Democratic legitimacy, political speech and viewpoint neutrality

The purpose of this article is to consider the question of whether democratic legitimacy requires viewpoint neutrality with regard to political speech – including extremist political speech, such as hate speech. The starting point of my discussion is Jeremy Waldron’s negative answer to this question. He argues that it is permissible for liberal democracies to ban certain extremist viewpoints – such as vituperative hate speech – because such viewpoint-based restrictions protect the dignity of persons and a social and moral environment of mutual respect. According to Waldron, well-drafted narrow hate speech bans are not democratically illegitimate, and they do not undermine systemic democratic legitimacy – that is, the legitimacy of a democratic political system. In contrast to Waldron, I will argue that democratic legitimacy requires viewpoint neutrality to respect persons as thinking agents. I will defend a civil libertarian doctrine of viewpoint neutrality, and this doctrine requires that citizens in liberal democracies ought to have a legal free speech right to do moral wrong – that is, a legal right to express and defend any political viewpoint or idea, even if it is morally wrong to express, or expose others to, such views. It will be argued that any viewpoint-based restriction on public discourse (including narrow hate speech bans) is democratically illegitimate, and that such restrictions undermine systemic democratic legitimacy.


Introduction
In contemporary liberal political philosophy, it is a widely held assumption that democratic legitimacy requires that all citizens have a right to freedom of political speech. However, there is deep disagreement about the issue of whether democratic legitimacy requires viewpoint neutrality with regard to political speech -that is, a right to express, hear and consider any political viewpoint, idea or doctrine within public discourse. This disagreement is evident in contemporary debates about extremist political speech, such as hate speech and advocacy of terrorism.
The purpose of this article is to consider the question of whether democratic legitimacy requires viewpoint neutrality with regard to political speech. The starting point of my discussion is Jeremy Waldron's negative answer to this question. He argues that it is permissible for liberal democracies to ban certain extremist viewpoints -such as vituperative hate speech -because such viewpoint-based restrictions protect the civic dignity of persons and a social and moral environment of mutual respect. According to Waldron, well-drafted narrow hate speech bans are not democratically illegitimate, and they do not undermine systemic democratic legitimacy -that is, the legitimacy of a democratic political system or a democratic constitution. 1 In contrast to Waldron, I will argue that democratic legitimacy requires viewpoint neutrality to respect the status of persons as thinking agents. 2 I will defend what can be called a civil libertarian doctrine of viewpoint neutrality -a doctrine Waldron and other defenders of hate speech laws reject. According to this doctrine, citizens in a liberal democracy -including unpopular extremist dissenters -have a basic right to participate in public discourse as speakers and listeners free from state-imposed viewpoint-based restrictions. This civil libertarian doctrine of viewpoint neutrality requires that citizens in liberal democracies ought to have a legal free speech right to do moral wrong -that is, a legal right to express and defend any political viewpoint or idea, even if it is morally wrong to express, or expose others to, such views. In contrast to Waldron, it will be argued that any viewpoint-based restriction on public discourse is democratically illegitimate, and that such viewpoint-selective restrictions undermine systemic democratic legitimacy.
My argument for the outlined position proceeds through three steps. First, I will develop and defend a version of a status-based theory of rights that provides a basis for a specification of the scope and strength of the right to participate in public discourse that supports the civil libertarian doctrine of viewpoint neutrality. According to this theory, respect for the status of persons as thinking agents and their sovereignty over their own mind requires viewpoint neutrality -that is, a basic right to express, hear and consider any viewpoint within public discourse. The idea of a thinking agent has two main aspects: (1) Thinking agents are adult persons who possess certain respect-warranting capacities, such as the deliberative capacity to make up their own mind in matters of politics and faith, as well as the ability to choose how they will live their own separate lives in accordance with their own understanding of what is a meaningful and valuable life. 3 (2) A thinking agent is a person who is sovereign over his/her own mind -that is, a thinking agent should have authority over his/her own beliefs and values. This second aspect is similar to Thomas Scanlon's idea of an 'autonomous agent' -that is, a person who 'is sovereign in deciding what to believe and in weighing reasons for action '. 4 Second, I will argue that the legitimacy of the democratic exercise of political power requires viewpoint neutrality (and the associated legal free speech right to do moral wrong) to respect the status of persons as thinking agents, under circumstances of deep disagreement. Third, I will argue that viewpoint-based restrictions such as broad and narrow hate speech laws undermine systemic democratic legitimacy, and that the enactment and enforcement of viewpoint-based restrictions fall outside the legitimate jurisdiction of majorities -that is, the scope of the majority's moral right to rule. With regard to this step of my argument, it is important to make the following clarification. I will argue that the illegitimacy of a hate speech ban -in an otherwise legitimate liberal democracy -does not remove the state or the majority's power-right to enact, and authorize the enforcement of, other laws against persons who have been prevented from publicly expressing their hateful viewpoints. In contrast to Ronald Dworkin, 5 I do not assume that hate speakers have a right to immunity from such state coercion. However, I will argue that hate speakers and their potential audience have a basic right to immunity from state-imposed viewpoint-based restrictions on public discourse, and that this means that such restrictions should be repealed through a system of judicial review.
This article proceeds as follows. In section 2, I will explain what I mean by political speech and give an account of the doctrine of viewpoint neutrality. Here, I will also give an account of different kinds of restrictions on speech and explain why hate speech laws (including the narrow bans Waldron proposes) are viewpoint-based restrictions. Section 3 sets out the understanding of democratic legitimacy that provides the starting point of my discussion. The aim of section 4 is to present Waldron's position and his most important argument for this position. In section 5, I will present in more detail the civil libertarian version of the doctrine of viewpoint neutrality that I will defend. In section 6, I will set out my case for the civil libertarian doctrine of viewpoint neutrality and the position that democratic legitimacy requires viewpoint neutrality. More precisely, it will be argued that viewpoint neutrality is a rights-based requirement of democratic legitimacy grounded in respect for the status of persons as thinking agents and their sovereignty over their own beliefs and values.

Political speech, viewpoint neutrality and hate speech laws
My point of departure is a broad understanding of political speech. Political speech will refer to all acts of expression that take place within public discourse -that is, deliberation on matters 'concerning the organization and culture of society'. 6 It is speech or other expressive conduct that is relevant to both intrapersonal and interpersonal deliberation 7 on issues concerning the organization and culture of society or matters of public concern. This will include religious or ideological views and ideas that are relevant to public discourse or political deliberation -for example, advocacy of holy war or Jihad.
The doctrine of viewpoint neutrality requires that all persons have a right to express, hear and consider any viewpoint, idea or doctrine within public discourse. This means that liberal democracies should impose no criminal or civil penalties upon the expression of political opinions or ideas. The doctrine of viewpoint neutrality plays an important role in US constitutional law, 8 and it requires that citizens in a liberal democracy should have a right to participate in public discourse as speakers and listeners free from stateimposed viewpoint-based restrictions.
To explain what this means, it can be useful to distinguish between three kinds of restrictions on speech. Content-neutral restrictions limit communication for reasons that are unrelated to the content or message of the expression -such as restrictions on time, place or manner of the exercise of speech (e.g. political demonstrations). Content-based restrictions limit communication because of the content or message of the expression. However, all content-based restrictions are not viewpoint-based. For example, a ban on all political speech in a certain place (e.g. airports or railway stations) is content-based but viewpoint neutral.
Viewpoint-based restrictions are a subset of content-based restrictions that restrict the communication of particular ideas or viewpoints. Such restrictions differentiate and discriminate between different political or ideological viewpoints. Typically, viewpoint-based restrictions are viewpoint discriminatory in the sense that they grant freedom to express and hear the state-approved viewpoint in public discourse but prohibit some forms of expression of the contrary viewpoint. 9 In a political context, viewpoint-based restrictions suppress at least some forms of expression of the viewpoints of one side in a political debate or disagreement. Justice Anthony Kennedy of the US Supreme Court describes the doctrine of viewpoint neutrality and its prohibition of viewpoint discrimination like this: At its most basic, the test for viewpoint discrimination is whether -within the relevant subject category -the government has singled out a subset of messages for disfavor based on the views expressed. . . . The First Amendment's viewpoint neutrality principle protects more than the right to identify with a particular side. It protects the right to create and present arguments for particular positions in particular ways, as the speaker chooses. 10 Viewpoint-based restrictions are often designed to restrict speech because the viewpoint it expresses is regarded as dangerous or poisonous -that is, because of a fear of the consequences of the reception of particular ideas or viewpoints. This can be a fear of the dangerous or poisonous pervasive effect or force of a viewpoint or a fear of the effect of a viewpoint or idea on the social and moral environment of the community.
Hate speech laws are examples of viewpoint-based restrictions. Although the content of hate speech laws varies from state to state, 11 the common denominator of most hate speech laws is that they in different ways ban speech which attacks others on grounds such as race, ethnicity, religion or sexuality. This also applies to the hate speech laws that Waldron defends.
Waldron's idea of dignity -the dignity of persons 12 -plays an important role in his understanding of the aim of hate speech and the aim of hate speech laws. On the one hand, the aim of hate speech is to attack the dignity of persons. Waldron regards hate speech as a form of group defamation -that is, an attack on group reputation. 13 The viewpoints expressed in hate speech are a 'poison' that attacks the dignity of persons and the reputation of minority groups, and this poison undermines or 'pollutes' a social and moral environment of mutual respect that is essential to sustain the conditions of 'a healthy working democracy'. 14 On the other hand, the aim of hate speech laws is to protect the civic dignity of persons (i.e. the legal, political and social status or standing of persons) and a social and moral environment of mutual respect. According to Waldron, well-drafted hate speech laws are narrow hate speech bans that are not solely based on viewpoints but also based on the manner in which a viewpoint is expressed, and the intended or foreseeable (i.e. likely) effect of the expression. Well-drafted hate speech laws prohibit viewpoints that attack the dignity of others on the grounds of their race, ethnicity, religion or sexuality -if and only if these viewpoints are expressed in a vituperative manner or language, and the speech is intended or foreseeable to have the effect of stirring up hatred and contempt against certain groups. 15 In a recent article, Waldron seems to question whether such well-drafted narrow hate speech bans are viewpoint-based restrictions at all. 16 Given the outlined account of the doctrine of viewpoint neutrality and viewpoint-based restrictions, there are, at least, two related reasons for rejecting this view. First, although these narrow bans only target or single out vituperative hate speech, they still prohibit viewpoints, such as extreme racist or islamophobic viewpoints expressed in a vituperative manner or language. 17 To determine whether a law is viewpoint-based or viewpoint neutral 'it is the speech that a law prohibits, not that which it leaves unregulated, that is relevant'. 18 Second, narrow bans on vituperative hate speech are viewpoint discriminatory, in the sense that they differentiate and discriminate between (a) messages and viewpoints that attack others on the grounds of their race, ethnicity and religion and (b) other sorts of messages and viewpoints. For example, they discriminate between racist viewpoints and non-racist viewpoints. Even narrow hate speech laws are, as Eric Heinze aptly points out, viewpoint-based restrictions, because they grant freedom to express, hear and consider the state-approved viewpoint in public discourse, while penalizing at least some forms of expression of a competing viewpoint. 'Hate speech bans . . . impose viewpoint-selective penalties on citizens who, within public discourse, express certain forms of hostility towards the state-proclaimed pluralist values expressed by the bans'. 19

Democratic legitimacy
In political and legal theory, it is a widely held assumption that political legitimacy (in the normative sense) is about the moral right to rule and the moral basis or justification for this right. The moral right to rule is a right to exercise political power -that is, the power to make and apply laws, as well as to use, or authorize the use of, coercion to enforce laws. Thus, one can say that political legitimacy concerns the moral right of A to rule or exercise political power over B in some context C and the moral justification for A's right to rule. The moral right to rule is a complex right that contains several Hohfeldian rights, 20 such as A's power-right to alter the legal rights and duties of B within certain limits. Various normative theories of political legitimacy differ on what conditions or requirements must be met in order for A to have the right to rule over B, and they differ on what the reasons are for such conditions or requirements. 21 In a democracy, A primarily refers to the people and their elected representatives, while B refers to citizens subject to A's rule. In a democratic state, political power belongs to citizens, in the sense that political power is citizens' power to impose laws and policies on one another under circumstances of disagreement. Each adult citizen has a right to run for office and vote in free elections that directly or indirectly determine what laws will be enacted and enforced and that select the public officials in top government posts. This means that the people (typically the majority of citizens and their elected representatives) exercise significant political power over the lives of those subject to their rule.
One can distinguish between two related aspects of democratic legitimacy that play an important role in the debate about whether viewpoint neutrality is a requirement of democratic legitimacy (or political legitimacy in a democracy). The first concerns the legitimate democratic exercise of political power -that is, the rightful exercise of political power by the people (or the majority of voters) and their elected representatives, who have been authorized by the people (or the majority) to act on their behalf. This is a matter of 'ruling legitimately'. A normative theory of democratic legitimacy provides a set of requirements that should apply to the legitimate democratic exercise of political power. These requirements have, at least, two related functions. First, they specify how democratic decisions should be made. As Waldron points out, '[i]t matters not just what laws we have but how they were enacted -by whose votes and under what conditions of deliberation'. 22 Second, they provide standards for assessing the legitimacy of democratic institutions and the decisions made within them as well as standards for evaluating the legitimacy of decision-makers (i.e. who should be the rightful source of authority or who should have the right to rule).
The second aspect of democratic legitimacy is about the legitimate jurisdiction of majorities -that is, the scope of the majority's moral right to rule in a liberal democracy. The idea is that the majority's moral power-right to alter the legal rights and duties of those subject to its rule has limits, and these limits define or specify the majority's legitimate jurisdiction. This is a matter of the scope of 'legitimate rule' (i.e. the scope of a state's or a majority's legitimate or rightful power). Strictly speaking, this second aspect of democratic legitimacy is contained in the first. A majority of the people and their elected representatives do not exercise political power legitimately if the majority makes decisions on issues that fall outside its legitimate jurisdiction. According to Arthur Isak Applbaum, an important question concerning the legitimate jurisdiction of majorities is what sorts of issues ought to be on and off the democratic table. 23 As we will see in sections 4 to 6, there is deep disagreement about the scope of the majority's legitimate jurisdiction. One important area of disagreement in the debate about democratic legitimacy and the permissibility of viewpoint-based restrictions is precisely about the question of whether majorities have a moral power-right to enact, and authorize the enforcement of, viewpoint-punitive laws.

Waldron on democratic legitimacy and freedom of political speech
In this section, I will first present Waldron's position on democratic legitimacy and permissible viewpoint-based restrictions. Thereafter, I will set out the core ideas of Waldron's argument for certain viewpoint-based restrictions on public discourse -that is, his defence of hate speech laws.

Waldron on democratic legitimacy and permissible viewpoint-based restrictions
According to Waldron's theory of democracy, democratic legitimacy requires that democratic institutions should show equal respect for 'persons as moral agents and moral reasoners' and their judgements. 24 In view of fundamental disagreement about which collectively binding decisions should be made in a democracy, equal respect for persons and their judgements provides a basis for, at least, three standards of democratic legitimacy. First, equal respect for persons requires political equality -that is, 'equal voice and equal decisional authority'. 25 Each person has a fundamental right to political participation on equal terms. This right is required for the constitution of a democracy. It is constitutive of the democratic process.
Democracy requires that when there is disagreement in a society about a matter on which a common decision is needed, every man and every woman in the society has the right to participate on equal terms in the resolution of that disagreement. . . . [T]here cannot be a democracy unless the right to participate is upheld, and unless the complex rules of the representative political process are governed, fundamentally, by that right. 26 Second, collectively binding decisions in a democracy should be made through majority rule, because majority rule is a respectful decision-procedure that shows equal respect for persons who disagree about what decision is right. In the first place, majority rule respects and takes seriously the views that individuals have arrived at after deliberation and their differences of opinion about justice, rights and the common good. 27 Secondly, majority rule accords equal weight to each person's vote and views. At the stage of decision-making, every individual is given equal power to affect the decision outcome. Waldron sums up the 'fairness/equality defence of majority rule' like this: 'Better than any other rule, MD [i.e. majority rule] is neutral as between the contested outcomes, treats participants equally, and gives each expressed opinion the greatest weight possible compatible with giving equal weight to all opinions'. 28 In the face of disagreement, Waldron assumes that 'majority-decision is the only decision-procedure consistent with equal respect', 29 but he doubts that one can prove this.
Third, equal respect for persons and their judgements and the right to political participation on equal terms require that everything is up for grabs in a democracy. More precisely, 'everything is up for grabs which is the subject of good-faith disagreement'including the rights of democracy itself. 30 From Waldron's normative ideas of equal respect for persons, political equality and the right to political participation, it follows that the legitimate democratic exercise of political power requires that all citizens should have equal political rights to participate in and influence political processes and decisions. In a liberal democracy, citizens should, among other things, be granted the right to vote and the right to freedom of political speech.
Although Waldron thinks that democratic legitimacy requires freedom of political speech, he rejects the doctrine of viewpoint neutrality (as this is outlined in sections 2 and 5). He assumes that it is permissible for liberal democracies to enact and enforce well-drafted narrow hate speech laws -that is, dignity-protecting laws that (at least) prohibit vituperative hate speech intended to stir up hatred and contempt against members of vulnerable minorities. Waldron claims that the effect on systemic (and nonsystemic) democratic legitimacy of well-drafted narrow hate speech bans is 'minimal or non-existent'. In other words, the loss to democratic legitimacy from such hate speech bans is insignificant. His defence of this claim can be called the 'bending over backward' argument, and it goes like this: [M]ost such laws [i.e. well-drafted narrow hate speech laws] bend over backward to ensure that there is a lawful way of expressing something like the propositional content of views that become objectionable when expressed as vituperation. They try to define a legitimate mode of roughly equivalent expression, a sort of safe haven for the moderate expression of the gist of the view whose hateful or hate-inciting expression is prohibited. 31 Waldron's bending over backward argument rests on what can be called 'the manner-matter distinction' (or the form-substance distinction), and one's assessment of this argument primarily depends on whether one accepts this distinction. Waldron's idea seems to be twofold. First, one can distinguish or separate between (a) the substance or matter of a speaker's viewpoint or thought and (b) the manner or form in which the speaker chooses to communicate his/her viewpoint or thought (e.g. 'how hatefully . . .views are expressed' 32 ). Second, narrow hate speech laws do not prevent hate speakers from communicating the substance of their viewpoints or thoughts, but only prevent hate speakers from communicating their viewpoints or ideas in a vituperative or extreme manner. Hateful viewpoints formulated in a moderate or civilized manner are left unrestricted.
There are good reasons for rejecting the manner-matter distinction. The main reason for this is that the manner in which a viewpoint or thought is communicated (e.g. formulated or phrased) is typically inseparably linked to the substance of the speaker's viewpoint or thought. As Peter Jones points out, the problem with the distinction is that it supposes that statements or viewpoints are capable of more or less extreme formulations (e.g. more or less hateful formulations) which are nevertheless identical in meaning. 33 Manner and meaning are not wholly separable in this way. Typically, manner and matter are so integrally related that it is impossible to distinguish the hateful, disrespectful or contemptuous manner from the hateful, disrespectful or contemptuous matter of a statement. At this point, it is also important to note that regardless of what one thinks about Waldron's manner-matter distinction, the narrow hate speech laws that he defends are still viewpoint-based restrictions, because they discriminate between different viewpoints in the sense explained in section 2.
Even if one accepts this rejection of the manner-matter distinction, one can still claim that the systemic delegitimizing effect of narrow hate speech laws is minimal or nonexistent. One can, as Waldron does, argue that the delegitimizing effect of narrow bans is minimal or non-existent because their effects on the opportunities of hate speakers to participate in public discourse are insignificant, since they merely ban the expression of extreme hateful viewpoints (i.e. vituperative hate speech), while moderate hateful viewpoints are left unrestricted. This argument is closely linked to Waldron's balancing approach 34 to justifiable infringements of rights in general and permissible viewpointbased restrictions on extremist speech in particular. The idea is that we need to weigh competing interests against each other and determine the relative strength or importance of the interests at stake. In the case of legal regulation of vituperative hate speech, this means that we should weigh the interests of hate speakers and their potential audience (i.e. the rest of us) against the interests of the targets of hate speech. Since narrow hate speech bans 'bend over backward', Waldron assumes that the interests of hate speakers and their potential audience are of little weight or insignificant compared to the importance or strength of the interests of the targets -especially their interest in being protected from attacks on their dignity or their elementary social reputation and their interest in living in a moral and social environment of mutual respect. 35 Thus, the interests of the targets of hate speech outweigh the interests of hate speakers and their potential audience. I will return to these issues in sections 6.1 and 6.3.
Waldron does not only claim that the prohibition of extreme hateful viewpoints does not undermine systemic democratic legitimacy. In a recent article, he also claims that narrow dignity-protecting bans on vituperative hate speech promote democratic legitimacy. He assumes that hate speech laws aim to secure the conditions of a 'healthy working democracy' and that such bans contribute positively to democratic legitimacy by helping sustain the conditions of democracy. 36 The conditions of a healthy working democracy include social peace, mutual respect and trust among those of different views and different communities. 37 According to Waldron, hate speech that stirs up hatred and contempt undermines these conditions, while '[w]ell-drafted hate speech bans are calculated to maintain social peace and secure dignity and respect among members of the community'. 38

The argument from civic dignity
Waldron's argument for the permissibility of certain viewpoint-based restrictions on public discourse can be called the argument from civic dignity. According to this argument, liberal democratic states and their citizens have a duty to respect the dignity of persons, and certain viewpoint-based restrictions on public discourse, such as narrow hate speech laws, are permissible in liberal democracies because such restrictions protect the dignity of persons and a social and moral environment of mutual respect.
The starting point of Waldron's theory of dignity is that dignity indicates a highranking legal, political and social status or standing. 39 It is a high-ranking status comparable to a rank of nobility. The modern idea of the dignity of persons is a normative idea of equal status -that is, the assignment of a high-ranking legal, political and social status to all persons. 40 The fitting or appropriate response to the dignity of persons is respect. Dignity requires respect for persons -that is, recognition respect. 41 In this connection, Waldron draws on Stephen Darwall who points out that recognition respect means that persons are 'entitled to have other persons take seriously and weigh appropriately the fact that they are persons in deliberating about what to do'. 42 According to Waldron, a crucial dimension of recognition respect is respect for 'the dignity of equal citizenship' for all members of all groups. 43 It is a matter of respect for the dignity of persons 'in the sense of their basic social standing, the basis of their recognition as social equals and as bearers of human rights and constitutional entitlements'. 44 Respect for the dignity of persons provides a basis for certain moral duties that should be legally enforced. First, the dignity of persons provides a basis for moral duties of respect -both public duties and private duties. Liberal democratic states have a public duty to respect the dignity of persons, and citizens in a liberal democracy have a similar private duty. 45 Second, hate speech laws (or group defamation laws) legally enforce the moral duties of respect. Citizens can 'reasonably be required to play their part in society by according the basics of recognition respect to each person'. 46 This is 'the obligation that is being enforced when we enact and administer laws against group libel [i.e. hate speech laws]'. 47 Third, the dignity of persons provides a basis for a public and private duty to protect what Waldron calls 'dignity-based assurance' as a public good. A liberal democracy should provide public and visible assurance to all its citizens that they will be treated in accordance with their dignity and the fundamentals of justice, and all citizens have a duty to contribute to the protection of this dignity-based assurance as a public good. 48 The idea is that in a 'well-ordered' liberal democratic society citizens should have and be able to rely upon the public and visible assurance of one another's commitment to the fundamentals of justice. The fundamentals of justice refer, among other things, to the following normative ideas: All are equally human and have the dignity of humanity; all have an elementary entitlement to justice; and all deserve protection from the most egregious forms of violence, exclusion, indignity and subordination. 49 Hate speech involves the express denial of the fundamentals of justice and the dignity of persons, and such speech undermines dignity-based assurance as a public 'environmental good'. Hate speech laws protect public and visible dignity-based assurance as a public good, and they aim to create and sustain a social and moral environment of mutual respect. According to Waldron, hate speech bans legally enforce certain 'civility norms' that are 'essential for society's ability to guarantee basic dignity'. 50

The civil libertarian doctrine of viewpoint neutrality and the right to do wrong
The doctrine of viewpoint neutrality requires that all persons -including unpopular extremist dissenters -have a right to express, hear and consider any viewpoint, idea or doctrine within public discourse. As John Stuart Mill points out, 'there ought to exist the fullest liberty of professing and discussing, as a matter of ethical conviction, any doctrine, however immoral it may be considered'. 51 With regard to extremist political speech, the civil libertarian doctrine of viewpoint neutrality that I will defend has two important dimensions. First, radical extremist dissenters (e.g. racists or islamophobes) have a right to express and defend their political convictions and ideas, however immoral they may be considered -except in cases where their speech acts violate the basic rights of other persons, such as their basic civil or political rights. Thus, all citizens have a right to participate in public discourse free from viewpoint-based restrictions, even if they express strongly worded and provocative views that attack the reputation of minority or majority groups in a vituperative language intended to stir up hatred or contempt. Second, the rest of us -their potential audience -have a right to listen and make up our own minds.
This civil libertarian doctrine of viewpoint neutrality requires a free speech right to do moral wrong that Waldron and other defenders of hate speech laws reject. The aim of this section is to explain in more detail the way in which this doctrine demands that citizens in a liberal democracy ought to have a legal free speech right to do moral wrong. My argument for this doctrine will be set out in section 6.
A right to do wrong is a moral or a legal right to do something that is wrong from a moral point of view. 52 P has a right to do moral wrong if P has a right to do something that P ought not to do. To put it differently, a right to do moral wrong is a right to do something one has a moral duty not to do -a right to violate one's own duty. 53 A right to do moral wrong is a choice-protecting right -that is, a right the aim of which is to protect the choices or the freedom of choice of the right-holder. Examples of choice-protecting rights are freedom of speech, freedom of association and the right to vote. Choice-protecting rights contain a claim-right that imposes a duty on others and the state not to interfere with the choices of the right-holder -that is, a duty of noninterference. Thus, to say that a person has a choice-protecting right to do moral wrong is to say that other persons and the state have a duty not to interfere with the rightholder's moral wrongdoing. As Ori Herstein points out, a right to do wrong 'is a right against enforcement of duty, that is a right that others not interfere with one's violation of one's own obligations'. 54 The function of a choice-protecting right is not to guide choices but to protect choices. In this connection, there is an interesting and important difference between moral duties and choice-protecting rights. Moral duties provide reasons for action. If P has a duty to do X, this entails that P has a reason to do X. Thus, if we have a moral duty to respect the dignity of persons in Waldron's sense, this provides a reason to do so. If we move to the case of rights, there is no such relation of entailment. Choice-protecting rights do not provide reasons for action. If P has a right to X, this does not entail that P has a reason to do X. Thus, if P has a right to express and defend extremist political viewpoints or ideas, this does not entail that P has a reason to express such views. Although choice-protecting rights do not give the right-holder reasons for action, they give others reasons against interference. The preceding considerations imply that a choice-protecting right to do moral wrong provides a reason for other people and the state not to interfere with the right-holder's moral wrongdoing, but it does not give the right-holder a reason to do moral wrong.
The civil libertarian doctrine of viewpoint neutrality requires that citizens ought to have a legal free speech right to do moral wrong. This right to do moral wrong contains, at least, two important and related elements. First, citizens should have a legally protected liberty to express and defend any political viewpoint or idea within public discourse, even if it is morally wrong to express, or expose others to, such views (e.g. racist or islamophobic views that can cause psychological harms such as inferiority, humiliation, fear or anxiety). Even if it is morally wrong for citizens to express political viewpoints that attack the dignity of persons, or the reputation of minority groups, with the intention to stir up hatred or contempt, the civil libertarian doctrine of viewpoint neutrality demands that they should have a choice-protecting legal right to do moral wrong -subject to the exception clause that their speech acts do not directly and demonstrably violate the basic rights of other persons. This exception clause limits the scope of the free speech right to do moral wrong. A specification of the exception clause demands an answer to the question of the nature and ground of basic rights. I will return to this issue in section 6.1.
Second, citizens ought to have a free speech right to do moral wrong that constitutes a right against legal enforcement of moral duty -that is, a legal claim-right against interference with moral wrongdoing. This means that it is not permissible for majorities in a liberal democracy to enact hate speech laws that legally enforce the moral duty to respect the dignity of persons or the moral duty not to defame ethnic or religious groups, because the state has a duty of non-interference. According to the civil libertarian doctrine of viewpoint neutrality, this is a matter of the legitimate jurisdiction of majorities in a liberal democracy. The enactment and enforcement of viewpoint-based restrictions on public discourse fall outside the scope of the legitimate jurisdiction of majorities -that is, the scope of the majority's moral right to rule.

Viewpoint neutrality as a rights-based requirement of democratic legitimacy
In contrast to Waldron, I will defend the position that liberal democracies should adopt the civil libertarian doctrine of viewpoint neutrality and that democratic legitimacy requires viewpoint neutrality. My argument for this position proceeds through three steps. First, I will develop and defend a version of a status-based theory of rights that provides a basis for a specification of the scope and strength of the right to participate in public discourse that supports the civil libertarian doctrine of viewpoint neutrality. Second, I will argue that the legitimacy of the democratic exercise of political power requires viewpoint neutrality (and the associated legal free speech right to do moral wrong) to respect persons as thinking agents, under circumstances of deep disagreement. Third, I will argue that viewpoint-based restrictions such as broad and narrow hate speech laws undermine systemic democratic legitimacy and that the enactment and enforcement of viewpoint-based restrictions fall outside the legitimate jurisdiction of majorities.

A status-based theory of the right to participate in public discourse
In the liberal tradition of political philosophy, it is a widely held assumption that the dignity of persons requires respect -that is, recognition-respect. This idea plays a central role in Immanuel Kant's moral and political philosophy that is an important source of inspiration for contemporary liberalism. According to Kant, a person possesses a dignity -an absolute inner worth -that commands respect from others. 55 However, there is no general agreement about what respect for persons or respect for the dignity of persons requires. This is evident in discussions about the right to freedom of speech.
Respect for the dignity of persons can be cited on both sides in debates about the right to freedom of expression and its scope and strength. 56 On the one hand, one can, as Waldron does, argue that respect for the dignity of persons requires certain viewpointbased restrictions. On the other hand, one can argue that respect for persons or the dignity of persons requires viewpoint neutrality.
My argument for the civil libertarian doctrine of viewpoint neutrality proceeds from the assumption that political institutions in a liberal democracy should show respect for persons as thinking agents. This is a normative idea that political institutions and procedures should instantiate and respect, and it provides the main justificatory basis of the basic rights of persons, as well as rights-based requirements of political legitimacy in a liberal democracy. The idea is that persons, regarded as thinking agents, have certain attributes or capacities that warrant or command respect, and these respect-warranting characteristics make it fitting to grant them certain basic rights that limit the scope of the majority's moral right to rule.
The status-based theory of basic rights that I defend has three noteworthy aspects. The first is that the ground of basic rights is respect for the status of persons regarded as thinking agents. 57 Basic rights are a response to, or an expression of, (1) the worth and the respect-warranting capacities of persons regarded as thinking agents; (2) the separateness of persons -that is, the morally significant fact that persons are 'distinct individuals each with his own life to lead'; 58 and (3) a person's sovereignty over himself/herself -that is, a person's authority over his/her body, mind (i.e. a person's beliefs and values) and separate life or existence. For present purposes, the most important respect-warranting capacities of persons as essentially thinking agents are as follows.
(i) The deliberative capacity to make up their own minds about 'what is good or bad in life and politics, or what is true or false in matters of justice and faith'. 59 (ii) The ability to express their most fundamental moral, political or religious convictions and defend them against criticism. (iii) The ability to choose their projects and how they will live their lives in accordance with their own understanding of what is a meaningful and valuable life. (iv) The capacity to take responsibility for their choices, the personal and political goals they aim for, and the manner in which they pursue them. These respectwarranting capacities of persons regarded as thinking agents and their sovereignty over their own beliefs and values provide the main reason for why it is fitting or appropriate to grant persons certain basic choice-protecting rights -such as the right to freedom of association, the right to freedom of thought and the right to express, hear and consider any political and religious viewpoint within public discourse.
It is worth noting that the idea of a person's sovereignty over his/her own mind plays a crucially different role in Scanlon's famous autonomy-based approach to the justification of viewpoint neutrality as a requirement of political legitimacy than in my statusbased approach. 60 In 'A Theory of Freedom of Expression' (1972), Scanlon sets out the argument from autonomy as a constraint on justifications of political authority. According to this argument, persons or citizens who regard themselves as sovereign in deciding what to believe and in weighing reasons for action would not be willing to grant the state the power to decide what arguments and viewpoints they should be permitted to hear and consider. 61 In contrast to my status-based approach, Scanlon's argument from autonomy has a contractarian character -in the sense that it asks what rational, autonomous agents would agree to. If the aim is to defend viewpoint neutrality, a problem facing Scanlon's theory is that it is far from clear that the outlined justification rules out viewpoint-based restrictions on public discourse. It is not unreasonable to assume that rational, autonomous hypothetical contractors would be willing to empower the state or the majority to enact and enforce certain viewpoint-based restrictions on extremist speech to protect themselves and others as potential victims of harms brought about by expressions of such extremist viewpoints. 62 To throw light on how Scanlon's autonomy-based theory differs from my statusbased theory, it is also interesting and important to point out that there are two different ideas of autonomy at play in Scanlon's theory. In 'Freedom of Expression and Categories of Expression', 63 he explains the relation between these ideas. According to Scanlon, the normative appeal of (1) the idea of 'autonomy as a constraint on justifications of [political] authority' 64 derives entirely from (2) the idea of 'the value of autonomy' (especially the value of audience autonomy) 'understood as the actual ability to exercise independent rational judgment, as a good to be promoted'. 65 'Its appeal [i.e. the appeal of (1)] derives entirely . . . from the importance of protecting central audience interests' -that is, 'interests in deciding for oneself what to believe and what reasons to act on'. 66 Scanlon's idea of the value of autonomy is, however, a problematic normative foundation for viewpoint neutrality. If the normative appeal of a legitimate political system that treats its citizens as autonomous agents 'derives entirely' from the value of autonomy, why should it not be permissible for a liberal democratic state to prohibit certain political viewpoints if such prohibitions promote the value of autonomy or the overall realization of this value and audience interests? For example, why should it not be permissible for the state to ban false and misleading political viewpoints or hateful political viewpoints if they diminish people's capacity for rational deliberation or their 'actual ability to exercise independent rational judgment'? In my argument for the civil libertarian doctrine of viewpoint neutrality, the idea of respect for the status of persons as thinking agents and their sovereignty over their own mind is not a value or 'good to be promoted' but rather a status-based side constraint on the pursuit of value and interests -including what Scanlon calls the value of autonomy and audience interests.
At this point, it should be noted that respect for the status of persons as thinking agents has justificatory priority over the protection and promotion of interests or the promotion of good consequences. For example, the reason why it is fitting to grant hate speakers and their potential audience (i.e. the rest of us) the right to participate in public discourse as speakers and listeners free from viewpoint-based restrictions is not that their interests outweigh the interests of the targets of hate speech. Rather, it is respect for the status of persons as thinking agents and their sovereignty over their own mind that provides the basis of this right -regardless of the importance or weight of their interests in expressing, hearing or considering extremist viewpoints. As Warren Quinn points out, [i]t is not that we think it fitting to ascribe rights because we think it a good thing that rights be respected [or that such an arrangement best promotes the interests of all affected parties]. Rather, we think respect for rights a good thing precisely because we think people actually have them -and . . . that they have them because it is fitting that they should. 67 This status-based approach to the justification of basic rights is crucially different from Waldron's interest-based approach that is inspired by Joseph Raz. Raz assumes that P has a right if P's interest is a sufficiently strong reason for holding some other person(s) to be under a duty. According to Raz, rights are the result of an evaluation of the interests or well-being of persons affected. ' [R]ights are based on evaluating the interests not only of their beneficiaries, but also of others who may be affected by respect for them'. 68 The second aspect of my status-based theory is that basic rights are deontological side constraints that provide a basis for assessing and constraining the actions and decisions of the state or the majority as well as individuals and groups. Status-based rights impose what Robert Nozick calls 'moral side constraints' on the pursuit of the interests and goals of other people and the state. These moral side constraints reflect the limited authority individuals, groups and states have over persons regarded as thinking agents with sovereignty over their own body, mind and life. A specification of our basic rights as deontological side constraints is a specification of our rightful freedom and our rightful powers and immunities with respect to one another.
The third aspect of the status-based theory that I defend is that the strength of basic rights is a reflection of the status of persons as thinking agents. The justificatory priority of respect for the status of persons over the protection and promotion of interests is crucial when considering the strength of basic rights. Although rights typically protect interests directly and indirectly, the strength of a basic right, such as the right to participate in public discourse, can be out of proportion to the interests it protects and promotes. As Frances Kamm points out, we can recognize that any given person's interest in speaking freely is not great, yet still argue that he has a strong [status-based] right to free speech, even when its strength is independent of serving (directly or indirectly) any other interest of his or anyone else's. 69 When we consider whether a basic right should be respected, its strength is not, as Waldron assumes, a function of the interests at stake -the issue of justifiable infringements of basic rights is not a matter of balancing competing interests.
According to the status-based theory of rights that I defend, basic rights are trumps in, at least, two important senses. First, basic rights trump moral considerations of aggregative welfare. The normative force of basic rights places them on a different plane from any aggregative calculus of interests. A basic right is not just another consideration to be weighed and balanced against conflicting interest-based or welfare-based reasonswhatever their strength. Second, a basic right is a trump in the sense that the overall promotion of respect for basic rights in a society cannot justify a violation of the right. As Robert Nozick points out, basic rights should not be treated as goals to be promoted, but as deontological side constraints to be respected -that is, as constraints that reflect respect for the status of persons and their sovereignty over their own body, mind and life. Consider the basic right to freedom of association. It is reasonable to assume that if a state respects this basic liberal right, this will have the consequence that some citizens form extremist political or religious organizations that provide breeding grounds for violent radicalization and that some members of these organizations will at some point in the future violate the basic rights of other persons. Nevertheless, the right to freedom of association is a constraint on the actions of the state that makes opportunistic violations of this right impermissible -even if violations would over time minimize the total amount of violations of basic rights in the society. 70 The proposed status-based theory of rights provides a basis for a specification of the scope and strength of the right to participate in public discourse that supports the civil libertarian doctrine of viewpoint neutrality. Respect for the status of persons as thinking agents and their sovereignty over their own mind (i.e. their own beliefs and values) requires that they are ascribed a basic right to participate in public discourse as speakers and listeners free from state-imposed viewpoint-based restrictions. A political system that does not respect this basic right as a deontological side constraint fails to respect persons. Persons regarded as thinking agents can complain about the indignity or disrespect of being told by the state or a majority what political or religious views they can express, hear and consider in processes of public discourse. To subject competent adult citizens to viewpoint-based restrictions is to treat them like children who need protection from being exposed to dangerous, poisonous or contagious viewpoints. A democratic state does not respect persons and their sovereignty over their own mind if it functions as a moderator of public debate and deliberation that suppresses or censors the political ideas and viewpoints that they are allowed to express, hear and consider. Viewpointbased restrictions on extremist speech within public discourse constitute an indignity and insult to all persons who are subject to them -both speakers and their potential audience. This also applies to members of minority groups who are targets of extremist speech, such as hate speech. I will return to this point subsequently.
The status-based doctrine of viewpoint neutrality that I defend faces the following questions. What about respect for the status of persons who are the targets of respectdenying speech, such as hate speech? Do persons have a basic right to dignity or recognition that protects them from hate speech that expresses viewpoints that attack their dignity and that are intended or likely to stir up hatred or contempt against members of certain groups? 71 If political institutions should respect the status of persons as thinking agents, should it not be permissible for a liberal democracy to ban the expression of such extremist political viewpoints if the views have a silencing effect, and the viewpoint-based restrictions will promote the freedom or opportunity to participate in public discourse for people overall? According to Alexander Brown's silencing effect argument, it is permissible or pro tanto warranted for a majority to enact viewpointbased restrictions on extremist speech such as hate speech, if they operate for the sake of ensuring that all citizens have the opportunity to participate normally in public discourse -that is, as 'ordinary deliberative democrats'. 72 My answer to these questions is threefold. First, persons regarded as thinking agents have no basic right that other individuals, who believe they lack dignity, not hold or express these convictions as individuals 73 -even if they intend to stir up hatred or contempt against them. This applies to all persons regardless of whether they belong to a vulnerable minority or whether they are communists, libertarians, social democrats, racists, religious fundamentalists, islamophobes, alcoholics or junkies. As Ronald Dworkin points out, '[l]iving in a . . . society whose government respects human dignity means that I must accept the right of others to hold me in contempt' or express viewpoints that attack my dignity. 74 According to the civil libertarian doctrine of viewpoint neutrality, a majority has no moral right to enact laws that prohibit or penalize hateful viewpoints that express an aspiration or wish to deprive members of certain groups of all or some elements of citizen status. This is a matter of the legitimate jurisdiction of majorities in a liberal democracy.
Second, political institutions do not show wrongful disrespect for people if they treat them as persons who have the deliberative capacities to evaluate, deal with and respond to political and religious viewpoints that attack their dignity or elementary social reputation. Rather, respect for the status of persons as thinking agents requires that they are treated as persons who have the ability to deal with such viewpoints and the ability to defend and stand up for themselves. In his famous defence of hate speech laws, Richard Delgado seems to defend the competing perspective that regards the targets of hate speech as weak and defenceless victims who need special state protection from 'words that wound'. 75 At this point, one might object (as an anonymous referee did) that my position seems to highlight the agential dimension of minority members' ability to respond to extremist speech targeting them, whereas the more structural aspects of this situation is neglected. With regard to the silencing effects of extremist speech, this objection focuses on what can be called the structure-agency relation. This concerns the relation between the structural background conditions of members of minority groups and their agency -that is, their capacity to act, such as their capacity to participate in political processes. The structural background conditions include, among other things, economic conditions, such as job opportunities and access to economic resources; educational conditions, such as educational opportunities; and security conditions, such as access to state protection from violence and insecurity. The idea is that there is a relationship between (a) the structural background conditions of members of minority groups who are targets of extremist speech and who are exposed to such speech and (b) silencing effects that inhibit target-group members' opportunity to participate 'normally or as other citizens' 76 in public discourse. The structural background conditions constitute a significant aspect of the environment of opportunities in which agency is exercised in a given society, and extremist speech pollutes or poisons this environment in a way that can silence members of targeted groups or reinforce silencing effects. However, all members of a given minority group in a state are not necessarily subject to the same structural background conditions. Thus, I assume that the outlined considerations about the structure-agency relation and silencing effects primarily apply to disadvantaged members of minority groups who are subject to unfortunate structural background conditions that in various ways negatively influence the environment of opportunities in which agency is exercised.
From an empirical point of view, these ideas about the structure-agency relation seem to be reasonable. I agree that unfortunate structural background conditions or inequalities constitute a problem that can play an important role when it comes to how extremist speech can have a silencing effect for targeted members of minority groups. However, from the normative point of view of the civil libertarian doctrine of viewpoint neutrality and the status-based theory of rights that I defend, the question of whether it is permissible to ban extremist speech that can have silencing effects is a matter of whether the speech acts at issue have a silencing effect that amounts to a direct and demonstrable violation of the basic rights of other persons (I will return to this subsequently). From this normative standpoint, it is not permissible for a liberal democratic state to prohibit the expression of viewpoints within public discourse as a means to mitigate or deal with problems related to unfortunate structural background conditions or inequalities. Having said that, the proposed doctrine of viewpoint neutrality does not imply that it is impermissible for the state to do something about these structural background conditionsespecially if they are unjust -by other means than banning political viewpoints within public discourse. For example, if the security condition of Muslims in a given state is a problem that makes them vulnerable to unacceptable levels of violence or insecurity that inhibit their ability to participate in political processes, the state should address the security problem and not violate the basic right to express, hear and consider viewpoints as a means of promoting Muslims' freedom or opportunity to participate in public discourse. This leads to my next and most important response to Brown's silencing effect argument.
Third, the proposed status-based theory regards viewpoint neutrality and the associated free speech right to do moral wrong as deontological side constraints that prohibit the state from violating this constraint even if a violation would better serve freedom of expression overall in the society. This means that the basic right to participate in public discourse free from viewpoint-based restrictions can be invoked against a democratic state that would ban the extremist political viewpoints of a Neo-Nazi group, even when the group, if allowed to express their hateful viewpoints, is likely to stir up hatred or contempt and psychologically 'deter or inhibit members of targeted groups from functioning as ordinary deliberative democrats' -that is, the targets are prevented 'from participating normally or as other citizens in the formation of democratic public opinion'. 77 If a state is prepared to ban political viewpoints to promote the freedom or opportunity to participate in public discourse for people overall, then it does not (as Nozick would have said) endorse the basic right to express, hear and consider any political viewpoint within public discourse as a deontological side constraint on its actions that reflects and respects the status of persons as thinking agents. Rather, it merely treats this right to participate in public discourse as a goal that should be promoted, and this would allow the state or the majority to violate or transgress this basic right when doing so promotes the overall freedom or opportunity to participate in public discourse. To treat viewpoint neutrality and the associated free speech right to do moral wrong as a deontological side constraint means that the state cannot justifiably fail to respect or instantiate this status-based constraint simply because an opportunistic breach or transgression promises to promote the overall realization of the freedom or opportunity to participate in public discourse better than respecting this constraint would do.
The proposed civil libertarian doctrine of viewpoint neutrality and status-based theory of rights do not imply that it is never permissible for a liberal democracy to ban speech acts that silence or aim to silence other persons. As I pointed out in section 5, the free speech right to do moral wrong does not cover speech acts that directly and demonstrably violate the basic rights of other persons. This exception clause means that if a speech act silences or aims to silence others in a way that constitutes a violation of other persons' basic right to participate in public discourse, then such speech acts fall outside the scope of the free speech right to do moral wrong. I believe that speech acts that fulfil one of the two following conditions can silence others in a way that amounts to a violation of the right to participate in public discourse. 78 1. The coercion via threat condition: A speaker A communicates a serious expression of intent to commit an act that will inflict serious harm 79 upon B or other people close to B to obstruct B's freedom to participate in public discourse or prevent B from exercising this right. 80 It is irrelevant whether A (or A's associates/co-conspirators) actually intends to carry out the threat, and whether A (or A's associates) has any intention of carrying out the threatened act imminently or not. This coercion via threat condition is fulfilled in situations in which A communicates a serious expression of intent to kill B or his/her children if B participates at a political demonstration or if B publishes certain political or religious viewpoints. In such cases, A's intention is to coerce B into acting against his/her will -the conditional threat is a form of coercion. 2. The incitement to imminent violence condition: A's speech act is intended and likely to incite C to imminent use of severe physical force or violence against B to obstruct B's freedom to participate in public discourse or prevent B from exercising this right. This condition is fulfilled if A incites a group of Neo-Nazis to imminent use of violence against B to prevent B from participating as a speaker in a public debate, and it is likely that the Neo-Nazis will be persuaded by A to commit imminent violence in the given context. The condition is also fulfilled in a relevantly similar situation in which A incites a group of radical Islamists to start a violent riot to prevent B and others from participating at an anti-Islamic political meeting or political demonstration. Although it is outside of A's control whether the Neo-Nazis or the radical Islamists will in fact be persuaded to use violence to prevent B, I think that A's act of incitement to imminent violence qualifies as a violation of B's basic right to participate in public discourse. What make the incitements at issue a violation of B's basic right is the combination of the intent with which the words are communicated and the high degree of certainty that the Neo-Nazis or the radical Islamists will act on the incitement.

The argument from democratic legitimacy
In contrast to Waldron, I will argue that democratic legitimacy requires viewpoint neutrality. My argument for this position can be called the argument from democratic legitimacy. According to this argument, the legitimacy of the democratic exercise of political power requires viewpoint neutrality (and the associated legal free speech right to do moral wrong) to respect persons as thinking agents, under circumstances of deep disagreement. The argument from democratic legitimacy proceeds from three assumptions. The first assumption is that political institutions in a liberal democracy should show respect for persons as thinking agents. As already noted in section 6.1, this is a normative idea that democratic institutions and procedures should respect and instantiate. Respect for the status of persons as thinking agents and their sovereignty over their own beliefs and values provides the foundation of the basic choice-protecting rights of persons, as well as rights-based requirements of democratic legitimacy, such as the basic right to freedom of political thought -that is, the right to make up one's own mind about political issues, problems and ideas.
The second assumption is that deep disagreement is a permanent feature of liberal democracies. For example, there is disagreement about how basic political, legal and economic institutions should be organized, and how these institutions should distribute benefits and burdens -that is, rights, duties, opportunities and resources. Deep disagreements about justice, the common good and the ideals of a good political society are, for instance, reflected in debates between liberal democrats, communists, fascists and various groups of religious fundamentalists. More generally, individuals and groups deeply disagree on matters of great importance to them in view of their fundamental moral, political or religious convictions.
The third assumption is that in a democracy political power belongs to citizens, and this means that they can exercise significant coercive power over one another. The way citizens and their elected representatives exercise this coercive power applied by the democratic state and its apparatus of enforcement has a significant impact on the interests, aims, opportunities and life-prospects of different individuals and groups, who have competing political doctrines and competing conceptions of what gives meaning and value to life. Despite the fact of deep disagreement, all members of a democratic society are bound by the decisions of the majority -even though they think that these decisions and the political institutions that underlie the decisions made are unjust or in clear conflict with their fundamental moral, political or religious commitments and convictions.
The outlined normative and empirical assumptions provide a basis for viewpoint neutrality as a rights-based requirement of democratic legitimacy. This rights-based requirement of democratic legitimacy has two aspects, and it can be formulated like this: In the process of political deliberation in a liberal democracy, all citizens -including unpopular and extremist dissenters whom many despise -should have (a) a basic right to express and defend their political thoughts, convictions and views and (b) a basic right to hear and consider the political viewpoints or thoughts of other persons. This viewpoint neutrality requirement of democratic legitimacy does not mean that liberal democracies should grant radical extremist dissenters the right to attain political power to turn their political ideas and objectives into law. 81 It only means that extremist political viewpoints and ideas should not be prohibited.
The proposed viewpoint neutrality requirement of democratic legitimacy rests on three normative ideas about what respect for persons as thinking agents requires in a political setting. First, in the process of intrapersonal and interpersonal political deliberation in a liberal democracy, respect for persons as thinking agents and their sovereignty over their own mind requires freedom of political thought -that is, a basic right to make up one's own mind about matters concerning the organization and culture of society. Second, there is an inseparable link between freedom of thought and viewpoint neutrality, and respect for the status of persons as thinking agents requires both freedom of political thought and viewpoint neutrality. The right to make up one's own mind about political issues, problems and ideas includes the liberty to develop, express and defend one's political thoughts or viewpoints in processes of deliberation or discussion with others free from state-imposed viewpoint-based restrictions. Third, viewpoint-based restrictions represent a form of state coercion that fails to respect the status of citizens as thinking agents, who have the ability and the right to develop their own political views. In the first place, viewpoint-based restrictions on political speech, such as hate speech laws, constitute a kind of political thought control -the aim of which is usually to prevent citizens from acquiring or developing dangerous or poisonous ideas or viewpoints that can bring about harmful changes in their subsequent behaviour. The state attempts to control (1) what particular political ideas, viewpoints or information people are exposed to and (2) how they think about certain political issues, ideas and aims -for example, to protect a society's moral environment or to prevent violent radicalization. Secondly, it seems odd or incoherent to claim that citizens in a democracy should have the right to participate in political processes and at the same time treat them as if they lack the necessary deliberative capacities to make up their own minds about politics so that they should be protected from (the exposure to) dangerous or poisonous political viewpoints. Thirdly, viewpoint-based restrictions constitute a form of state coercion that is especially problematic with regard to respect for dissenters, who are subject to institutions and laws that they reject, and that are in clear conflict with their fundamental moral, political or religious convictions.

Systemic democratic legitimacy and the majority's moral right to rule
If viewpoint neutrality is, as I argued in the preceding section, a fundamental requirement of democratic legitimacy, what is the delegitimizing impact of viewpoint-based restrictions, such as broad and narrow hate speech laws? My answer to this question is twofold. First, since viewpoint neutrality is a fundamental requirement of democratic legitimacy, any viewpoint-based restriction on public discourse is democratically illegitimate. This means that both broad and narrow hate speech bans are democratically illegitimate, because they violate the basic right to participate in public discourse as speakers and listeners free from state-imposed viewpoint-based restrictions. Viewpointbased restrictions on public discourse are democratically illegitimate, and this is an allor-nothing matter -it is not a matter of degree. This means that I reject Waldron's two central claims about the compatibility of democratic legitimacy and viewpoint-based restrictions on vituperative hate speech. The first is that narrow hate speech bans do not undermine democratic legitimacy because they have little or no impact on hate speakers and their potential audience's interest in participating in public discourse. As we saw in section 4.1, this claim is closely linked to Waldron's balancing approach to justifiable infringements of rights. Since narrow hate speech bans 'bend over backward', he assumes that the interests that the targets of vituperative hate speech have in being protected from attacks on their dignity clearly outweigh the interests hate speakers and their potential audience have in expressing, hearing and considering extreme hateful viewpoints. Waldron's second claim is that narrow bans promote democratic legitimacy, because they protect and promote the 'conditions of a healthy working democracy', such as social peace, mutual respect and trust.
My argument for rejecting Waldron's compatibility claims proceeds from two assumptions. The first is that viewpoint neutrality and freedom of thought are rightsbased requirements of political legitimacy in a democracy grounded in respect for the status of persons and their sovereignty over their own mind. The second is that these rights-based requirements of democratic legitimacy should be regarded as deontological side constraints on the actions and powers of the state or the majority, in the sense outlined in section 6.1. From this deontological point of view, it is impermissible to violate these status-based constraints to protect and promote the interest individuals and groups have in being protected from attacks on their dignity; their interest in living in a moral and social environment of mutual respect; or their interest in living under conditions that sustain a healthy working democracy. As we have seen, Waldron argues that these interests provide strong (first order) reasons for action -that is, interest-based reasons for the enactment and enforcement of certain viewpoint-based restrictions on extremist speech within public discourse. However, a basic right is a deontological side constraint that provides an exclusionary reason -that is, a second-order reason to disregard, and refrain from acting for, certain conflicting reasons, 82 such as the firstorder interest-based reasons Waldron appeals to. The basic right to participate in public discourse free from viewpoint-based restrictions provides a second-order exclusionary reason for others and the state to disregard these interest-based reasons for legal regulation, regardless of their strength or weight. More precisely, this basic right provides a reason for non-interference by others and the state, 83 as well as an immunity against being subject to the majority's power to enact viewpoint-based restrictions on public discourse, and that reason is an exclusionary reason not to act on certain conflicting reasons, such as those outlined above and in section 6.1. 84 At this point, it is important to note that the strength of the exclusionary reason is not put to the test in these cases. It prevails in virtue of being a reason of a higher order. 85 This understanding of basic rights as deontological side constraints that function as exclusionary reasons rules out Waldron's balancing approach to justifiable infringements of basic rights or rights-based requirements of democratic legitimacy. The basic right to express, hear and consider any viewpoint within public discourse provides an exclusionary reason that is not just another consideration or reason to be weighed and balanced against conflicting interest-based reasons. This basic right supplies a second-order reason which excludes rather than outweighs the conflicting interests-based reasons that Waldron appeals to.
Second, even if any viewpoint-based restriction on public discourse is democratically illegitimate, systemic democratic legitimacy nevertheless is a matter of degree. The democratic legitimacy of a state's political system or its democratic constitution is not an all-or-nothing matter. A political system or a democratic constitution can be more or less legitimate. In an otherwise legitimate democratic state, a viewpoint-based restriction -such as a hate speech ban -does not in itself or in isolation destroy the overall democratic legitimacy of the state's political system. The viewpoint-based restriction undermines (or diminishes) the democratic legitimacy of the state's political system or its democratic constitution, but the state still has a moral right to rule or exercise coercive political power over its subjects. Thus, in an otherwise legitimate liberal democracy, the illegitimacy of a hate speech ban does not remove the state or the majority's power-right to enact, and authorize the enforcement of, other laws against persons who have been prevented from publicly expressing their hateful viewpoints. Hate speakers do not have a right to immunity from such state coercion. This aspect of my theory of political legitimacy in a liberal democracy differs significantly from Dworkin's theory. According to Dworkin, '[t]he majority has no right to impose its will on someone who is forbidden to raise a voice in protest, or argument or objection', and this includes people who are banned from expressing their hateful viewpoints -that is, 'people [who] have no right to pour the filth of pornography or race-hatred into the culture in which we all must live'. 86 He points out that we cannot ban hate speech 'without forfeiting our moral title to force such people [i.e. hate speakers] to bow to the collective judgements that do make their way into the statute books'. 87 Dworkin's idea seems to be that if hate speakers have no right to express their hateful viewpoints or convictions, then a liberal democratic state has no moral right or moral title to rule or exercise coercive political power over them -that is, the majority cannot legitimately impose its will on them.
Although it is a central idea of my theory of democratic legitimacy that any broad or narrow hate speech law is a viewpoint-based restriction that is democratically illegitimate and that such restrictions on extremist speech undermine the legitimacy of a state's political system or its constitution, the illegitimacy of a hate speech law does not remove the state or the majority's moral right to rule over racists or other hate speakers in an otherwise legitimate liberal democratic state. For example, hate speakers do not have a right to immunity from laws that protect the basic political and civil rights of all citizens -even if they are prevented from using hate speech to express their view or conviction that members of certain groups (e.g. Muslims or Jews) should be denied some of these basic rights or to express their aspiration or wish to deprive members of these groups of all or some elements of citizen status. The protection of basic political and civil rights is a fundamental requirement of a legitimate liberal democratic political system, and such political regimes can legitimately impose such laws on hate speakers. The illegitimacy of a hate speech law does not remove the state's moral right to rule over hate speakers, but it does require that the state removes the illegitimate viewpoint-based restriction on public discourse. Hate speakers and their potential audience (i.e. the rest of us) have a basic right to immunity from state-imposed viewpoint-based restrictions on public discourse, and this means that such restrictions should be repealed. More precisely, I believe that the best available institutional mechanism for protecting viewpoint neutrality and other requirements of democratic legitimacy is a system of judicial review of the constitutionality of legislation. First, requirements of democratic legitimacy should be incorporated into the constitution of liberal democracies. Second, constitutional courts 88 should be empowered to review the constitutionality of legislation and invalidate viewpoint-based restrictions on public discourse, such as hate speech laws that are clearly in conflict with the civil libertarian doctrine of viewpoint neutrality.
On the basis of Waldron's theory of democracy, one can object that the proposed system of judicial review faces problems of democratic legitimacy. The main reason for this is that a system of judicial review has a counter-majoritarian character, in the sense that constitutional courts (or unelected judges) are granted the power to impede the will of the majority. In view of Waldron's theory of democratic legitimacy, this is problematic because he assumes that all decisions about ordinary and constitutional laws should be made through majority rule, since this decision rule shows equal respect for persons and their judgements. 89 I have two responses to this objection. First, a countermajoritarian institution like judicial review is compatible with requirements of democratic legitimacy and their justificatory basis if it is designed to serve as a system of checks and balances, the aim of which is to ensure that democratic institutions and procedures respect fundamental requirements of democratic legitimacy over time.
Second, one can accept Waldron's assumption that majority rule shows equal respect for persons and their judgements in the sense that majority rule accords equal weight to each person's vote and views 90 but reject the assumption that constitutional constraints on majority rule necessarily show wrongful disrespect for persons in a political context. The moral right to rule over others is a limited and conditional power-right, and the majority has no moral right to exercise political power in a way that violates fundamental requirements of democratic legitimacy, such as the right to participate in public discourse as speakers and listeners free from state-imposed viewpoint-based restrictions, and the right to freedom of political thought. These requirements of democratic legitimacy are basic rights that individuals hold against the majority in a liberal democracy. This position can be elaborated through an analysis of the nature and ground of the majority's moral right to rule in liberal democracies. People -both individuals and groups -have no natural or basic moral right to rule and exercise significant coercive power over the lives of other people. 91 Likewise, the majority has no basic moral right to rule over the minority -that is, a basic moral power-right to alter the legal rights and duties of the minority. Rather, the majority's moral right to rule is a conditional powerright, in the sense that certain conditions have to be satisfied for its possession and rightful exercise. The conditions I have in mind are as follows: (1) A political system of democratic institutions and procedures must be in place that sufficiently satisfy certain fundamental requirements of democratic legitimacy -that is, requirements that can be justified on the basis of the normative idea that political institutions should respect the status of persons as thinking agents. (2) The decisions made by a majority of citizens and their elected representatives must respect the requirements of democratic legitimacy and their justificatory basis. 92 From this point of view, a system of judicial review on the basis of the outlined rights-based requirements of democratic legitimacy does not show wrongful disrespect for persons if a court invalidates laws that ban certain political viewpoints, since the majority does not possess a moral right to rule over others in this way in the first place. Rather, the proposed system of judicial review protects systemic democratic legitimacy -that is, a democratic constitution that respects the status of persons as thinking agents.

Conclusion
In this article, I have considered the question of whether democratic legitimacy requires viewpoint neutrality with regard to political speech. The point of departure of my discussion has been Jeremy Waldron's interesting negative answer to this question. I have defended a civil libertarian doctrine of viewpoint neutrality that Waldron and other defenders of hate speech laws reject. Proceeding from the idea that political institutions in a liberal democracy should respect persons as thinking agents, I have argued that any viewpoint-based restriction on public discourse -including narrow hate speech laws -is democratically illegitimate, and that hate speech bans are viewpointbased restrictions that undermine systemic democratic legitimacy. 56. The scope of a right refers to what the right is a right to. The strength of a right refers to the weight of the right when it comes into conflict with competing ethical considerations. 57. Some aspects of my status-based approach to the justification of basic rights are inspired by the status-based theories of Kamm (2007), Nagel (2002), Nozick (1974) and Quinn (1989). It should be noted that all rights are not basic or fundamental rights. For example, a number of legal rights that define institutional roles (e.g. the legal power-right of judges) are not based on respect for the status of the right-holders as persons. 58. Nozick (1974,34). 59. Dworkin (1996, 200). It should be noted that to say that persons are thinking agents who have this deliberative capacity does not mean that they always or mostly use this ability in a good or rational way -for example, deliberate thoroughly or rationally about political issues. Rational deliberation and decision-making require both information and the ability to process it rationally. The extent to which thinking agents are informed and rational vary from person to person, and it is a matter of degree. A thinking agent is rarely, if ever, fully informed or fully rational. One important reason for this is that we as thinking agents are subject to cognitive biases that prevent us from believing, thinking or doing what we ought to believe, think and do in view of the information we have available. 60. See 2003, 95-98). I am grateful to an anonymous referee for encouraging me to make these clarifications about how my theory differs from Scanlon's. 61. The aim of Scanlon's argument is to defend what he calls the 'Millian principle': 'There are certain harms which, although they would not occur but for certain acts of expression, nonetheless cannot be taken as part of a justification for legal restrictions on these acts. These harms are as follows: (a) harms to certain individuals which consists in their coming to have false beliefs as a result of those acts of expression; and (b) harmful consequences of acts performed as a result of those acts of expression, where the connection between the acts of expression and the subsequent harmful acts consists merely in the fact that the act of expression led the agents to believe (or increased their tendency to believe) these acts to be worth performing' (Scanlon 1972, 213). 62. For interesting discussions of the contractarian aspect of Scanlon's argument from autonomy, see Greenawalt (1989, 32-33, 115) and Brison (1998, 328-29). 63. This article was first published in 1979 in University of Pittsburgh Law Review. 64. This is the idea of autonomy that provides the normative basis of Scanlon's argument in 'A Theory of Freedom Expression' (1972). 65. Scanlon (2003, 97-98). 66. Scanlon (2003, 98 and 97). 67. Quinn (1989, 312, see also 309). 68. Raz (1994, 35-36). 69. Kamm (2007, 246-47). 70. As Nozick points out, the side-constraint view forbids the state or the majority to violate these constraints even if a violation would 'lessen their total violation in the society' (Nozick 1974, 29). 71. Richard Delgado claims that persons have such a right. '[A] racist insult is always a dignitary affront, a direct violation of the victims right to be treated respectfully' (Delgado 1982, 143). 72. See, for example, Brown (2015, 195, 198). 73. See Dworkin (2012, 342).