Moral Rights,Human Rights and Social Recognition

Conventionalism

In this article I examine a way of conceiving rights that rejects the notion that rights can be ‘moral’ in the way I have just described. It does not claim that there is no legitimate sense in which rights can be described as ‘moral’, but it rejects the view that rights can be rights in virtue of morality alone and without reference to convention. I shall describe this way of conceiving rights as ‘conventionalism’, since it holds that rights are always and only conventional rights.

Within conventionalism, rights are conceived first and foremost as legal rights, but conventional rights can take other forms. They may be conferred by the rules of non-state institutions, such as universities or trade unions or churches. People can also have rights arising from their society's positive morality, that is, from the mores that are actually shared by the members of a society and that are upheld by the sanction of social disapproval.

Another way of describing the conventionalist position is through the language of social recognition and social practice. In the full-blooded version of conventionalism, people have rights only in so far as they are recognised as possessing rights by the society to which they belong or in which they reside. Rights are recognised primarily through being incorporated in a society's law, but they may also be recognised in a society's conventional morality. Accordingly, rights are always social practices; in the absence of the relevant social practice, we can argue that a right ought to exist, but the right will not exist unless and until it exists as a social practice.

In examining this view, I begin by distinguishing between two forms of conventionalism. While these two forms are sometimes elided, they are significantly different. Only the second presents a fully conventionalist understanding of rights and it is upon that understanding that I focus in the remainder of the article. Having set out its main features, I comment critically on its conceptions of legal rights and human rights, before going on to mount a defence of the orthodox conception of moral rights – rights conceived as morally grounded entitlements. However, before turning to that defence, I pause to consider T. H. Green's well-known recognition theory of rights and suggest that his thinking on rights differs significantly from that of the contemporary conventionalists I criticise.

Two forms of conventionalism are extant in current work on rights and their difference is best articulated with reference to how each understands the term ‘moral right’.

Conventionalism (i): Sumner and Campbell

The first form of conventionalism is exemplified by the work of Wayne Sumner (1987) and Tom Campbell (1983, pp. 18–26; 1996, pp. 164–72; 2006, pp. 27–30). They propose that we should understand a ‘moral right’ as a conventional right that people ought to have. If we say, for example, that people have a moral right not to be tortured, that should mean that they ought to have a conventional right not to be tortured. As a matter of fact, they may or may not have that conventional right but, whether or not they actually have it, we can still affirm their moral right not to be tortured because, through that affirmation, we present the right not to be tortured as a conventional right that they ought to have. Similarly we should understand a human right to be a conventional right that everyone ought to have.

Thus Sumner (1987, p. 137) holds that ‘I have a moral right just in case my possession of the corresponding conventional right is morally justified’. His phrasing here might suggest that I can have a moral right only if I already possess the corresponding conventional right, but he makes clear that the relevant test for a moral right is whether the corresponding conventional right can be morally justified, irrespective of whether it actually exists as a conventional right (Sumner, 1987, pp. 136–51). Because we should only ever think of moral rights as morally justified conventional rights, ‘the concept of a moral right analytically presupposes, and is therefore logically parasitic on, that of a conventional right’ (Sumner, 1987, p. 141). But we can still use the term ‘moral right’ to describe conventional rights that ought to but do not exist, as well as those that ought to and do exist. ¹ Campbell (2006, pp. 28–9) holds similarly that ‘To claim that something is a moral right (if it is not a statement about existing societal rights) can be understood as a claim that certain social or legal norms ought to exist. … such rights can be seen as prescriptive affirmations about what rights there ought to be’. ²

Both Sumner and Campbell would therefore allow us to say we have moral rights even when we do not actually possess the corresponding conventional rights, and, to that extent, their stance on rights is not fully conventionalist. On the other hand, they do not think of a moral right as a right that can stand independently of a conventional right, such that it can, qua right, justify the conventional right. The expression ‘moral right’ is simply a way of speaking of conventional rights that are morally justifiable and which ought therefore to exist.

The Sumner/Campbell account is prey to two objections. First, in so far as it aims to capture what we normally mean when we say either that a person has a moral right to x, or that morally a person has a right to x, it fails. If I say someone has a moral right not to be tortured, I normally mean to say that others are duty bound not to torture him and that, if they do torture him, they violate a right that he has and not merely a conventional right that he ought to, but may not, have. If the unfortunate victim has no conventional right not to be tortured and I protest that he has a moral right not to be tortured, I would mean to say not merely that the torturer's act, while it does not currently violate a conventional right, would violate a conventional right if the legal system were as it should be. I would mean to insist that, irrespective of the laws or conventions of the society in which the torturing takes place, the torturer violates a right of his victim. ³

Second, and relatedly, the usage proposed by Sumner and Campbell conflates the ideas of ‘rights we have’ and ‘rights we ought to have’. The understanding of ‘moral rights’ they propose is similar to that captured by the phrase ‘ideal rights’. D. G. Ritchie famously proposed that when people assert moral rights, and especially natural rights, they simply assert rights that they believe people would ideally possess:

Natural rights, when alleged by the would-be reformer, mean those rights which in his opinion would be recognised by the public opinion of such a society as he admires, and would either be supported or at least would not be interfered with by its laws, if it had any laws; they are the rights which he thinks ought to be recognised, i.e. they are the rights sanctioned by his ideal society, whatever that may be (Ritchie, 1903, p. 80; see also pp. 75–6, pp. 80–1, pp. 85–7, pp. 101–3).

Conventionalism (ii): Darby and Martin

If we reject the idea that moral rights can be rights simply in virtue of their being grounded in moral principles or moral reasons, we would do better to reject altogether the notion that there can be moral rights that are not also conventional rights. That is the position adopted by Derrick Darby (2001; 2003a; 2003b) and Rex Martin (1993). They too retain the term ‘moral right’ but their understanding of its proper meaning places them in a different camp from Sumner and Campbell. For Darby and Martin, conventional rights are the only rights there are, and a moral right is simply a morally justified conventional right. While Sumner and Campbell would allow us to speak of a moral right to x even in the absence of a conventional right to x, Darby and Martin would not. On their view, if there is no conventional right to x, there can be no moral right to x, simply because there is no right to x that we might describe as moral (or immoral). We can, of course, still put forward well-grounded moral claims about the rights that people ought to have, but people should not be said to have those rights unless they have them as conventional rights. So on this understanding, the ‘moral’ in ‘moral right’ describes the ethical quality of a conventional right and contrasts with ‘immoral’ or ‘amoral’ or ‘non-moral’. It does not describe the type of foundation in virtue of which the right is a right, such that it contrasts with ‘legal’ or ‘conventional’; nor should it be used to describe a right that ought to, but does not, exist. ⁶

On this view, then, putative rights are really rights only if they are socially recognised; ‘it is the fact of social recognition that is the essential rights-making feature’ (Martin, 1993, p. 53). ‘All rights including moral ones are products of social recognition’ (Darby, 2009, p. 1). Darby (2001; 2004) describes this position as ‘rights externalism’ since it holds that people have rights in virtue of the recognition bestowed upon them by others and not because they themselves possess certain ‘internal’ right-endowing properties. He also describes its view of moral rights as ‘moral rights positivism’, since it conceives the existence of a right, including a moral right, as a matter of social fact (Darby, 2003b). The adjective ‘moral’ has a critical rather than a positive meaning, but the existence or non-existence of the right remains a matter of social fact and one on which the moral quality of the right has no bearing, just as for a legal positivist the moral quality of a law has no bearing on its existence or non-existence as law.

Clearly, what drives Darby's and Martin's view is the thought that people can be said to have rights only in so far as they are actually able to enjoy and exercise them. Likewise, people are rights-holders and can function as rights-holders only if that is a status that their society actually accords them. If, for example, we say that someone has a right to a fair trial in circumstances in which they are tried manifestly unfairly and there is nowhere and no-one to whom they can turn to secure a fair trial, we risk mocking the notion of having a right. In these circumstances, people do not have a right to a fair trial, and its being a right that they ought to have does not alter that fact.

A conspicuous feature of Darby's and Martin's account of rights, including legal rights, is their insistence that maintenance and enforcement matter as much to a right's being a right as does formal recognition (Darby, 2003a, pp. 64–7; 2003b, pp. 2–5; 2009, pp. 86–8; Martin, 1993, pp. 58–60, pp. 64–72, pp. 82–7; 2003). A right that exists in a society's statute books or which has a foundation in its customary law but which the society does nothing to maintain and secure, such that people cannot actually enjoy it, is not really a right at all. At best it is a nominal right. For Darby and Martin a right is a form of ‘social practice’: what matters are not mere legal formalities but whether that practice is actually present and operative in a society. Rights ‘are always actual practices based on social recognition and social support’ (Martin, 1993, p. 53). That is why they treat legal rights as the paradigm case of rights. They allow that people may have rights in virtue of the non-legal social mores of a society that are upheld by the sanction of social opinion (Darby, 2009, pp. 85–6; Martin, 1993, pp. 24–9, p. 86, p. 96). But rights that rely upon social opinion for their existence will often be fragile or uncertain. Legal rights, by contrast, are normally upheld by a society's judicial system and enforced by the use of sanctions, which is why they are normally the clearest and most complete instance of rights. It is also why Darby holds that the rights that people ought morally to enjoy they should usually enjoy as legal rights: if rights are morally justified, they should be fully secured to those who should possess them and that is best achieved by making them legal rights (Darby, 2001, p. 405; compare Martin, 1993, pp. 86–7).

What then, on this view, are we to make of the idea of human rights? If human rights are to qualify as genuine rights, they must meet the same criteria of recognition, maintenance and enforcement as other rights (Darby, 2003c; Martin, 1993, pp. 73–97; 2003). They too must constitute established practices that are recognised by the society in which they are enjoyed and that are protected and promoted by that society so that they constitute genuinely operative practices. Without recognition, maintenance and enforcement, we can have human rights claims – claims that there ought to be human rights – but no actual human rights. Martin accepts that human rights are targeted mainly at governments but, rather than providing us with reason to ground human rights independently of governments, he sees that feature of human rights as providing additional reason for treating government recognition and promotion as necessary conditions for there being human rights. Human rights are distinguished from other rights by their being rights to which all persons have morally valid claims and therefore rights that ought to be secured universally. But for any particular individual, human rights claims constitute claims to be provided with those rights by the government or political system of their particular society. To that extent, human rights are ‘special’ rather than ‘general’ rights; a person possesses them as a member of a particular society, rather than as a member of the human race. That is why Martin holds that all human rights are really civil rights, albeit civil rights distinguished by a certain kind of moral justification.

Conventionalism, Legal Rights and Enforcement

While Darby and Martin challenge the orthodox conception of moral rights, they may seem to stay within the mainstream in their account of legal rights. Yet even their understanding of legal rights demands more than the notion of a legal right is often thought to require. On their understanding, a right is not something that ought to be maintained and enforced; rather maintenance and enforcement are constitutive of a right such that, without them, there is no right. ⁷ Yet ordinarily, we would conceive a legal right separately from its maintenance and enforcement, such that the right was something that might or might not be maintained and enforced, rather than something that maintenance and enforcement brought into being. Similarly, we might present the fact that a right is a legal right as a justification for its maintenance and enforcement and, if our government failed to uphold our rights, we might still insist that it was our rights that it failed to uphold. But on the Darby/Martin view, that would not be possible, because maintenance and enforcement are practices that contribute to the possession of a right, rather than practices that should follow upon its possession.

That indicates that their ‘social practice’ conception of rights entails more than just a rejection of the traditional notion of moral rights. It entails a more general rejection of rights conceived as titles or entitlements. A legal right is not a legal title, which, as a title, might justify its maintenance and enforcement. At best, a legal title will be a token of official recognition and will contribute to the existence of a legal right only as part of a larger and more inclusive social practice. While we normally think of rights as providing reasons, it is hard to see how they can if we adopt the Darby/Martin view. On their view, there may be reasons for rights, but it is difficult to see how rights themselves, conceived as social practices and therefore as states of affairs, can be reasons for anything. ⁸

J. S. Mill, unlike Darby and Martin, did embrace the idea of purely moral rights. But like them, he made enforcement, or something akin to it, part of the concept of what it was to have a right: ‘When we call anything a person's right, we mean that he has a valid claim on society to protect him in the possession of it, either by the force of law, or by that of education or opinion’ (Mill, 1910, p. 149, emphasis added). The maintenance and enforcement that are relevant to Mill's conception of a moral right are those which there ought to be rather than those which exist. Nevertheless, he runs into the same sort of objection as do Darby and Martin. It is not what ought to be maintained and enforced that defines the rights we ought to have; it is the rights that we ought to have that indicate and explain what (normally and in part) ought to be maintained and enforced. ⁹

Conventionalism and Human Rights

The criticism to which the conventionalist understanding of human rights is most obviously subject is that it is at odds with the primary purpose of those rights. That purpose is to set limits to how people may and may not be treated by those who wield political power. If rights are to be adequate for that purpose, they cannot be ordinary legal rights since governments can grant, withhold or withdraw legal rights as they choose. They need to be rights that, in conception, are grounded independently of governments. That is part of the point of describing them as ‘human’ and also why, in previous ages, people like Locke identified some rights as ‘natural’. If human rights are no more than the creations of governments, they cannot perform the role they are designed to perform. As Martin himself phrases this criticism (1993, p. 96), his view ‘makes the human rights people can have wholly dependent on what government chooses to provide’.

Martin (1993, p. 96) gives two reasons why he believes this criticism to be ‘wide of the mark’. First, we can still say that governments are under a moral duty to meet morally justified human rights claims, understood as claims to be given human rights. Second, the rights we think of as human rights may be recognised as rights in a society's conventional morality even if they remain unrecognised in its law. However, neither of these points concedes much to ordinary human rights thinking. Even if a government owes moral duties to its citizens those moral duties do not, for Martin, answer to human rights. Whether human rights are recognised in a society's conventional morality is an entirely contingent matter and, if possession of those rights is dependent upon that recognition, ‘human rights’ will do little for individuals and minorities who are threatened by hostile governments or majorities. It is difficult to see how Martin's view could enable us to say that the Holocaust or the Rwandan genocide violated human rights, even though such genocidal episodes are typically regarded as extreme violations of human rights. ¹⁰

We might expect Martin to seek refuge from this criticism in the laws and institutions of the international world. Internationally recognised human rights are conventional or ‘positive’ forms of right and so would seem legitimately described as ‘rights’ by a conventionalist. But that way of rescuing the traditional role of human rights is not open to Martin (1993, pp. 59–60), since he insists that maintenance and enforcement are constitutive of the practices we call ‘rights’. Thus, if a government successfully flouts the international community and denies its citizens the rights that the international community alleges to be human rights, those will be neither rights that its citizens have nor rights that their government violates.

This feature of Martin's position applies not just to human rights but to any form of morally grounded right that we might want to deploy for critical purposes. It is just most vivid in the case of human rights because of the political role we expect those rights to perform. In his work, Darby has sought to resist the claim that rejecting the notion that people can have rights in virtue of moral reasons alone will result in ‘moral impoverishment’ (Darby, 1999; 2009, pp. 62–73). That is not a case that the idea of social recognition can be used to make. Whatever virtue there may be in making social recognition a necessary condition of rights possession, social recognition cannot compensate for any moral deficit that results from eliminating rights from our moral thinking. If rights can disappear from morality without moral loss, it must be because there are other moral concepts, or moral theories to which rights are not integral, that can do the job just as well. Of course, we might take the view that the substantive claims associated with moral rights ought to be rubbished. There is a strong tradition of hostility not just to the concepts of moral rights, natural rights and human rights, but also to the sorts of substantive claims rights have been used to make. But while Darby (2003c, pp. 210–1) occasionally dallies with that substantive criticism, he is mostly concerned not to reject the moral claims associated with the orthodox concepts of moral rights and human rights, but to deny that we need those concepts to make them.

T. H. Green on Rights and Recognition

Before attempting to defend the orthodox concepts of moral and human rights against Darby's and Martin's claims, I want to pause to consider T. H. Green's celebrated social recognition thesis about rights. While Jeremy Bentham is the historical influence most frequently cited by those who are sceptical of orthodox notions of moral rights, T. H. Green is for Martin a much more significant influence (Martin, 1993, p. 353). He describes Green's Lectures on the Principles of Political Obligation as ‘perhaps the finest book in the philosophy of rights written to date’ (Martin, 1986, p. 104). Darby too finds sustenance in Green's work (Darby, 2009, pp. 30–7, pp. 142–69). How then do Green's claims about the relationship between rights and recognition compare with those of Darby and Martin? I want to preface my answer with some observations on the idea of recognition itself.

‘Recognition’ of the sort with which we are concerned here can come in two forms. ¹¹ Sometimes it is a form of acknowledgement or acceptance. In conformity with this sense, I may recognise that Mozart was a better composer than Haydn, or that Britain is no longer a world power, or that I have been fortunate, or that I have behaved badly. Recognition can be negative; we may, for example, recognise someone as a failure or as untrustworthy. More commonly, however, it is deployed positively and certainly, when people demand or crave recognition, it is positive endorsement that they seek. In this more positive mode, we may recognise, for example, that Charles Dickens was a great novelist or that Usain Bolt is a great athlete. Most societies have processes through which this sort of recognition is formally given; universities accord it through the award of honorary degrees, militaries through the award of medals, and so on. This sort of recognition can also be straightforwardly factual. I may recognise, for example, that (at the time of writing) Usain Bolt has run 100 metres faster than anyone else.

The feature of this sort of recognition to which I want to draw attention is that it does not understand itself to create what it recognises. If, for example, we recognise Dickens as a great novelist, we do not think that Dickens becomes a great novelist through our recognition. On the contrary, we think that Dickens' achievement is independent of our recognition; we simply notice and duly acknowledge it. Similarly, if an army recognises the bravery of a soldier by awarding him or her a medal, it does not understand itself to be making him or her brave through the award of a medal; rather, the award provides formal and approving acknowledgement of his or her bravery. This feature of recognition is even more readily evident when we recognise simple matters of fact, such as the speed of Usain Bolt's running. ¹²

Sometimes, however, recognition takes more creative forms. A group of individuals may recognise one of their number as their leader. Here the act of recognition creates what it recognises: the leader becomes leader in virtue of being recognised as such by the other members of the group. A person might, through his or her activism and force of character, come to assume a leading role without anyone's design and, at some point, the other members of the group may simply acknowledge what is already a matter of fact. But normally, when we speak of a group's recognising someone as their leader, we think of their recognition as an act of conferral; an individual is constituted leader through the group's act of recognition. Accordingly, if they withdraw their recognition, he or she ceases to be leader.

When Darby and Martin speak of recognition in relation to rights, they must mean recognition as creation or conferral. If legal recognition, or some other form of conventional recognition, of a right were to mean ‘acknowledgement’, that would imply that, in some form, the recognised right existed independently and in advance of the recognition, and that implication is precisely what Darby and Martin mean to deny. For them, a right must be a right in virtue of its being recognised by the laws or the other conventional rules of the relevant society. The rules confer the right; they do not notice its pre-existence.

In a well-known couple of sentences, Green (1941, s.136) declared that ‘rights are made by recognition. There is no right but “thinking makes it so”’. That would seem to place him in the same camp as Darby and Martin, but Green's position on rights and recognition is both different from and more complex than theirs. I will comment on it here only cursorily, though not I hope in a way that misrepresents Green.

Green distinguishes between state and society and his claim that rights entail recognition relates primarily to the members of a society rather than to the state. The state has a role in the recognition of rights: it should enact and uphold a system of legal rights, but the rights that it institutionalises should be rights that are already recognised by the society it governs:

[T]he state, or the sovereign as a characteristic institution of the state, does not create rights, but gives fuller reality to rights already existing. It secures and extends the exercise of powers, which men … had recognised in each other as being capable of direction to that common good, and had already in certain measure secured to each other in consequence of that recognition (Green, 1941, s. 132).

What then is this prior form of recognition that yields rights? The members of a society each possess certain powers (i.e. capacities or capabilities) that they should exercise and should therefore be free to exercise. ¹³ The freedom to exercise those powers should be secured for each member because their exercise is essential for the attainment of the ideal good that the members of a society share as their common good. That common good consists in their mutual attainment of self-perfection. It is their capacity to recognise and to be determined by that good that distinguishes individuals as moral persons, and only moral persons – only beings possessing a will – can possess ethical (as distinct from legal) rights. It is the contribution of those rights to the realisation of their holder's will that justifies their possession (Green, 1941, s. 27).

If moral personality and moral rights are linked in that way, why is recognition necessary for the possession of rights? Green explains that, when he says that ‘rights are made by recognition’ and that there is no right ‘but thinking makes it so’, he means that rights exist ‘solely in consciousness’. ‘Nothing is more real than a right, yet its existence is purely ideal’ (Green, 2003, s. 136); it exists only in idea, only in the moral consciousness of those who recognise it. If that were all he meant, virtually all proponents of moral rights could agree with him. However, he clearly means more. Social recognition matters because that recognition secures the freedom of each to realise his powers and it is that ‘securing’ that realises rights; ‘the free exercise of his powers is secured to each member [of a society] through the recognition by each of the others as entitled to the same freedom with himself’ (Green, 1941, s. 25; see also ss. 26, ss. 29–31, ss. 114–7, s. 132, ss. 138–9, s. 142). While social recognition has a cognitive character, it entails more than just noticing and accepting something about the nature of others and ourselves. It entails acting in consonance with what we notice and accept. To recognise that persons have the powers that render them capable of possessing rights is to recognise that they ought to enjoy those rights and that recognition, in turn, spawns the mutually respectful attitudes and conduct among a society's members that serves actually to secure rights for them. The members of a society accord rights equally to one another as a necessary part of their recognising the rightfulness of that state of affairs. It is in that more developed sense that their ‘thinking makes it so’.

How does Green's idea of social recognition map on to the two types of recognition that I distinguished above? Does it acknowledge or does it create? While it seems to do both, it works, in the first instance, through acknowledgement. The members of a society come to acknowledge one another, albeit gradually and slowly, as beings possessed of moral powers that destine them to realise a common good and it is that acknowledgement that leads them to treat one another as possessors of rights. They do not bring rights into existence in the manner of either a legislator or a social contractor. Rather, they arrive at rights by recognising a series of moral truths. They acknowledge what morally is the case and what morally they have no option but to acknowledge. Indeed, it would seem that, in their own eyes, the recognisers must understand themselves to acknowledge rather than to create the rights that they recognise. How else could they come to treat one another as rights-holders and, through that treatment, ‘secure’ rights for one another? Of course, the ‘securing’ part of the process remains essential for Green; it is by securing certain freedoms for one another that the recognisers ‘confer’ rights upon one another. Nevertheless, the recognisers must think of one another as possessing rights (or, if we are to avoid that term, morally grounded entitlements) independent of their recognition, if their recognition is to secure the rights it recognises. If they did not conceive of one another as possessors of rights, why would they treat one another as possessors of rights? Thus, in the recognising process, the conferral of rights seems to depend upon a prior conception of people as right-holders. Perhaps that is how, for Green, the ‘thinking’ of the recognisers makes it so.

We might also ask what role morality plays, for Green, in constituting a right. In Darby's and Martin's thinking, it plays none. Moral considerations may distinguish a right as morally justified, but they will contribute nothing to its being a right. For Green, the answer is again less straightforward. His socially recognised rights are like non-legal conventional rights to the extent that they must be embraced and respected by the society whose members possess them; they cannot be rights that an individual has but that no-one recognises. On the other hand, for the recognisers, the rights they recognise are not merely conventional in that the moral considerations that drive the recognising process imply that they are necessarily moral rights – they are morally grounded entitlements which the members of a society ought to enjoy equally and which each is duty bound to respect. In the absence of those moral considerations, there would be nothing that the recognisers could recognise as rights. Thus, for Green's recognisers, the moral quality of a right is as essential to its having the status and character of a right as is anything that might equate with its conventionality. ¹⁴

In Defence of Moral Rights

As I previously mentioned, the incorporation of rights in morality has frequently been attacked for substantive as well as conceptual reasons, especially when it has led to the assertion of natural or human rights. Bentham rejected natural rights thinking for its wrong-headed morality as well as for its erroneous conception of rights. Raymond Geuss' (2001) and Alasdair MacIntyre's (1985) contemporary assaults on moral rights and human rights have a similarly comprehensive character. More generally, misgivings about the egoism and atomism that rights thinking allegedly promotes are widely entertained and frequently voiced. That sort of substantive critique of rights thinking is not, however, at issue here. On the contrary, a common part of the conventionalist case is that we can insist that rights can be only conventional without sacrificing any of the substantive moral claims that moral rights theorists are so keen to make. ¹⁵

Can we, then, dispense with the idea of moral rights without moral loss? By ‘moral rights’ here and in the remainder of this section, I shall mean rights that are rights in virtue of their moral character or foundation only; I do not mean rights that exist only as conventional rights but that also happen to be morally justified.

What rights bring to morality is a subject-centred perspective such that, in John Rawls' phrase (1993, p. 32), we treat those who possess rights as ‘self-authenticating sources of valid claims’. That is captured in two related ways of thinking that go with rights. The first is the concept of owing: there are certain sorts of conduct, certain sorts of doing or not doing, that we owe to others. What makes that conduct morally rightful and morally required is that it is owed to the person who possesses the right. That does not mean that moral rights claims must be logically primitive, such that there is nothing to be said beyond insisting that there are some things we just ‘owe’ others. A good deal of moral argument may need to go on before we reach a right but, having reached the right, the concept of owing explains how using the idea of a right makes a difference to our moral thinking.

The second concept is that of wronging. When we violate a right, we do not merely do wrong, we wrong someone. Introducing the idea of rights means that ‘wrong’ ceases to be a noun or an adjective only; it can also function as a verb and as a transitive verb. We have the idea of not merely acting wrongly or committing a wrong but the more specific idea of wronging someone. Of course, wronging is really only the flip side of owing, but considering it as an idea in its own right helps to bring out the moral thinking special to rights.

These ideas also indicate the idea of moral status associated with rights thinking. We ascribe that status only to beings to which we think that we can owe certain forms of conduct and that we believe we would wrong if we failed to treat them accordingly. That, for example, provides the conceptual test for whether we should extend rights to more than just human persons.

It is this central feature of rights that, I believe, lies behind the choice (or will) and interest (or benefit) ‘theories’ of rights, and the rivalry between them. Each is trying to capture the idea of owing: in one case owing is signalled by the right-holder's having control over the performance of another's duty; in the other it is indicated by the right-holder's being the intended beneficiary of a duty. Neither ‘theory’ does the job adequately, which explains the interminable wrangling over which theory is correct. ¹⁶

Now we could, of course, take the term ‘rights’ out of morality and replace it with a synonym. While conventionalists reject the notion of moral rights (as I am using it here), they are quite happy to speak of ‘claims’ that are ‘morally valid’ or ‘morally justified’ and to accept that some if not all such claims should be instantiated as legal rights. Proponents of moral rights, like Rawls (1993) and Joel Feinberg (1970), often use the same terminology. The trouble with these proposals is that we do not think that all ‘claims’, not even all ‘morally valid’ or ‘morally justified’ claims, amount to moral rights. Something more is needed; hence the efforts of Ronald Dworkin, Robert Nozick and Joseph Raz to characterise rights respectively as ‘trumps’, or ‘side-constraints’ or ‘exclusionary reasons’ (Dworkin, 1978, pp. 90–4, pp. 364–8; 1985, pp. 335–72; Nozick, 1974, pp. 28–33; Raz, 1975, pp. 35–48; 1986, pp. 186–7, pp. 195–6). ¹⁷ But the more important objection to substituting another language for the language of rights is that it is singularly pointless. ‘Entitlement’ is a better synonym for ‘right’ than ‘claim’, but what would we gain if we substituted it for ‘right'? If conventionalists want to hold that rights can be removed from morality without loss, they have to show not merely that we can avoid the language of rights but that we can strip out of morality, without loss, the whole idea of rights and all that it entails.

Circumlocutions, which aim to get to the same place as moral rights, but by way of reference to moral duties or moral obligations instead, are equally pointless. For example, saying that people owe duties to others and that failure to perform those duties wrongs the people to whom they are owed, does not avoid rights; it simply talks about them in another way. Not all duties answer to rights, but many do and it is rights that give them their point. Again, we could no doubt give an explanation of why those moral duties were duties without making any express reference to ‘rights’, but the idea of rights would still be present in, or implied by, our explanation.

It is a curious feature of some who oppose the concept of a moral right that they have no equivalent misgivings about the concept of a moral duty. That can make sense in so far as duties logically do not entail rights, but it is puzzling in so far as duties, as relational duties, are coupled with rights. Some legal duties correlate with, or arise from, legal rights and the conventionalist can insist that, if they are to count as legal duties, they must meet the same criteria of recognition, maintenance and enforcement as legal rights. But in so far as a moral duty correlates with, or arises from, what we would normally describe as a ‘moral right’, what is it supposed to correlate with, or to arise from, that is other than a right? ¹⁸ Martin's answer is a ‘morally valid claim’, which, if it is no more than a morally valid claim, falls short of a right because it lacks the social recognition necessary for it to be a right (Martin, 1993, pp. 85–6, pp. 92–7). But why does the same logic not apply to duties? Why should our thinking about what is ‘due from’ and what is ‘due to’ be so radically different? ¹⁹

So it seems to me that the idea of rights, as opposed merely to the language of rights, does contribute something distinctive to our moral thinking. That distinctiveness explains why many of us continue to think of human rights as moral rights first and foremost. There are certain things that it is morally wrong to do to, or to not do for, any human person. There are certain actions or inactions that are owed to anyone who is a human person and whom we wrong if he or she is not treated accordingly. It is that conception that makes human rights such a potent idea when we seek to constrain political power. It is a conception that is wholly at odds with the claim that human rights, properly so called, can be granted only by governments. I do not pretend that this conception must be present in every usage of ‘human right’, but it is very widely shared and its substance would be difficult, if not impossible, to capture without the idea of moral rights.

None of this is to deny that social recognition matters. If rights matter, their recognition must matter. Nor is it to deny that rights grounded in law or custom can be sources of moral obligation. For example, in the absence of legal title a person may have no other title to a good but, once the person has legal title to it, theft of the good can, ceteris paribus, wrong the title-holder morally as well as legally. The same can hold if custom rather than law is the source of title. Thus, while we commonly suppose that the moral should shape the legal and the customary, the legal and the customary can also shape the moral. My purpose has been to deny only that moral rights must always be parasitic upon legal or customary norms.

If we are to speak of moral rights, including moral human rights, we must not, of course, imply that people ‘have’ them in the same manner in which they ‘have’ conventional rights. People's having or not having conventional rights is a matter of fact. Their having or not having moral rights is not. Moreover, the moral rights they have will be controversial to the extent that the moral reasons that underlie them are controversial (compare Frey, 1980, pp. 14–7; 1983, pp. 46–66; Nielsen, 1968; Young, 1978). But moral rights are controversial only because morality is controversial and that controversy provides no more reason to excise rights from our moral thinking than to excise other concepts such as duty, justice or virtue.

The language of ‘existence’ is particularly unfortunate when used in relation to moral rights. Sumner approaches the question of what should count as rights by investigating and setting out ‘existence’ conditions for rights (Sumner, 1987; compare Geuss, 2001, p. 145; Martin, 1993, p. 74). Unsurprisingly, both moral rights and natural rights (as these are ordinarily understood) perform poorly when tested against these existence conditions. People often ask whether human rights ‘exist’, and proponents of human rights are frequently as keen to answer ‘yes’ as are opponents to answer ‘no’. In fact, the question should be neither asked nor answered. In so far as we think of human rights as moral rights, the question is no more appropriate than asking whether the categorical imperative or the principle of utility ‘exists’. None of these moral notions exists in any ordinary sense of ‘exist’, but that renders them neither unintelligible nor indefensible as moral ideas.

Conclusion

There are obvious attractions in the conventionalist understanding of rights. The most compelling in Darby's and Martin's thoroughgoing version, though not in Sumner's and Campbell's halfway house, is that it removes any ambiguity from claims that people ‘have’ rights: people will ‘have’ rights only if their rights are a reality, that is, only if people are able, as a matter of fact, to enjoy them. That enables us to avoid confusion between the rights that people have and the rights that they ought to have, which loose talk about moral rights or human rights sometimes promotes. However, the conventionalist view is a revisionist view in that both moral rights and human rights (as these are ordinarily understood) have a long-established and well-entrenched place in our moral vocabulary. There seems little prospect that the conventionalists will succeed in dislodging them from that vocabulary. But more to the point, we would lose something significant and distinctive in our moral thinking if their revisionist project were to be successful, and the loss would outweigh anything that we might gain – or so I have tried to argue.