state response.)
But in addition, questions of fact are deeply implicated in both inquiries. Rae Langton concludes her famous essay on the silencing effect of pornography by noting unresolved empirical questions. On her argument, the power of speech to silence depends in part of the authority of the speaker in the relevant domain. As she concludes:
These are not really questions to be settled from the philosopher’s armchair. To answer them one needs to know about the role pornographers occupy as authoritative speakers about the facts, or supposed facts, of sex.
The same can be said about “silencing” on the other side of the ledger. The extent to which hate speakers are truly “silenced” by a hate speech law will depend upon many features of the law targeting hateful speech including the degree to which speakers can express their views in other ways as well as the mechanisms for its enforcement, the remedies or penalties that it imposes.
Similarly, although not entirely an empirical question, the question of whether the silencing caused by hate speech should be addressed by law in part depends questions about the propensity of governments (including prosecutorial authorities and courts) to overstep their powers, selectively enforce laws, or otherwise misuse power in a way that undermines or frustrate any legitimate role for hate speech laws.
- Reasonable Disagreement?
Weinstein calls for more attention to these empirical questions (especially the question of silencing) but he also concludes that in the absence of evidence there is good reason to favour freedom of speech. First, he says, the silencing of “hate speech” laws is certain whereas the silencing effect of hate speech is merely putative. More fundamentally, he insists that we treat speakers in this situation equally, even in the face of evidence as to silencing:
[E]ven if it could be definitively shown that bigoted speech prevented others from participating in public discourse, it is not clear what principle would justify shutting up A (or a group of As) so that B (or a group of Bs) can speak. Or put in terms of political legitimacy, why should A’s sense of alienation and lack of identification with the state be of lesser concern than B’s?
In this respect Weinstein position is in tension with other aspects of the argument. First, Weinstein writes as if there is neutral position in which no party is silenced. But if the silencing effect is real, as he is prepared to assume, then we have a choice between either silencing the hate speaker or the victim. (The question he poses could equally be reversed: “why should B’s sense of alienation and lack of identification with the state be of lesser concern than A’s?”) If he is truly prepared to accept the claims of silencing, there is no way of avoiding the dilemma. If Weinstein’s view is that silencing by the state is more problematic than silencing by virtue of private action, then that position needs to be justified.
Second, the argument for treating all hate speakers and their victims equally appears to belie his first position as to moral worth of hate speech. Weinstein agrees that elemental commitments to racial and sexual equality and religious tolerance are beyond contest. However, as the argument develops it appears that his central concern is that individuals not be denied the opportunity to express their views on matters about which there can be reasonable disagreement. It is this reason that the illustrations given in the second part of the paper all involve questions speech critical of homosexuality and anti-Islamic speech, matters on which there is “a close moral question, pitting fundamental liberty interests . . . against each other.” This aspect of Weinstein’s argument appears to narrow his claim in ways that are not clearly acknowledged. If the objection from legitimacy applies only where speech on matters of reasonable disagreement is prohibited, then we need an account from Weinstein as to where reasonable disagreement arises and why. Moreover, the argument should acknowledge a realm (where reasonable disagreement is absent) in which it does not apply.
Finally, it is no answer to these arguments to reassert that political legitimacy requires that all citizens are able to participate in public debate regardless of their viewpoint. For one thing, it does not resolve the dilemma. Members of affected groups are likely to be the most motivated, insightful and effective advocates for their points of view and therefore hate speech can reasonably be supposed to silence in a viewpoint discriminatory way. Indeed, the viewpoint discriminatory effect of this silencing may be magnified in comparison with hate speech laws. If hate speech truly “silences” its victims, it does not preserve alternative ways for those victims to express the “propositional content” of their views.
But in any event, the question we have been considering is why viewpoint discrimination should be though to undermine political legitimacy. I hope to have shown, along with others, that enforcing viewpoint neutrality by precluding hate speech laws could have costs to legitimacy equal to and as specific in their application as the costs Weinstein attributes to viewpoint discriminatory hate speech laws. If this is true, there is no reason, in the absence of evidence on silencing, to err on the side of freedom of speech unless Weinstein contradicts his earlier acceptance (for the purposes of argument) of the silencing effect.
- Conclusion: The Double-Sided Nature of Freedom of Speech
This debate about the legitimacy of hate speech laws has, in the end, a familiar rhythm. It is a constant feature of debates about freedom of speech that the values that apparently underlie it can be wielded both for and against the right. It is, for instance, especially evident in the comparative scholarly dialogue about the protection of hate speech in Canadian law and under the First Amendment. Typically, the decisions of Canadian Supreme Court upholding hate speech laws are explained as revealing fundamental and distinctive Canadian constitutional values of “equality” and “multiculturalism” by discouraging discrimination and racial violence; counteracting silencing; and protecting cultural identity and an individual’s capacity to take part in the communal life of cultural groups. I have elsewhere explored how First Amendment inspired accounts of freedom of speech rely on equality and multicultural diversity as values for protecting speech, and in particular for protecting hate speech.
This debate on legitimacy and hate speech shows us yet another way in which the values underlying freedom of speech can militate for and against freedom of speech. Moreover, it demonstrates that satisfactory resolution of the argument requires close engagement with matters of fact. It seems unlikely that the harm question and the abuse of power question can be entirely avoided. Nor can we avoid distinguishing more carefully amongst kinds of hate speech by identifying where (if at all) it expresses reasonable disagreement and where it does not. For all its illumination, this debate should encourage philosophers and legal scholars to engage seriously and systematically with the fine detail of what hate speech is, how it is regulated, how those laws operate in practice and the evidence as to what hate speech does to its victims.
* Redmond Barry Distinguished Professor, Kathleen Fitzpatrick Australian Laureate Fellow, Director of the Centre for Comparative Constitutional Studies, Melbourne Law School.
. For a discussion of the First Amendment in the comparative context, see Frederick Schauer, The Exceptional First Amendment, in American Exceptionalism and Human Rights 29 (Michael Ignatieff ed., 2005); Adrienne Stone, The Comparative Constitutional Law of Freedom of Expression, in Research Handbook on Comparative Constitutional Law 406 (Rosalind Dixon & Tom Ginsburg eds., 2011).
. In Canadian law, if a law has the purpose of limiting expression it is, by virtue of that purpose alone, taken to infringe section 2(b) of the Canadian Charter of Rights and Freedoms. The need for a factual inquiry into the effects of the law is thus obviated. However, this finding does not invalidate the law, rather the Court proceeds to apply section 1 of the Charter to determine whether the law is reasonably necessary in a free and democratic society. See Irwin Toy v. Quebec,  1 S.C.R. 972. In Australian law, the High Court of Australia has adopted a distinction between a “direct” and “incidental” burden on freedom of expression that appears at least partially to track the distinction between content-based and content-neutral laws. Hogan v Hinch,  243 CLR 506, para 95. While this classification may lead to a stricter form of review, it does not amount to a presumption against invalidity. On the contrary, a flexible proportionality-style test remains the dominant approach. McCloy v N.S.W.,  257 CLR 158; see also Adrienne Stone, The Limits of Constitutional Text and Structure, 23 Melbourne U. L. Rev. 668 (1999).
. Adrienne Stone, The Canadian Constitutional Law of Freedom of Expression, in Canada in the World: Comparative Perspectives on the Canadian Constitution (Richard Albert & David R. Cameron eds., Cambridge University Press) (forthcoming 2017).
. This rule is an aspect of a more general prohibition in First Amendment law on “content-based” laws regulating speech. Within the general rule, a law involving viewpoint discrimination is regarded as the most problematic form of a content-based law and therefore even expression within “low value categories” is protected from viewpoint discrimination. See R.A.V. v. City of St. Paul, 505 U.S. 377 (1992).
. One question aired, which I will put aside, is whether it is correct to treat laws such as Waldron defends as viewpoint based (given that they are not targeted at viewpoints per se but at the harmful consequences of hateful speech).
. Alexander Brown, Hate Speech Laws, Legitimacy, and Precaution: A Reply to James Weinstein, 32 Const. Comment. 599 (2017). For other discussion in this Symposium, see Jeremy Waldron, The Conditions of Legitimacy: A Reply to James Weinstein, 32 Const. Comment. 697 (2017) and Katharine Gelber, Hate Speech—Definitions & Empirical Evidence, 32 Const. Comment. 619 (2017).
. See Jennifer Hornsby, Speech Acts and Pornography, in The Problem of Pornography (Susan Dwyer ed., 1995); Rae Langton, Speech Acts and Unspeakable Acts, 22 Phil. & Pub. Affs. 293 (1992). On the question of empirical support for this claim, see Gelber, supra note 12; Katharine Gelber & Luke McNamara, Evidencing the Harms of Hate Speech, 22 Soc. Identities 324 (2016).
. Id. at 312. (questioning whether pornography “is authoritative for those hearers in the domain of speech about sex . . . people, men, boys, who . . . want to know which moves in the sexual game are legitimate.”).
. I have elsewhere suggested that positions on this question, in part, depend on assumptions and commitments that are not empirically derived in any systematic way but constructions of a constitutional culture. See Stone, infra note 23.
. Id. at 539 (“[W]here the morality of a law cannot just be reasonably questioned but where its moral status is both contestable and highly contentious the lack of an opportunity to participate can have grave consequences for political legitimacy.”).
. See Stone, supra note 4; see also Adrienne Stone, How To Think about the Problem of Hate Speech, in Freedom of Speech and Hate Speech in Australia (Katharine Gelber & Adrienne Stone eds., 2007).
. Weinstein’s discussion of hate speech laws in a number of European jurisdictions is highly informative, but it principally illustrates how many hate speech laws go much further than the narrow form of hate speech law that Waldron would defend. Additionally, Katharine Gelber points to its limitations as a systematic study of hate speech laws in operation.