Ethics Discussions at PEA Soup: Leif Wenar’s “The Nature of Claim-Rights” with Critical Precis by Arthur Ripstein

35 Replies to “Ethics Discussions at PEA Soup: Leif Wenar’s “The Nature of Claim-Rights” with Critical Precis by Arthur Ripstein”

1. My reaction to the article is that Wenar is not close enough to a justificatory interest theory. If I’ve understood the kind-desire theory correctly, it seems to me to have a number of counter-examples.
   The first kind of apparent counter-example is when the R kind-desire forms no part of the reason for the duty held by Ds to phi Rs. In Wenar’s “Prison Warden” example (p. 10), the prison warden wants the prisoner to obey his duty to not escape, but has no claim-right that the prisoner not escape. Wenar avoids the problem by excluding the warden from the content of the prisoner’s duty. I don’t think that is an adequate response to the problem raised by the example. We can replace “Prison Warden” with a revised example of the same point:
   Bribery
   Officer Oliver has stopped Driver Denise for speeding and is about to issue her a ticket.
   1. Denise (qua member of the public) has a legal duty to not offer a bribe to Oliver (qua law enforcement official). [The law specifically prohibits offering bribes to law enforcement officials.]
   2. Oliver (qua law enforcement official) has a desire that Denise not offer him a bribe. [Law enforcement officials, as such, want members of the public to obey the law.]
   3. Denise’s legal duty to not offer a bribe to Oliver is appropriately enforceable. [He can arrest her for offering him a bribe, etc.]
   But Oliver surely does not himself have the legal claim-right that Denise not offer him a bribe. She does not owe her legal duty to not offer him a bribe *to him*. His kind-desire is, so to speak, entirely downstream of her legal duty: it exists as a consequence of the fact that she has that legal duty, and is not itself part of the direction of the duty.

2. Here’s another example of the same basic point:
   Custody Battle
   Paul and Paula are two parents locked in a court battle for custody of their child Charlie, before Judge Judy.
   1. Judy has a duty (qua judge) to modify the custody rights of Paul (qua parent) in the way that best serves Charlie’s interests. [Suppose that the relevant norms make the child’s best interests the decisive consideration in the current circumstances.]
   2. Paul (qua parent) wants Judy to modify his custody rights in the way that best serves Charlie’s interests. [Parents, as such, want what is best for their children, and therefore want, qua parents, custody rights over their children to be modified in whatever way best serves their children’s interests. (Imagine how you would react to a parent who expressed the desire that his custody rights be modified on the basis of his own interests, and not those of his child.)]
   3. Judy’s duty to modify Paul’s custody rights in the way that best serves Charlie’s interests is enforceable. [Suppose she might get thrown off the bench if she makes a decision on arbitrary grounds, etc.]
   Suppose that Judy correctly believes, and openly announces, that Paul is a terrible parent, and that Charlie’s best interests would be served if Paula were awarded full custody. Nevertheless, Judy does not particularly like Paula, and flagrantly awards full custody to Paul. In this case, I think Judy has violated a claim-right of Paula’s, but I do not think she has violated any claim-right of Paul’s. Paul has not been wronged by Judy, even though his desire (qua parent) that Judy decide the case on the basis of Charlie’s best interests has been wrongfully flouted. With respect to that kind-desire, however, the relation of Paul and Paula to Judy’s duty is symmetrical.
   Of course, Paul and Paula also have the kind-desire, qua parents, to actually have custody over Charlie, and only Paula’s desire here has been flouted. So there is a plausible statement of kind-desires for which Wenar’s analysis seems to work. But the problem is that there is also a plausible statement of kind-desires where it does not work.
   The initial problem in both Bribery and Custody Battle seems to me to be that the R kind-desire that Ds perform the duty to phi Rs is not itself any part of the reason why Ds have that duty. It is either derivative of the duty, or in some other way incidental to it. We could tighten the analysis a little by including a justificatory element — that Ds have the duty in part because Rs have the relevant desire. Perhaps this tightens the analysis too much, but it cannot generate false positives that are not already generated by Wenar’s original analysis, where no further connection between the duty and the desire, other than their matching kind-specificity, is asserted. However, even with this move, I think the theory is subject to further false positive counter-examples. (More later …)

3. The first condition of Wenar’s formalizations concerns me. These require that some “supported” statement of S specify, in an appropriate way, the bearer of the claim-right. I suspect that a person sometimes has a claim-right without being specifiable in the way required by condition one. Here’s an example. To keep the sidewalks clear of ice, the city may require each homeowner to remove the snow in front of his or her house. As a homeowner, my failure to fulfill this requirement will inconvenience the disabled person up the street. Because of the allocation of responsibility, this disabled person seems to have a claim-right against me that I deal with my snow. However, the kinds of statements supported by this system of norms simply say things like this: “Homeowners must remove the snow from the portion of the sidewalk in front of their properties.” The particular victims of my norm violations aren’t specified. Generally, the difficulty concerns cases in which duties are allocated in order to secure some public good, and where one person’s failure to perform her allocated duty can deprive another person from enjoying the public good. In such cases, I think we will want to say that the deprived person had a right to the other’s performance of the allocated duty even though the connection between these two people isn’t specifiable (at least by the immediate system of norms) in the way required by condition one.
   I am also curious about the normative system behind promising, and about how we determine which statements are supported. Suppose that A promises B to benefit C (according to C’s own specifications). Why wouldn’t S support the following statement: “If D makes a promise to R1 to benefit R2, then D must keep D’s promise to R1 by benefitting R2″? It seems that Wenar must either deny that such a statement is supported by our normative system of promising, or he must affirm that C has a claim-right against A to fulfill the promise to B. The first option would suggest stricter limits on which statements a system of norms can be said to support. What would justify those limits?

4. Leif Wenar’s “The Nature of Claim-Rights” is an original and important contribution to the analysis of rights. It offers the possibility of theory of rights that breaks from conflict between Will Theories and Interest Theories. Let me ask three questions that may spark some discussion.
   A Proposed Friendly-Amendment
   However, before posing those questions, I would like to propose a friendly-amendment to Wenar’s view. It attempts to resolve a small technical problem. The first condition of the Kind-Desire Theory is:
   1. Some D (qua D) has a duty to phi some R (qua R) where “phi” is an verb phrase specifying an action, such as “pay benefits to,” “refrain from touching” “shoot,” and so on (Wenar 2013, 18).
   On this view, it seems that D’s duty must be a duty to do something to R (e.g., pay benefits to R, refrain from touching R, shoot R, etc.). However, there are cases in which an R has a claim-right that a D do some phi that is not doing anything to R. R’s father rests in the Père Lachaise cemetery in Paris. She travels to Paris once a year to lay flowers at her father’s grave. However, due to work-related conflicts, R cannot go to Paris this year. In conversation with D, R discovers that he will be traveling to Paris. R asks D to lay flowers at her father’s grave. D agrees to do so. R has a claim-right that D lay flowers at the grave. But D’s duty is not to do anything to R. I think that this problem can be fixed by simply removing “some R (qua R)” from the first condition.
   1. Some D (qua D) has a duty to phi where “phi” is an verb phrase specifying an action, such as “pay benefits to,” “refrain from touching” “shoot,” and so on.
   One might worry that removing this phrase will leave Wenar’s analysis unable to account for the directed nature of the duties implied by claim-rights. However, the second condition of the analysis,
   2. R’s (qua R’s) want such duties to be fulfilled (Wenar 2013, 18),
   provides the needed directionality. D duty to lay the flowers is to R because it is R that wants (in Wenar’s technical sense of “wants”) the flowers to be laid.
   Question One: Can Wenar’s analysis be extended to other kinds of rights?
   One virtue of Will and Interest theories is that they provide accounts not only of claim-rights, but also of immunity-rights, liability-rights, etc. For example, Joseph Raz holds that R has a claim-right that D phi iff some aspect of R’s interests is sufficient reason to hold D to be under a duty to phi (Raz 1988, 166). Raz’s account of immunity-rights is clear and coheres well with his account of claim-rights. R has an immunity-right that D phi iff some aspect of R’s interests is sufficient reason to hold D to be under a disability to phi. The notion of one person’s interest being a sufficient reason to justify the view that another person has a duty/disability crosses the accounts of claim-rights and immunity-rights.
   If we try the parallel move with Wenar’s account, we have
   Consider a system of norms S that refers to entities under descriptions that are kinds, D and R. If and only if, in circumstances C, a norm of S supports statements of the form:
   1. Some D (qua D) has a disability to phi; and
   2. R’s (qua R’s) want such disabilities to be fulfilled; and
   3. Enforcement of this disability is appropriate, ceteris paribus;
   Then: the R has an immunity-right in S that the D fulfill this disability in circumstances C.
   The account of immunity-rights produced by this simple-minded substitution of “disability” for “duty” seems strained. The notion of an R wanting a disability to be fulfilled seems strained because the notion of fulfilling a disability seems strained. Perhaps we should replace
   R’s (qua R’s) want such disabilities to be fulfilled
   with
   R’s (qua R’s) want D’s (qua D’s) to have such disabilities.
   However, this would seem to motivate a parallel change in the account of claim-rights.
   R’s (qua R’s) want D’s (qua D’s) to have such duties.
   That seems odd. Claim-right-holders want duty-bears to fulfill their duties. They do not merely want them to have the relevant duties. The notion of enforcing a disability also seems to raise questions. On Condition 3, the duties implied by claim-rights “are appropriately enforced, either by coercing performance or by penalizing nonperformance” (Wenar 2013, 13). It does not seem possible to coerce performance of a disability. The most promising route for explicating the enforcement of immunities would be to consider the notion of penalizing nonperformance of an immunity. Will and Interest theories provide a coherent account of the rights corresponding to each of the Hohfeldian relations. Wenar’s may well be able to as well, but more work needs to be done in this area.
   Question Two: Is Wenar’s theory a version of the Interest Theory?
   I hesitate to pose this question because my first reaction to it is: Who cares? I care about whether Wenar’s theory is a good theory, not whether it is an instance of this or that family of theories. Nevertheless, this question naturally comes to the mind if one has been reading recent work on rights.
   It would be easy to redescribe Wenar’s theory as an interest theory. As Wenar notes, “there is a way in which it is perfectly natural to speak of interests when carrying through a kind-desire analysis of rights” (Wenar 2013, 26). One could recast Wenar’s view as an attempt to defend an Interest theory of rights by suggesting that we understand an interest as “the satisfaction of a role-based desire” (Wenar 2013, 27). I doubt that Wenar would have much objection to such a recasting of his view.
   Question Three: Do rights attach to roles or rules?
   Wenar writes that “the historical unity of rights (I just assert here) is that rights enabled role-bearers to do their jobs” (Wenar 2013, 6). As Arthur Ripstein highlights, in his Precis, “Wenar analyzes the concept of a right in the first instance in terms of the concept of a constructed role.”
   One issue here is historical. My review of the history of rights does not incline me to the view that the historical unity of rights is found in constructed roles. I am inclined to the view, also noted by Wenar, that rights find “their origins in the law” (Wenar 2013, 13). In particular, it is the law as a system of rules that creates the historical unity of rights. Rules are more fundamental than roles. Roles are created by rules, not the other way around. I think it is no accident that Wenar’s description of the role of goalie in soccer begins with a quote from “the official rules of soccer” (Wenar 2013, 2, emphasis added) and that the rules of soccer are called The Laws of the Game. As I see it, the history of rights is not, at a fundamental level, a history of roles, but a history of rules that were first developed and made precise in legal systems and then migrated to morality.
   Another issue is philosophical. If even the historical unity of rights was in constructed roles, rights have left roles behind and moved to rules. As Wenar notes, “history has moved on” and there is “a whiff of the premodern in talk about roles” (Wenar 2013, 17). I do not think it is a
   mistake…to take the analysis of these latecomers (detheologized humans, children, animals, etc.) as its paradigm, and to attempt to explain rights-ascriptions beyond these cases with the independent value of well-being. This is like Puerto Rico invading the United States for the purpose of colonization. The homeland of rights, which is roles, resists the attempt (Wenar 2013, 27).
   After the American Revolution, the French Revolution, the Emancipation Proclamation, Seneca Falls, the Universal Declaration of Human Rights, I Have a Dream, and Stonewall, these latecomers are now paradigms. The notion that rights are not based on roles is a fundamental moral change. It is not like Puerto Rico invading the United States. It is like the United States invading Britain. The country that was once the colony of a super-power is now the super-power.
   Wenar’s theory strains to account for these latecoming paradigm rights. Thus it is natural for him to see it as a mistake to take them as paradigms. Will and Choice theories do not strain to account for these latecoming paradigms. Both can provide natural and plausible accounts of the view that some of the most important rights are independent of roles. While I think that there are other problems with Raz’s theory of rights, his view that the unity of rights is in the “sufficient reasons” found in systems of rules provides a less-strained account of this important feature of modern rights-talk and I think the same can be said of Will theories.

5. My sense is that if role- or kind-desires are determinants of a duty’s direction towards a claim-right holder, they must figure as interests. On p. 27, Wenar claims that we sometimes ascribe interests to individuals, not on the basis of an independent objective theory of well-being, but rather as the expression of a role-based desire: “it is the role-based desire that is the primary concept, and `interest’ means nothing more than the fulfillment of such a desire.” This is all well and good. I don’t think it matters at all which of the two concepts is primary. Perhaps we sometimes ascribe interests to agents only because they are already assumed to have certain role-based desires, and we take the interest to be the satisfaction of that desire. And perhaps we sometimes ascribe role- or kind-desires to agents because we already assume them to have certain kinds of interests as beings of a certain sort. Whatever the case here, it seems to me essential that, once the satisfaction of the desire is taken as an interest, that it bear on the duty as an interest, and not in some quite different way. For instance, the role-based desire might bear on the duty as a permission, as a vice, or as a virtue.
   Here’s a whimsical example where a role-based desire may be relevant as a permission. I assume *everyone* has read Douglas Adams:
   Dish of the Day
   Arthur, Ford, Trillian, and Zaphod arrive at Milliways, the restaurant at the end of the universe, where they ask to see the dish of the day, i.e., a creature specifically bred to want to be eaten, and to be capable of saying so clearly and distinctly. (“We’ll meet the meat, says Zaphod.) Suppose we modify the scenario a little, so as to include a duty on the part of Arthur to not give offence to the restaurant owners, or the other diners, or whoever, by refusing to order and eat the dish of the day. If so, the system of norms in the restaurant could now be such that:
   1. Arthur has a duty (qua diner) to order and eat some tasty part of the creature (qua dish of the day). [Ordering anything else would be offensive and in poor taste.]
   2. The creature (qua dish of the day) has a desire that Arthur (qua diner) order and eat some part of it. [This is what a dish of the day, as such, most wants diners to do.]
   3. Arthur’s duty is appropriately enforceable. [If he orders a green salad, he may be thrown out of the restaurant.]
   I take it that Arthur would not wrong the creature itself by ordering a green salad, and that it certainly has no claim-right to be eaten by him, even though it has a role-based or kind-desire (qua dish of the day) that he honour his duty (qua diner) to eat it. Moreover, diners may have the duty in part because the dish of the day is assumed to have the relevant desire — were it to not want to be eaten, perhaps this would defeat Arthur’s duty as a diner to eat it.
   The content of the Dish of the Day example is of course fantastical, but it is the structure of the case that strikes me as important. A kind-desire may be essential to the justification of a duty to phi beings of the relevant kind, but only insofar as it is taken to defeat (as a permission) what might otherwise defeat the duty, and not as an interest of the being that the duty is there to promote or protect.

6. In response to George Rainbolt’s third question–Do rights attach to roles or rules?–I take Wenar’s position on the _history_ of rights to be correct.
   As I understand the history of, for instance, Roman law and Protestant natural law (i.e. Grotius, Pufendorf, Hobbes, Locke, etc.)–law was organized by roles or offices (Ciceronian officia). So, in early modern systems of practical ethics in which duties, rights, and virtues were enumerated, the roles come first. The most important roles were those of human being, member of a family, and member of a state. Each role had an end or ends that constituted the role, and rights enabled the fulfillment of those ends.
   In Roman law (which I know much less well), the law of persons begins with the distinction between free persons and slaves (related to the distinction between independent and dependent persons). The rules and rights come after one identifies the relevant roles or offices.
   As an historian of early modern ethics and political philosophy, I think Wenar’s work is pretty exciting–it both captures the way in which rights used to be conceptualized in many systems of philosophy and law and it offers a way in which that history might contribute to our present-day debates about rights.
   All that said, I’d be very curious to hear from George how he understands the history such that rules seem more fundamental than roles. Are you thinking about, say, rights as determined by statutory law? Common law? What?

7. Enjoyed the paper, Leif, and got a lot out of it. I have a couple of concerns.
   (1) By focusing on desires, your theory grounds rights in aspects of our psychologies. The will theory does this as well, and consequently faces problems with individuals with limited or non-existent psychologies, such as new-borns or comatose. You claim your theory can accommodate these as right-holders because:
   “Our duties to young children not to abuse them correlate with rights in those young children not to be abused, because young children (like everyone, or perhaps even more than others) have reason to want not to be abused. And the same, mutatis mutandis, for the comatose.”
   But it seems to me that new-born babies and the comatose don’t have reasons because they don’t have psychologies that are responsive to reasons. I’d also be somewhat inclined to say that new-borns aren’t yet capable of desires, and I’m strongly inclined to say this about some comatose patients — imagine someone with sufficient irreparable brain damage.
   (2) As you note later on in the paper, there’s a way of conceiving of “interest” according to which a role/kind-based desires theory comes very close to a parallel interest theory. Consequently, I’m worried that your theory will inherit problems facing the interest theory. In particular, I’m worried that third-party beneficiaries will count as right-holders.
   Consider the example of the parent’s right to child support. The child is a third-party beneficiary, and as such has reason to want the parent receive the child support. So the analysis incorrectly predicts the child is a right-holder.
   Similarly, consider a teacher’s right to decent working conditions. Their students have role-based reasons to want this (since it affects their education), but the students aren’t the holders of this right.

8. Thanks for the paper, and to Arthur Ripstein for the commentary.
   I share some concerns with Ripstein and Dougherty, having to do with the introduction of desires and interests into the account. Ripstein wonders whether desires are superfluous in the account, and the same work couldn’t be done with just a focus on the normative framework of roles, so that the account becomes about “what sort of roles right bearers occupy in relation to each other.”
   Dougherty adds in his comment that if we understand role-based desires as tied to (“semi-solid”) interests, then the role-based account might suffer from counter-examples that plague the interest theory of rights.
   I want to add to this discussion another worry that might push back towards concentrating on roles rather than desires. Consider the parent case: we can assume that parents want the best for their children, but I’m not convinced that this is enough to get to the further premise that parents qua parents desire government provided child support. It depends on whether those parents believe child support to be what is in the best interests of their child. One can imagine, for example, strong libertarians who believe that relying on the government for support for their child is worse – for the child – than being hungry or wearing second-hand clothes or the like. I don’t argue here that this is a good reason, but only that we can imagine parents who do not consider government provided child support to be good for their child, but who still desire the best for their children.
   In order to present this case, I’ve had to appeal to qualities of the parents that go beyond their role as parents – I’ve had to appeal to a political view. But is their enough within the account of the role of parent to rule out this sort of case? It depends on the details of our account of the role of parent. What my case shows, if anything, is that we may not be entitled to read off more concrete desires from the more abstract desires that seem uncontroversially to attach the the role of parent. However, we can avoid this problem entirely if we remove from the analysis of the concept of right the dependence on desire, and concentrate instead simply on the normative framework of roles and right bearers. The parents have a right to government sponsored child support because the legal structure of the family includes this entitlement for any person who occupies the role of parent and meets other appropriate conditions (perhaps including an income-based condition, and so on).

9. I am inclined to agree with Simon’s point about the crucial role that justification plays in rights.
   Regarding Colin’s question, my knowledge of the history of law is not strong enough for me to have a firm view. Which came first roles or rule? It hard to say. Historical records may not allow knowledge of the matter. There may not be a single answer. Different legal systems have different histories. Until I do a lot more reading, I’ll defer to folks like Colin on the historical issue.
   I have a more firm view that, whatever its history, the concept of rights is now tied to rules, not roles. In my post, I listed some of the key documents that take rights to be role-independent. Moreover, I hold the (controversial) view that there is moral progress in societies and that the notion of rights held independently of roles is a key advance. Thus some cases I see as paradigms are classified by Leif’s account odd balls.
   Nevertheless, I want to emphasize that I think Leif’s paper is great. A genuinely interesting and novel approach.

10. The contributors to the forum have been very generous with their attention and suggestions; I am very grateful to them. Arthur Ripstein especially has my gratitude for his penetrating, constructive contribution.
    Let’s warm up to discussing the deeper issues by exploring how the kind-desire theory handles the examples raised by the contributors to the forum. These are excellent examples; going through them brings out strengths of the theory that the article lacked the space to bring out.
    In at least one of the examples, when we attend more closely to where the normative system at issue contains a claim-right, we find the analysis locating a claim-right there. In Craig Nishimoto’s example of the snowy sidewalk, neighbors as such do not have a legal claim-right that the homeowner removes the snow from in front of his house. (Imagine a neighbor citing the shoveling statute, and then asserting that he himself has a right that the homeowner gets to it.) However, the homeowner does have a legal duty to pay compensation to those who are injured by slipping on his unshoveled sidewalk, and those so injured are assumed to want such payment. So the analysis yields the correct result: those injured by slipping have a claim-right for compensation against the homeowner.
    In other cases, we can just stick to rights close to hand instead of questing for rights more exotic. George Rainbolt describes the case of R, whose father rests in the Père Lachaise cemetery in Paris. R asks D to lay flowers at her father’s grave. D agrees to do so. R has a claim-right that D lay flowers at the grave. This is because D has made a promise to R: D has a duty to do what he promised to R to do.
    Both Nishimoto and Tom Dougherty wonder about third-party beneficiaries. As the article says on p. 213, the analysis finds that third parties have claim-rights when they are explicitly beneficiaries (as the British legal system, for example, affirms).
    In the other examples offered of third-party beneficiaries, the formalization gets the correct answers because the third parties are not explicit beneficiaries. The state has a duty to provide decent working conditions to teachers (condition 1), but students cannot have a claim-right to the state’s performance because they do not come into the specification of this duty. The state has a duty to provide decent working conditions for teachers, that is all. Students, and any others who benefit from teachers working in decent conditions, are not mentioned. So students (and others) have no right that teachers be provided with decent conditions.
    Dougherty also does not find a legal claim-right in children that their parents be provided with child support. The analysis produces the same answer, for the same reasons (mutatis mutandis) as we have just seen. The state has a duty to pay child support to parents, and since parents are assumed to want child support, parents have a right to it. The formalization does not allow us to sneak an extra entity (e.g., children) in condition 1 within the cloak of a possessive phrase (e.g. “the parents of children”). That is why the analysis gets the right answer.
    Pete Murray worries about the very assumption just stated: that parents are assumed to want child support payments from the state. What, he asks, if particular parents do not? Perhaps these parents hold libertarian political views, for example, and are actively hostile to state payments of any kind.
    As it happens, Tamar Meisels challenged kind-desire theory with nearly this same point during the conference in honor of HLA Hart in Cambridge in 2009. Meisels’s challenge resulted in the insertion into the article of the last paragraph on p. 216 (“The formalization shows…”) The response to the challenge is that the formalization gets the correct results when we remember which normative system we are working within. When considering legal rights like parents’ rights to child support, we must take up the perspective (as Hart would say) internal to the law. From this legal perspective, parents are assumed to want the support payments from the state. It may well be true that some individuals who are parents have very different preferences when considered from other points of view (when seen as voters, for example, or as activists). Yet when we stay within the relevant normative system, we get the correct result—parents have legal rights to child support payments.
    Simon Cabulea May (who has in years prior been tremendously generous in helping to develop kind-desire theory through acute suggestions) presents two good examples. (I confess the flaw of not having read the works of the philosopher Douglas Adams, so will have to ponder the normative system behind May’s third example at greater length.) May’s examples bring out two vital points.
    The first point is that the formalization is a precision instrument, and works best when used exactly as it is designed. Seemingly problematic cases often resolve themselves when disciplined by being put into its canonical form. The second point flows from the theme developed in the article that a good conceptual analysis will mirror uncertainty on one side of the biconditional with uncertainty on the other. When we find ourselves unwilling to ascribe a certain desire to members of a kind, we should find ourselves unwilling to ascribe a corresponding right to those kind-members—as May’s examples illustrate. (In the following discussion I will assume all the duties mentioned are enforceable, and so omit condition 3.)
    In the Bribery case, the correct formalization is:
    1) Denise (qua member of the public) has a legal duty not to offer a bribe to Oliver (qua police officer);
    2) Police officers (qua police officers) want not to be offered bribes.
    May does not believe that police officers have a claim-right not to be offered bribes, and the analysis pinpoints the source of his skepticism. The problem is that condition 2 looks very dubious. Who would affirm this proposition confidently? We are much more confident that police officers want more money, and this conviction already makes us feel tepid about 2). More, it may be that some police officers want members of the public to obey the law. Yet do police officers as such want this? It seems doubtful. More likely they are more like meter attendants, doing their job but not having further desires about the public’s compliance. In this example the analysis is working well: we hesitate to ascribe anti-bribe rights in police officers to just the same extent that we hesitate to ascribe anti-bribe desires to them.
    Custody Battle presents a case where we are reluctant to ascribe a desire to a role-bearer because of a competing desire held by such role-bearers. Here is the formalization:
    1) Judy (qua judge) has a duty to assign Paul (qua parent) the custody rights most beneficial to Paul’s child;
    2) Parents want to be assigned the custody rights most beneficial to their children.
    May does not find a claim-right in Paul, and again the analysis agrees, because of our resistance to the assertion in the second condition. As May notes, parents as such want custody of their children. The desire of a parent for custody of his or her children is strong. When this desire conflicts with the desire indicated in 2), we lose any confidence in affirming 2). This is an especially vivid case, of a type not emphasized in the article, where a conflict of desires within a certain kind of role-bearer makes a big difference to the rights that role-bearer has. Again the analysis shows how our reluctance to ascribe desires on one side of the biconditional matches our reluctance to ascribe a claim-right on the other.
    Kind-desire theory is a surprising theory, I know (it was certainly a surprise to me). But when one lives with it, and becomes accustomed to reading ordinary rights-assertions in its canonical form, its strengths reveal themselves more and more. I hope this is emerging from the discussion above, and I welcome more of what look to be hard cases.
    More soon on the contributors’ remarks on the formalization, and on the more philosophical dimensions of kind-desire theory.

11. Both Rainbolt and May offer friendly suggestions for modifying the first condition of the formalization. Above we have seen why such modifications are not needed for analyzing the examples that raised their concerns. Showing now how the proposed modifications would weaken the analysis will illustrate how tightly constructed the analytical framework is in its published form. Here is its condition 1:
    1. Some D (qua D) has a duty to phi some R (qua R) where “phi” is an verb phrase specifying an action, such as “pay benefits to,” “refrain from touching” “shoot,” and so on. (p. 18)
    Rainbolt’s friendly amendment is that R can be omitted entirely from this first condition. This, however, would be like removing the brake pedal from a sports car—there will be a great danger of that the analysis will speed out of all control.
    An example: Arizona employs a contractor to build a highway through the desert by a specified date. The contractor now has a duty to build the highway by that date, and with Rainbolt’s modification that statement would be the entirety of condition 1. Then in condition 2: small business owners close to the planned exits on the planned highway very much want the contractor to complete the highway by the specified date. But now we reach a false conclusion. We would not say that these small business owners have a right that the contractor completes the highway on time. And innumerable examples like this one can be generated—what we see here is like the interest theory’s problem with third-party beneficiaries on turbo. Without the R in the first condition, all of the R’s who (as R’s) want some duty to be completed will emerge as correlative rightholders. That way of designing the analysis will rack up rightholders much too fast.
    May, by contrast, proposed not to eliminate R from condition 1, but to tie R more tightly to D. “We could tighten the analysis a little by including a justificatory element,” he writes, “that D’s have the duty in part because R’s have the relevant desire.” (emphasis added)
    This proposal for changing condition 1 is like coupling the gas pedal to the brake pedal: it risks breaking the engine that makes the analysis go. If May is suggesting that a Raz-style ‘justificatory element’ be added to kind-desire theory, then the reply will point to the objections to Raz’s theory in the article. Yet May might be suggesting something different. Some older versions of the interest theory attempted to base their analysis on the rationale for the duty, or (as it used to be said) ‘the point of the norm.’ The difficulty of designing the analysis this way is that it is often quite difficult to say why duty-bearers have a duty. Take legal norms that assign a duty. Where do we go to find why that law assigns that duty? Perhaps to the legislative history? But there we often find that even the members of the legislative majority have voted for quite different reasons. The point of norms in games can be even harder: do opposing players have a duty not to obstruct goalies in part because goalies have a desire not to be obstructed? It’s tough to say. And this perplexity in the realm of political theory is often even greater. (Consider, e.g., Kant’s or Mill’s political theory on the question of whether duty-bearers have their duties in part because certain others, all of whom will turn out to be rightholders, have some desire.)
    If one looks very carefully over the blueprints of the kind-desire theory, one can see how the design of condition 1 accomplishes what the old ‘point of the norm’ analysis attempted. Condition 1 works with the current design, and does so because it does not tie D and R together with a justificatory element. Kind-desire analysis needs both the gas pedal and the brakes to find all and only the right rights—and these must operate as independently as the unmodified condition 1 allows them to.

12. Leif, thank you for your replies. In light of them, I think I should try to clarify my prior comment.
    In the snow-shoveling case, I imagined that the disabled person up the street would have a moral claim-right against me to shovel my snow, and that she would have this claim-right because of the allocation of individual responsibility for the public good of ice-free sidewalks. Although municipal fines enforce my individual responsibility, I suspect that the ordinance helps to determine or to clarify what I also have moral responsibility to do (perhaps by determining and clarifying my fair share of a collective responsibility for securing the relevant public good). Suppose that I do this moral reason to remove the snow on account of my responsibility for the public good. In such a case, the system of norms might still only support statements, both legal and moral, like this: “Homeowners, qua homeowners, have a duty to remove the snow in front of their properties.” Nevertheless, it seems to me that the system of norms can still yield moral claim-rights of particular individuals against against particular homeowners. The disabled person up the street seems to have a moral claim-right against me that I shovel the snow. (If, however, there were no allocation of individual responsibility for maintaining the sidewalks, then I doubt that she has any such a moral claim-right.)
    In the promising example, I meant to suggest that our system of norms surrounding promising might well support statements of a kind that would satisfy condition one in such a way that third-parties to a particular kind of promise might, by your formalization, have a claim-right against the promiser. Our system of norms surrounding promising would seem to support this statement: “The person who promises B to benefit C, (qua the person who promises B to benefit C), has a duty to benefit C (qua the person whom the promiser has promised to benefit).” Since C (qua the person whom the promiser has promised to benefit) presumably wants to be benefitted, C presumably wants the duty to be fulfilled. So, unless there’s trouble with condition three, this third-party to the promise should have a claim-right against the promiser to fulfill the promise. Counterintuitively, this would hold for our actual system of norms surrounding promising.

13. Thanks for your thoughtful response Leif. I don’t think I am persuaded by your reply to Bribery or Custody Battle, however. Here is a slight revision of the former example:
    Resisting Arrest
    Officer Oliver says he has to arrest Denise for offering him a bribe. She jumps out of the car and runs away, shouting “You’ll never take me alive, copper!” He pursues and eventually captures her after a short chase.
    1) Denise has a legal duty (qua member of the public) not to resist Oliver (qua police officer) in his act of arresting her.
    2) Oliver (qua police officer) wants Denise not to resist him in his act of arresting her.
    3) … appropriate enforceability …
    1) is surely true, since what Denise has now done compounds her legal problems. 2) is also surely true. If Oliver is trying to arrest Denise, he wants to arrest her, and ceteris paribus wants nothing to interfere with the successful execution of the arrest, including her running away. But my intuition is that Oliver has no legal claim-right that Denise not resist arrest (although he does have the legal claim-right that she not do so by inflicting violence on him). She has not wronged him by running away, even though she has wrongfully flouted a desire that we surely wish to attribute to him. At the very least, he has no more legal claim-right that Denise not resist arrest than the prison warden has a legal claim-right that the prisoner not escape from his jail. Oliver is also surely part of the content of Denise’s duty: members of the public do not have the legal duty to not resist disembodied acts of arrest, but rather the acts of arrest performed by *police officers*, or similarly empowered officials.
    More generally, if you find it dubious that police officers, as such, want members of the public to obey the law, is there any office that we could identify or define that does include the desire that members of the public obey the law? Perhaps judges are the most obvious example — it seems a little ad hoc to say that judges only want people to obey the law in cases that come before them, and not in other matters. But if we can identify or define some such office, then we can also surely then pass a law specifically prohibiting attempts to bribe those individuals who occupy that office. And yet I have a strong intuition that judges (or whoever) do not have a legal claim-right that members of the public not offer them bribes. (Or at least that if they do, this is because some or other much more complicated story has to be told about how offering a bribe to a judge could seriously set back her interest in being seen as honest and trustworthy.)
    Here is another case structurally similar to Resisting Arrest:
    Compulsory Vaccination
    An epidemic of Serious Disease sweeps the country. Fortunately there is a vaccine. The government orders compulsory vaccination of all people in area X, with fines imposed on people who refuse to let special certified anti-epidemic medics inject them with the vaccine.
    1) Nervous Norris has a legal duty (qua person in area X) to let Medic Maddy (qua certified anti-epidemic medic) inject him with the vaccine.
    2) Medic Maddy (qua certified anti-epidemic medic) wants Nervous Norris to let her inject him with the vaccine.
    3) … appropriate enforceability …
    But Maddy has no legal claim-right that Norris not refuse to be vaccinated. She is frustrated by his nervous wriggling, but not ipso facto wronged by him.
    What divides us here is that I think an individual’s claim-rights are necessarily grounded, in some way, in part, either directly or indirectly, in the non-instrumental importance of her interests, whereas we can ascribe kind-desires to individuals that either have nothing to do with their interests or are not relevant as interests. Once we identify such kind-desires, we can start to gerrymander examples that satisfy the other conditions of the theory. In both Resisting Arrest and Compulsory Vaccination, the interests of the police officer and medic are not, in themselves, any (non-instrumental) part of the reason for the duty in question. I think it is this that really explains our reluctance to ascribe claim-rights to Oliver and Maddy. (I should add that I think reports of the demise of a Raz-style justificatory interest theory have been greatly exaggerated — I tried to show how at least some counter-examples might be avoided at the end of my paper on Sreenivasan’s hybrid theory.)

14. Let me say a bit about other parts of the discussions raised above before returning to the examples. Rainbolt asks whether the kind-desire theory is just a version of the interest theory. No – the interest theory is a broke-back version of the kind-desire theory. ‘Interest’ only appears to be the right concept when one squints together two different senses of ‘interest’: what Parfit would call the desire-satisfaction sense, and what he would call the objective-list sense. Kind-desire theory naturally spans both senses. Insofar as it seems appropriate to link claim-rights and interests, this is only because we are sensing kind-desires at a deeper level.
    To push this point through, we can remember that the interest theory construed strictly (in terms of what objectively makes a person’s life go better) is plagued by counterexamples. Individuals have many claim-rights to things that are not in their interests. The mechanic has a right to delivery of the engine that he bought, even though (as he doesn’t realize) he can’t use it and it will cost a lot to get rid of it. Neil MacCormick makes the same point for the case of property owners thus: “Some hereditates may be damnosae.” Having a right is not necessarily in an individual’s interests (it doesn’t necessarily make the individual’s life go better).
    Even if we move the interest theory from individuals to roles or kinds, the interest theory goes wrong. The examples of the journalist who has a right against having his notes searched, and the parent who has a right to receive child support, show that the rights of role-bearers cannot rest on interests of those role-bearers sufficient to justify those rights. This objection has now been a standing challenge to the main (Razian) version of the interest theory for several years, and to my knowledge it has not been answered. (Matthew Kramer has promised a version of the interest theory that does not turn on the rightholder’s interests being sufficient to ground the duty-bearer’s duty. Yet for reasons set out on p. 205 of the article, I suspect that Kramer will have difficulties in constraining the extension of his theory in non-rigged ways.)

15. After we sublate the interest theory, some may still question whether ‘desire’ is the right concept for our analysis. Tom Dougherty worries that analysis in terms of kind-desires will be too restrictive, and not allow us to ascribe rights to beings whose desires are incoherent, weak, or even absent. This concern is valid, even after we remind ourselves that the analysis turns on the desires of kinds, not on the desires of any extant individuals. (The analysis “is not sensitive, as it might appear to psychological states.” p. 209)
    On the analysis, whether young children have moral rights against being abused will depend on whether young children have reason to want not to be abused. It seems natural to say that children do indeed have such reasons, and similarly with comatose adults. Even a person with severe brain damage might plausibly be said to have reason to want not to be abused, if being abused might make his condition worse. If we move to considering a human with genuinely irreparable brain damage, we have reached a hard case. Some might begin to lose their conviction that this human indeed can possibly have such reasons and such rights. (They might think that only persons, not human beings as such, have such rights—on which more below.) Others might argue rather that it is the person in their earlier, undamaged life that has the right that their later, damaged self not be abused. (The structure of the debate around this will be like the familiar debate about whether the dead can have rights.) In the debate over this hard case, the analysis shows its strengths. We disagree about the desire-attribution in the hard case, so as the analysis predicts we disagree about the corresponding rights. The analysis “mirrors uncertainty on one side of the biconditional with uncertainty on the other, as a good conceptual analysis should.” (p. 223)

16. ‘Desire’ is a more felicitous concept than ‘interest’ for the analysis of claim-rights. And I continue to believe strongly that the desires of role-bearers, and of kind-members more broadly, are basic in our making sense of rights assertions. Even I admit, however, that the idea of a kind-desire is easy to misunderstand, which is one place where the presentation of the analytical framework can still be improved. This is why I am very grateful to Arthur Ripstein for his suggestion of a supplemental concept within this part of the analytical framework. Ripstein speaks of the ‘constitutive ends’ of role-bearers, and (in an alternative formulation) their ‘characteristic ends.’
    This may be a real advance in the presentation of the theory. Speaking of the constitutive or characteristic ends of journalists, parents, goalies, humans and so on appears to be quite natural, and it appears to fit the various types of kind-desires catalogued in the article (duty-derived, stipulative, implanted-by-god, kind-normative). It may be that using these locutions can eliminate some of the presentational hurdles with kind-desires that have tripped up more than one reader. Those who think kind-desire theory is worth exploring might consider whether Ripstein’s phrases might indeed be more than supplemental—whether it should come first within the presentation of the central concepts of the analysis. The main question about Ripstein’s phrases is whether they really are apt for the analysis of all claim-rights (for example, those of the property-owner and, especially, of the promisee). Still, Ripstein’s suggestion opens a line of inquiry well worth pursuing.

17. Turning now from the concept of desire to the concepts of role and kind. Rainbolt says that “Rules are more fundamental than roles. Roles are created by rules, not the other way around.” As far as the historical debate is concerned I am much encouraged by Colin Heydt’s remarks. It is indeed exciting, as he says, to see conceptual analysis and legal history converging in this area, and one could only hope for more historical scholarship to help us deepen our understanding of the connections between roles and rights.
    On the conceptual point about whether rules or roles have priority, it may be that in many cases the question is moot (which came first: the pieces of chess or the rules of chess?). All I would want to emphasize is that roles are quite deep in our conceptualizations of the social world—often as deep as our sense of our own and others’ identities. And, crucially, roles can persist through changes in rules: we can still be the same ‘players’ even when the game changes. Catfish is still a baseball pitcher—only now he doesn’t have to bat. Sheila is still your secretary—only now you must not harass her. I suspect that in our cognition, at least, what some being is—its role or natural kind—comes first, and the rules by which it properly acts and is properly treated come after.

18. Arthur Ripstein raises questions about the analysis of rights that go very deep. Since Ripstein has his own project in this area, we can look forward to the further development of his account and to considering how far kind-desire theory is compatible with it.
    In his comment, Ripstein posits that the extension of the article’s analytic framework from roles to kinds may be unnecessary. ‘Human being’ might be a role, and more generally (as he says) it may be that roles are always available as the load-bearing term in the analysis.
    One question that Ripstein’s suggestion raises is how to draw a line between social kinds (like roles) and natural kinds. As the philosophical literature shows, this question is far from trivial. My own sense is that in order to be a role-bearer within a certain normative system, a being must act with at least latent awareness that it is acting within that normative system. But clearly there is more to be said on this topic.
    A second question is whether human beings are really the bearers of human rights, or whether we should rather say (as a reviewer for the journal suggested) that human rights are instead held by persons. This again pushes forward the question of whether ‘person’ is a natural kind or rather a social kind. If it is the latter then, as Ripstein makes clear, it is not a ‘constructed’ personae in a sense that carries with it individual (or perhaps even collective) discretion over how the role is to be understood.
    It is of course in one way indifferent to kind-desire theory if Ripstein is correct that roles are always available for the analysis. Roles are kinds, so the theory would still be universally applicable. I was convinced to expand the role-based theory to include kinds by Jerry Gaus at the APA Eastern in 2007, for some of the reasons that Rainbolt cites in his remarks. If that extension to kinds proves unnecessary, then the theory can again take a more compact form.
    However, one challenge that Ripstein’s suggestion faces will be to find a role-based analysis that always locates the right rightholder. Consider for example what Ripstein says about animals in the livestock example. As noted above, I believe that in order to occupy a role a being must act with at least latent consciousness of the relevant normative system, which would rule (at least most) animals out as rightholders right away. Putting that point aside, and granting that animals can be role-bearers, why would they be rightholders?
    In Ripstein’s construal of the livestock example, there is a manager who enforces the rule against abuse of the animal on behalf of the animals. The challenge in this reading is to come up with a general analysis which shows that it is indeed the animals–instead of the manager–who is here the rightholder. If that sounds like a peculiar challenge, recall that this challenge arises for the will theory that has inherited the Kantian tradition of thinking about rights. In Hillel Steiner’s will theory, for example, citizens do not have a right against being assaulted on the streets. It is, rather, the district attorney who has rights that citizens not be assaulted on the streets (since the DA has the power to enforce the relevant duties). This is a weakness in the will theory insofar as it aims for fit with ordinary usage. I suspect that Ripstein will need to attend to the parallel challenge of finding the right rightholder (the animals, not the manager) as he sets out his general account of what rights are. Granting that both the manager and the animals have roles, why is ‘animal’ here the (or at least a) role that carries with it the right that the animals not be abused?

19. (My apologies to Simon Cabulea May – I’d not seen the article he’d mentioned before posting one of the messages above. I’m giving a series of lectures in Sicily this week, grabbing internet access where I can. Will return to the new examples once I get back to my desk. Many thanks in the meantime to all of the contributors for their careful scrutiny and generous suggestions.)

20. Thanks for the replies Leif.
    I’m not so sure it is natural to say that new-born babies have reasons for desires; I think it’s more natural to say that only agents with developed psychologies have reasons for their mental attitudes. We may have different intuitions here.
    Thanks for clarifying about third-party beneficiaries. I understand better now how the formalisation is intended to work and my earlier comment was confused on this point. You say that “students cannot have a claim-right to the state’s performance because they do not come into the specification of this duty.” I think this brings out that before formalising the norm, some prior work is needed in deciding how to specify the duty. Suppose (no doubt artificially) that the legislation puts the parent’s right to child support this way:
    “The state has a duty to make it the case that if any person is a parent of a child, then this person receives financial support if she so wants.”
    The legislation mentions two roles / kinds who have reason to want the duty enforced. To apply the formalisation in the way that you intend, we would need to know how to identify the parent as the relevant role / kind for the “R” role. Given that both the parent and child have reasons to want the duty to be performed, how are we meant to select the parent in a non-ad hoc way?
    It’s true that in this example English has a handy verb-phrase for formulating the duty as:
    (1) The state has a duty to provide child support to any parent
    By contrast, we’d have to “cook up” an awkward verb-phrase for formulating the duty the other way:
    (2) The state has a duty to provide a financially supported parent to any child
    But this is simply an artifact of the contingencies of the English language. An account of the concept of a right should presumably generalise over different languages. And there’s nothing to stop us coining a new term to mean “provide a financially supported parent to” and using this term widely so it became familiar to everyone.

21. I’d like to join the discussion by saying that Wenar’s approach seems to me especially promising because of how it handles socially created rights such as property rights, rights in games, legal rights – as opposed to uncreated ‘natural’ rights. For the latter, Simon Cabulea May’s insistence that “an individual’s claim-rights are necessarily grounded, in some way, in part, either directly or indirectly, in the non-instrumental importance of her interests” seems correct. But for the former it seems doubtful, given the many purposes for which we can create systems of rights, and the ensuing wide range of examples (of which interest theorists like Kramer, MacCormick and Raz are aware, of course) of rights that are distant from their holders’ interests. Wenar’s role-based approach seems, to me, to be a big advance by capturing the sense in which rights purport to serve their holders (the sense in which there is some description under which they necessarily serve their holders) while detaching this from the holders’ well-being and psychology, and from the rights’ justification.
    But May’s examples, and the others above, put pressure on this. A further pressure comes from the fact that if we allow something to be my ‘right’ when its justification is independent of what it does for me, then it is unclear why violating the right disrespects me. My worry, though, is that if we accept May’s point and return to an extended Razian theory of the type outlined in May’s article, then we will be compelled to deny that many property rights and rights in games are genuine rights, because it seems to me that many of these rights are justified entirely on common good grounds that are independent of what they do for the right-holder. Of course, some will deny this point and find Razian individualistic grounds for the rights in question, but it seems odd to make whether something really is a property ‘right’ (as opposed to a set of duties with no correlative right) depend on debates about the justification or grounding of property. If a property system involves duties of the type within which we live now, then I’m tempted to think it must contain property *rights* no matter what justifies the system.
    One more point: I share the worry about whether ‘desires’ is the right concept. Ripstein’s ‘characteristic ends’ is suggestive but not quite correct because, as Wenar points out above, there don’t seem to be characteristic ends for owners or promisees or, I would add, modern citizens. One option mentioned to me when I was discussing Wenar’s paper recently was a role-bearer’s legitimate expectations: we would say a duty was owed to whoever, in or because of their role, legitimately expects the duty fulfilled. But this errs because, for example, criminals qua criminals legitimately expect to be pursued by the police, but the police’s duties of pursuit are not owed to them. The claim-right-holder must gain something ‘positive’ from the duty. Should we stick with (suitably role-qualified) desire or interest?

22. Many thanks to Leif Wenar for his long meditated, rich and ambitious article on claim- rights. Here are a few rough thoughts that the article provoked on a first reading, with a few headings to separate out separate points.
    The aim
    I take Leif’s aim to be an analysis of the concept of a claim-right as this is expressed in the claim-rights assertions that we make. An obvious question, then, is what counts as a claim-right assertion for purposes of the pursuit of the aim and evaluation of the result.
    Leif writes on page six “A person holding a claim-right is owed a duty by some other person(s)” From this one might think that the claim-rights assertions that we make are those in which someone is said to be “owed a duty” by some other person(s). This seems a bit odd. For one thing, one would think that the most indubitable claim-rights assertions that we make would involve the word “right” as in “Jane has a right to your cleaning up.” For another, the phrase “owing a duty” is a on the technical side. Talk of “owing” may be closer to the vernacular, but then the term “duty” does not appear and cannot, perhaps, be assumed to be implicated. Perhaps Leif can clarify his sense of the data here.
    Its achievability
    For the sake of argument I’ll assume that the aim of an analysis of claim-rights, according to Leif, is to offer a set of conditions that are individually necessary and jointly sufficient for the truth of assertions such as “Jane has a right to your cleaning up”. I do not mean to impugn this aim; something like it is, as Leif indicates, common among rights theorists. Nonetheless, it is not obvious that it is achievable.
    Not to make too confident a statement here, it is quite possible that claim-rights talk—along with “owing” talk and “duty to” talk—is ambiguous. The different meanings in question could be related in various ways but nonetheless distinct. Again, we could be dealing with a single family resemblance concept, along the lines indicated by Wittgenstein, such that there is no set of individually necessary and jointly sufficient conditions for the applicability of the concept.
    So, should Leif or a will or interest or other theorist of claim-rights not succeed in realizing the aim just specified, or something similar, they will not necessarily have failed to achieve the achievable. Rather, the enterprise itself may have been doomed from the start. Still, there’s no harm in trying, and it remains to consider Leif’s theory in light of his aim—as I am construing it. In the rest of this comment I make a few points related to this.
    Some cases from moral theory
    There are cases commonly offered in moral theory as cases that involve duties but not claim-rights, where the kind-desire theory may say that there are such rights. I will simply describe the cases for Leif to clarify his position with regard to them, should he wish to.
    A. Rescue
    Anne is a non-swimmer who will drown unless rescued from the deep pond she has fallen into. Ben is a champion swimmer, dressed for swimming, without much else to do, who ho finds himself on the side of the pond at this time. He notices Anne floundering about helplessly and pauses a moment. He has a moral duty to rescue Anne. Chris, a passer-by who does not swim would be justified in attempting to enforce this duty at least to the extent of putting pressure on Ben to get moving.
    B. Gratitude
    Don has done a tremendous favor to Ellen. She has a duty to express gratitude to Don in some way. Her friends and family, at least, would be justified in urging her to do this.
    Promising
    Sometimes in discussing promising Leif is clearly talking about promises or contracts in law. Promises can, however, be made without any intent to create legal relations and, in that case, promisees have rights—non-legal rights. My points about promises here relate to cases of the latter type.
    My first point follows up a thread started by Craig Ishimoto. Competent speakers would presumably say that if Jack promises Jill that he will buy some milk at the store this evening, then Jill has a right to Jack’s going to the store this evening. Equivalently, Jack has a duty to Jill to go to the store this evening—whatever precisely that means. As to the content of this duty, it can plausibly be represented as a duty to phi, not as a duty to phi Jill. Leif may reply that Jack only has the specified duty to phi because he has a duty that can—perhaps a little awkwardly—be represented as a duty to phi Jill: the duty to fulfill his promise to Jill. Still, it would be good for a theory of claim-rights somehow to mark the fact that many promisors are in Jack’s situation: they have duties that are not representable as duties to phi their promisee, and their promisees have the correlative rights.
    Next: what does a promisee qua promisee want? Leif supposes that promisees as such want the fulfillment of the promise. I am not sure why one should buy this supposition. It seems to me to be a contingent fact that some promisee wants, or ever wanted, fulfillment of the promise. Presumably the promisee wanted, for some reason, to accept the promise. In other terms, the promisee wanted the promise. That does not mean the promisee wanted, or wants it fulfilled. Leif’s theory may fare well if the promisee qua promisee wants fulfillment of the promise, but that does not mean that promisees qua promisees do want this. It seems to me that promisees qua promisees do not want anything in particular, or have any particular constitutive end. That they have rights, then, cannot be explained by reference to any such want or end.
    Promisees’ rights and Wenar’s general theory
    Somewhat following Hart in “Are There Any Natural Rights?” (1955), Feinberg in “The Nature and Value of Rights” (1970) and others, I have argued in various places that the right of a promisee is such that the promisee has the standing to demand the action he has a right to from the promisor and to rebuke him for not performing that action after the time for performing is past.
    I take there to be an important class of claim-rights here such that one has such a right if and only if one has the standing to demand the action one has a right to. For further elaboration see e.g. my “Three Dogmas about Promising” (2011) or, with less focus on promising, “Giving Claim-Rights their Due” (2012). Let us suppose this is correct. Let us refer to the species of claim-rights in question as “demand-rights” just to keep things clear.
    The question arises: what is the relation of demand-rights to claim-rights according to Leif’s kind-desire theory—“kind-desire rights” for present purposes? One possibility is that demand-rights are a proper subset of kind-desire rights. Another is that they are not kind-desire rights, perhaps because they lack one or more of the feature necessary for kind-desire rights. I take this question be a large and important one, an important test case not only for kind-desire theory but for any other general theory of claim-rights. If the theory cannot encompass demand-rights it is not a truly general theory.
    On the face of it, demand-rights are not kind-desire rights. Not if one accepts, as I suggested earlier, that promisees as such don’t want anything in particular: they are just promisees—people who have by their acceptance co-created a promise with their promisors. Since wanting something as a member of some kind is essential to the having of a kind-desire right, promisees’ rights appear not to be kind-desire rights, whatever else is true of them. If this is so then a central kind of claim-right falls outside Wenar’s theory, however capacious it may be.
    History
    I think it most likely, by the way, that our perception and responsiveness to demand-rights, at least, antedates such sophisticated legal systems as are often deemed to be the cradle of rights thinking.

23. Taking up the new examples, before going on to the larger themes:
    Resisting Arrest and Compulsory Vaccination. In the 2005 article “The Nature of Rights” (p. 234, here) I noted in a footnote that there are a few seemingly active verbs that are in fact “false friends.” For example, the verb “to inherit” appears to be an active verb, yet in fact one can inherit without performing any action. (A baby, for example, can inherit an estate while still in the womb). In setting up the analysis of a rights-assertion, a sentence with a falsely active verb should be rephrased to bring out the true agent. So “The infant has a right to inherit the estate” should be rephrased as “The infant has a right that the executor pass the estate to him.” (In the 2005 article, this rephrasing from a falsely active to a truly active verb reveals that the Hohfeldian incident at stake in this assertion is not a Hohfeldian privilege or a power, as it might first appear, but in fact a Hohfeldian claim.)
    May has uncovered a related set of “false friend” verbs with his latest two examples. The verbs in his two examples are, “not to resist the police officer in his act of arresting her,” and “to let the medic inject him with the vaccine.” These verbs are false friends because their primary activity is not performed by the subject of the verb, but rather by a second party (the police officer, and the medic, respectively). The subject of these verbs remains passive as the second party does something to her or him. In this set of false friend verbs, the subject of the verb is passive as the primary actor acts upon him or her.
    Observe that these false friends are not defined by containing a negation (“not to touch” is not a false friend). Nor is this set of verbs defined by their reference to an action of a second party (“follow the orders of” is not a false friend). It is the passivity with respect to the action of another that gives these verbs the wrong shape to fit into “verb phrase specifying an action” in the kind-desire formalization.
    This may seem like a technical point, yet in fact it provides another perspective on the structure of the relation between duty-bearers and right-holders that the article’s analysis reveals. Claim-rights concern what duty-bearers, D’s, must do. The verbs in May’s set are false because they primarily refer not to what D’s do, but rather to what R’s do to D’s. In this set of verbs, the active shoe is on the wrong foot. It is an interesting feature of English that it contains such verbs that so generously point to the actions of a second party, but we should not allow these false friends to lead us astray.
    May is quite correct to bring up such cases in his comment, since the current article does not contain a footnote equivalent to the one in the 2005 paper that shows how to identify these false friend verbs. Future presentations of kind-desire theory will include such a note, and I am grateful to May for highlighting such excellent examples of verbs within the relevant set.

24. Third Party Beneficiary. In his follow-on comment, Nishimoto explains that his third-party beneficiary example was not meant to be an example within a legal system, but rather within “our system of norms surrounding promising,” which most would identify as a system of moral norms. This move to moral norms gives occasion to remark on a feature of the article’s analytical framework not yet mentioned: its identification of a universe of discourse.
    The role-theoretic formalization, for example, begins: “Consider a system of norms S that refers to entities under descriptions that are roles, D and R…” The system of norms in question must actually refer to entities as role-bearers for those entities to be potential right-holders within that system. The article notes that the legal systems that contain third-party beneficiary rights do explicitly refer to entities under the description of the role “third-party beneficiary.” However, I suspect that Nishimoto will find no such role within our system of moral norms concerning promising. Our moral system contains the roles of “promisor” and “promisee,” but lacks the role of “third party beneficiary.” I conjecture that it is Nishimoto’s coming up empty while trying to locate such a role-bearer within our system of moral norms that explains his not finding a claim-right in this location.
    Snow Removal. Nishimoto’s clarification of this example shows that what is at stake in the homeowner’s snow shoveling is a duty to do one’s fair share in the provision of a public good. Insofar as it seems to me that there is indeed a moral right at stake in the clarified example, I agree that Hart’s Principle of Fairness (or an improved variant) is the best explanation. (To get Nishimoto’s example to work, we can postulate that the disabled neighbor pays to have the snow in front of her own home shoveled, and so does her fair share in the provision of the public good instead of being a free rider herself.)
    Interestingly, Hart formulated his Principle of Fairness in terms of claim-rights: “When a number of persons conduct any join enterprise according to rules and thus restrict their liberty, those who have submitted to these restrictions when required have a right to a similar submission from those who have benefited by their submission” (“Are There Any Natural Rights,” (1955), p. 185.) And Rawls, in his discussion of the principle in TJ §18, makes explicit that he believes that the obligations generated by this principle are owed to definite individuals (they are directed duties), again supporting the idea that the principle generates claim-rights (though note that Rawls’s discussion on TJ p. 113 is in fact rather nuanced).
    Whether the Principle of Fairness states a valid moral principle is famously controversial. However, from the perspective of kind-desire analysis it is clear how the Principle of Fairness generates claim-rights within any normative system of which it forms a part. The key to the analysis here is that those engaged in the joint enterprise appear in condition 1, so are eligible to be the right-holders in the conclusion. The duty of the potential ‘free rider’ is to do what the ‘paid riders’ have already done.
    The kind-desire analysis of Hart’s principle will run: define “Paid Riders” as “those who submit to the restrictions of a joint enterprise when required” and “Beneficiaries” as “those who have benefitted from the submission of Paid Riders.” Then we get: When Paid Riders have conducted a joint enterprise, (1) Beneficiaries have a moral duty to do what Paid Riders have already done, and (2) Paid Riders want Beneficiaries to fulfill this duty. Adding (3) enforceability completes the kind-desire analysis of why Hart’s Principle of Fairness generates a moral right—in the case at hand, the moral right of the neighbors that Nishimoto has detected to the homeowner’s snow-shoveling.
    Since the Principle of Fairness is on the table, I might mention that someone might wish to deploy it to explain the justification of some of the “socially created” claim-rights that concern Rowan Cruft—such as the claim-rights within property rights. Whether this is the correct way to explain such claim-rights is an engaging question. Yet it is a question in substantive normative theory, and so it is in a different domain than the conceptual analysis of claim-rights. The conceptual analyst says: ‘Give me any normative system described without claim-rights, and I will show you where the claim-rights are.’ Whether the Principle of Fairness is part of, say, the system of genuinely valid moral norms is to the analyst irrelevant. The system of valid moral norms is merely one more system that he can (if he has a good theory) find the claim-rights within.
    A related point applies to Cruft’s mention of disrespect. If it is correct that violating a person’s moral rights always disrespects that person, then this is a fact to be explained by a substantive moral theory. If there is indeed a correlation between moral rights violations and disrespect, the conceptual analysis of rights will not account for this all by itself—although of course the best substantive moral theory is likely to draw on the best conceptual analysis when setting out the full explanation of a connection between rights and respect.

25. Thanks for the replies, Leif. It is interesting to see how much work is being done by the first (‘doing something to someone’) clause.
    I’m inclined to agree with Wenar on the distinction between the analysis and the moral justification of rights, though I also see the appeal of the Razian position May was pushing which (at least on one natural interpretation of Raz’s notion of ‘an interest that grounds a duty’) builds moral justification into the analysis. I have a paper of my own (draft here: http://www.aristoteliansociety.org.uk/pdf/cruft.pdf) which starts with the thesis that it is morally disrespectful to a person to violate any morally justified right she holds, even when that right is not justified in any sense by what it does for her. (Thus, I suggest, it is disrespectful to you to paint graffiti on your barn even though your ownership of the barn is not justified by what it does for you but rather by common good considerations in which you need play no part.) This general link between rights-violation and disrespect is difficult to explain. Raz’s account, which excludes something from being a right unless it is justified by what it does for its holder, would make this thesis much more easily explicable, but it would exclude too many candidate rights.
    Note that the Principle of Fairness could be understood in a way that delivers Raz-style accounts of the rights it justifies. According to this understanding, the Principle of Fairness would tell us that Beneficiaries’ duty to make their contribution is grounded in or justified by Paid Riders’ interests. I think any right that fits Raz’s theory will also fit Wenar’s, for whenever someone’s interest is sufficient to ground someone else’s duty, the former person will have an appropriate role- or kind-based desire for the duty’s fulfilment. But by breaking the link with justification, Wenar’s theory – as he notes – encompasses more than Raz’s. For instance, the just-sketched Fairness-based approach is implausible as a justification of property rights: while theft is perhaps unfair on ‘law-abiding’ owners, that is not the fundamental explanation for why theft wrongs its victims, and property rights are morally binding even on those who do not benefit from a property system. Instead, property rights cannot be made to fit Raz’s theory, though Wenar’s accommodates them happily.
    Gilbert’s point about demand theories is an important one. In assembling the symposium for Ethics, I was aware of it as a bit of an omission. I think there were two reasons for this absence. First, the symposium focuses primarily on directed duties, and many directed duties do not give their recipients standing to demand their fulfilment. Consider Gilbert’s example of gratitude: I have no standing to demand that you thank me for helping you cross the road, even though you owe me a duty of gratitude. I’m attracted to the thought that standing to demand is something that one adds to a directed duty in order to turn it into a right. Compare Wenar’s suggestive similar idea – focused on enforcement rather than standing-to-demand – that “enforceability turns directed duties into rights” (‘The Nature of Claim-Rights’, footnote 24). If this is correct, then demand-rights will not be simply a subset of claim-rights as Gilbert suggests; rather, all claim-*rights* will be demandable, but not all directed duties will. Secondly, it is I think very difficult to give an account of a distinctive form of demandability borne by right-holders and not others, even though many try this approach (Gilbert mentions Hart and Feinberg; we could add Darwall, Harel, Skorupski). Third parties can very often permissibly demand that rights be respected by correlative duty-bearers. For example, I can permissibly demand that someone not assault you, or that they not steal your property, or that they provide appropriate care for their children. Is there a distinctive form of demand made available specifically by one’s being a right-holder?

26. I appreciate Wenar’s generosity in helping me to better understand his proposal. I overlooked his stipulation that “R” must be a description already used in S (that the bearer of the claim-right must fall under a kind-description that is already used in the system of norms). This stipulation prevents Wenar’s view from generating claim-rights for third-party beneficiaries of promises even for the special promise I imagined, where the promise is simply “to benefit” some third party (or to otherwise satisfy some kind-based desire of some third-party).
    We can adjust the case of the special promise so that it satisfies Wenar’s stipulation about S containing “R”. Wenar accepts that our system of moral norms surrounding promising contains the kind-description “promisee.” So suppose that A makes a promise to B that she will benefit every promisee. On account of A’s promise to B, the moral norms of promising now support this statement: “A has a duty to benefit some promisee, as promisee—namely person C.” Let this be the statement that satisfies condition one, and let C be some promisee other than B. If promisees, as promisees, desire to be benefitted, then Wenar must apparently say that C has a claim-right against A to fulfill A’s promise to B (assuming such desire fulfills condition 2, and condition 3 is also met). That we can in this way meet Wenar’s stipulation suggests to me that one shouldn’t rely upon it to block counterexamples.
    This suspicion increases when I consider how we are to determine whether a given system of norms already uses a given kind-description. Perhaps a system of norms uses a kind-description iff an adequate description of the system requires us to employ that kind-description, perhaps because we must refer to someone under that kind-description. It seems plausible to me that, for any system containing claim-rights, an adequate description of that system would require using a description of the category of persons who hold the claim-rights. So we might therefore expect that any system of norms that contains a given claim-right must also use the description of the category of persons who hold such a right. If such categories are always kinds (if claim-rights attach to kinds), then Wenar’s stipulation will be fitting. However, the order of explanation may be different from what Wenar’s proposal seems to suggest. It may not be that the person has a claim-right because the system of norms happens to employ the relevant kind-description; rather, it may be that we are willing to concede that the system employs the relevant kind-description simply because we intuitively think that the system ascribes rights to persons of that kind.
    In the snow-shoveling case I see how condition one might be met under Hart’s Principle of Fairness if the disabled neighbor is a paid-rider. Since I suspect that the disabled neighbor would have the claim-right even if she where an eight year-old child, I think it would have to be easy to qualify as a paid-rider. Thoughts in this direction lead me to think that every community member is a paid-rider so long as he/she hasn’t unacceptably violated any of his/her the public-good related requirements. Instead of describing the relevant kind as “paid-rider” we might rather say “community member in good standing.” The eight-year-old child needn’t have done anything. If this is so, then we would need a different statement to satisfy condition one. Perhaps we can still formulate a new phi phrase that suitably relates the community member to homeowner with respect to the homeowner’s duty to remove snow, but it isn’t obvious what it should be. I am concerned that the task of formulating the needed phi phrase seems to require the kind of creative process that could just as well be used to over-generate statements satisfying condition one.
    So I am still puzzled by condition one. Condition one requires that the system of norms supports statements of the duty that suitably specify the holder of the claim-right. Whether or not the right-holder is suitably specified is sometimes difficult to determine, and this difficulty isn’t always reflected in any uncertainty about whether or not there is such a right-holder. Moreover, the question of whether or not a person is suitably specified in condition one doesn’t seem to resonate with the significance of whether or not the person has a claim-right. If the question of claim-rights comes down to the question of whether or not the system of norms supports statements that satisfy condition one (with regard to the possible right-holder), should we really care about whether or not a given person has the claim-right?

27. The analysis of rights is hard; it has defeated some of our best minds. Who now accepts Bentham’s account of what rights are? Or Mill’s? In our day, Hart’s version of the will theory is known to suffer several serious defects, and Scanlon’s constraint theory from at least as many (on Scanlon’s view, see here).
    No doubt there are different reasons that these excellent theorists have struggled with this analysis. One reason may be that they (like all of us) are subject to a kind of occupational hazard. A theorist can spend his days meditating on a certain class of rights (legal rights, say, or human rights, or speech rights). After long study, those rights become familiar, more deeply known. He can begin to think that the objects of his study are especially important, even exemplars, of all rights. He tailors his conceptual analysis of rights to fit his class, and when presented with rights that burst through its strictures he defends his analysis by saying that his special class defines the “rational kernel” of the concept, or that these are the “paradigm” of what all rights are — and perhaps also by saying that the counter-examples do not even offer “real” rights at all.
    The difficulty, of course, is that over the years different theorists have fixated on quite different classes of rights, resulting in a potpourri of quite varied characterisations of “what rights really are.” (A collection of examples is here.) Each theorist has his own paradigm, none can say on independent grounds why his favored class should be thought the real paradigm. Once theorists develop this kind tunnel vision, their debates over rights tend to become exchanged salvos of counter-examples. Since all of the participants to the debates are focused on a limited class of rights, most of these counter-examples hit their targets. The debate turns into a contest over which theory can grasp the prize of being the least counter-intuitive. (More on this theme is here.)

28. I suspect that both George Rainbolt and Margaret Gilbert are in danger of succumbing to this occupational hazard. Each of them has a favored class of rights, each tries to suggest that their favored class of rights is distinctive, especially important, or even the paradigm of rights. If they continue down this path, they may begin to urge us to think that certain rights are in the “core” of the concept, and to turn us away from other rights as more “peripheral,” and so on. We have seen this move made many times, over decades and even over centuries. Up to now this move has not led to success in any of its many variants, and we might demur when offered the chance to go down such a path yet again.
    Rainbolt, for example, is especially concerned with a certain class of rights, which the kind-desire article identifies the “latecomer” rights of detheologized human beings. Rainbolt says, “After the American Revolution, the French Revolution, the Emancipation Proclamation, Seneca Falls, the Universal Declaration of Human Rights, I Have a Dream, and Stonewall these latecomers are now paradigms.”
    It may be that certain of the rights that Rainbolt is pointing to here are less what the French Declaration would call the rights of man, and more what it would call the rights of citizens. Still, taking this class as Rainbolt describes it, we can agree wholeheartedly that it contains extraordinarily important rights. As moral and political theorists we might spend our careers studying such rights: human rights, women’s’ rights, minority rights, gay rights, and more. We can devote our working lives to exploring the content and justification of such rights, and we may even try, in our way, to get our fellow men and citizens to better appreciate what these rights require. Such rights are crucial if we are to have decent societies in modern times. They are not, however, the paradigms of rights. They are just the rights that we moral and political theorists often think the most about while engaged in our own peculiar occupation.
    When we open our ears to what our fellow reasoners say, and to what we ourselves say, in everyday life we hear an enormous variety of rights-assertions. The best analysis of rights will accept from its inception this enormous variety as defining the phenomenon to be explained. We are looking for a conceptual analysis that can capture how competent speakers use the concept of a right in all of the sundry contexts in which they do. We ask: what is a right, such that all of these myriad rights-assertions make sense to people like us?

29. On the first day of this forum opened I did an internet search of news sources for the term “right.” Below is a random selection of what I found: examples of how competent language-users are using this concept around us right now. First we see two examples that appear to be pointing directly to a claim-right:
    “Laymen and Women Have a Right to be Properly Informed” (GhanaWeb)
    “New Laws for the Internet: Do We Have a Right to Be Forgotten?” (care2.com)
    Next are examples where analysis would likely show that a claim-right is a part of what is being asserted, along with other Hohfeldian incidents:
    “American Drivers Have a Right to Choose E15 [a gasoline supplement]” (KTIC News)
    “Residents Have a Right to Park” (South Wales Argus)
    “I pay my segregated fees, just like anyone else, and I should have a right to utilize any building that I wish.” (Dunn County News)
    “Protecting Reality TV’s Right to be Fake” (Salon)
    All of these are perfectly natural uses of our language, phrased in terms of rights, and we understand them with minimal if any reflection. They speak to different types of normative concerns, and it would not be difficult to collect many further examples that speak to other concerns still. This is “rights talk” as we actually hear and speak it, this is what conceptual analysis should help us understand. I see no reason to think that Rainbolt’s special class of rights defines some sort of paradigm, measured against which the rights above that fall outside of that class are deviant or inferior or unreal. All of the examples above deploy the concept of a claim-right, and that is the concept that we are trying to capture with our analysis.

30. Lest anyone think, by the way, that the claim-right itself is somehow the paradigm of rights it is also easy to find assertions that are focused on other Hohfeldian incidents, such as the privilege:
    “Americans have a right — indeed a duty — to make reasoned, fact-based criticisms … But the critics don’t have a right to invent their own facts” (National Review Online)
    Indeed we often hear and make statements using the concept of a right when we are not even talking about rights of conduct at all, but rather, for example, affective rights (“All students have a right to feel safe,” said one article. “Do Oilers Fans Have a Right to be Angry?” asked another.) In these domains of rights, there are no claim-rights at all. (More on affective, epistemic and conative rights is here.)
    One final assertion about rights on the day I searched the web might serve as the motto for this forum (were it not too parochial for its contributors):
    “We Need a National Conversation on What Rights Actually Are” (IMAO)
    Whatever rights are, we should start the conversation by embracing the variety of rights-assertions that we hear and make every day. Our analysis will only succeed if we resist the tunnel-vision that sometimes afflicts us, and when we welcome what look like counter-examples from our news sources and from our everyday conversations. This is a matter of being faithful to our own ideas.

31. Gilbert says she has found what she calls an important class of rights in the exact area that she has been researching for many years. She calls these “demand-rights.” Her strategy for advertising the specialness of this class of rights is to suggest that claim-rights, which most theorists would assume was the larger class of which demand-rights are a subset, may in fact be more obscure, and less unified, than has previously been recognized. Her favored class of rights is, she says, likely ancient and weighty, while the class of rights that others have been speaking about for decades may actually be poorly defined and fragmented. The difficulty for Gilbert is that she has given little reason for believing that demand-rights are anything other than what they first appear to be: another species of claim-rights that some moral and political theorists may choose to study especially intensively.
    Taking up an early point in her comment, Gilbert is certainly correct that the most indubitable claim-rights assertions that we know involve the word “right.” This is in the same way that the most indubitable neutron star assertions that we know involve the word “star”. Yet we are doing conceptual analysis, so we must press on to search for something more informative. The standard formulations are that a person who has a claim right is owed a duty by another, or has another’s duty directed toward her, or is the object of another’s duty. Gilbert worries that these phrases may be odd, yet the weight of the tradition of thinking about claim-rights in terms of these phrases presses toward asking for more explanation as to why.
    Moving on, all of Gilbert’s problem-cases for kind-desire analysis can be handled by it. In her rescue case and her gratitude case there is no enforceability (urging is not enforcing), so the third condition of the analysis is unmet and we reach Gilbert’s result that there is no right. Gilbert herself gives the answer to her third case, which is also very close to the answer given in the article. When Jack makes a promise to Jill then he has a duty to do what he promised to her. If that assertion sounds awkward to Gilbert then perhaps she and I just disagree on this, and we can leave it to other members of the forum to decide whether the assertion is out of bounds or not. Is Jack’s duty to do what he promised to Jill, as Gilbert asserts, “a duty that is not representable as a duty to phi [his] promisee?”
    Gilbert next raises the question of the ascription of desires to promisees, wondering whether we do in fact assume that a promisee (who has not waived the promise, etc.) wants the promise to be fulfilled. It does seem to me that everyday social interaction supports this assumption, and that we make this assumption all the time. Once we make a promise, we act on the premise that our promisee wants us to do what we offered and they agreed that we do. The example from the article is the illustration of the driver delivering the engine to the amateur mechanic. The driver quite naturally assumes that the mechanic wants the engine that he ordered. The driver tells the mechanic that if he didn’t want the engine he shouldn’t have ordered it or he should have cancelled the order. The ascription of the desire to promisees is nothing more extraordinary than that, and indeed it may be that such attributions are so ordinary that they can be missed.

32. So far it looks like no serious problems for kind-desire theory. Now it could be, as Gilbert suggests, that the project of finding a unified analysis of the claim-right is doomed from the start. It could be that the class of claim-rights is in fact too diverse for this. This is certainly a possibility– just as it is possible that there is no unified grammar for sentences of English, or no unified physics for the origin of planets. Still, what this article sets out is a precise formalization of the individually necessary and jointly sufficient conditions for the existence of a claim-right within a normative system. This is a formalization that appears to capture all known claim-rights, leaving (as yet) not a single counter-example. If the analysis is doomed, it should be possible to show where it fails. Gilbert has not herself offered to take on the role of prophet of doom for claim-rights analysis, but someone tempted by the role might heed what Scripture says about prophets: By their fruits ye shall know them.
    The rights analyst who studies claim-rights can agree that exploring Gilbert’s demand-rights and what may be special about them could be a fascinating study, just as the rodentologist who studies the genus Rattus can agree that exploring what is special about the species Rattus Rattus can be a fascinating study. Of course we should heed Gilbert’s own worry about demand-rights themselves. Talk of demand-rights may be ambiguous, and there may be no unified theory of demand-rights there to find. Certainly we have not yet seen any analysis of demand-rights that is as rigorously specified and extensively tested as kind-desire theory is for claim-rights. Still, we may hope that such an account of demand-rights may someday be developed, and as Gilbert says, “there’s no harm in trying.”

33. Thanks very much to Rowan Cruft for his comments on the difference between kind-desire theory and Raz’s interest theory, and also for raising the concerns about the demand theory. All of what he says seems quite plausible, and I’d recommend the Aristotelian Society paper in his comment to anyone interested in exploring the relations between rights and respect.
    Thanks also to Craig Nishimoto for further questions about the formalization of the kind-desire theory. His question about how we are to determine whether a system of norms uses a kind-desire description is fair enough in the abstract. However when we go to cases, I can’t find a one where the concern has bite. When we think within a particular normative system (such as morality) we can ask ourselves whether that system contains a certain role (such as third party beneficiary). The answers appear to come in a straightforward way (as Nishimoto agrees, there is no role of third-party beneficiary within morality). This transparency of normative systems is as we should expect. In order to play a game we need to know who the players are; in order to make judgments within any normative system we need to know what kinds of beings inhabit that system. This is basic, and within the cases I can think of, not difficult.

34. In Nishimoto’s adjusted snow-shoveling example, it could be that he and I have different views about what moral rights there are and aren’t. Nishimoto says that after a community’s law specifies that all homeowners must shovel the snow in front of their houses, each community member in good standing has a moral right that each homeowner shovels. He goes so far as to say that even an eight-year-old child who is a community member in good standing has a moral right that each homeowner shovels.
    This I must admit I can’t see. Say it the city of Cambridge Mass. that passes the snow-shoveling ordinance. Does an eight-year old living near Lechmere Square really have a moral right that a homeowner on the far side of Fresh Pond (some five miles away) shovels his walk? My view is that if not ‘obviously not,’ then at least ‘not obviously.’
    Nishimoto started the shoveling example with a right-holder that looked like someone who would be put at risk by an unshoveled walk. He then moved to a right-holder who could plausibly be thought to be covered by the Principle of Fairness. We have seen how the analysis covers those kinds of cases. If morality really does present us with a third kind of right-holder in the guise of a distant eight-year old, perhaps more could be said to illuminate him or her.

35. Nishimoto’s adjusted promisee example again shows the analysis working correctly. In trying to generate this counterexample Nishimoto appeals to the hypothesis that ‘promisees, as promisees, desire to be benefitted.’ Yet this cannot be correct. Imagine walking up to someone in the street and saying:
    “You’ve accepted a promise, so I’m assuming you want money. Here’s $1000.”
    Your startled passerby might accept the cash, but would likely think you a bit mad. Contrast this with the desire that (as we’ve seen) it is rational to assume that promisees have:
    “You’ve accepted a promise, so I’m assuming you want performance. Here’s what I promised you.”
    Promisees as such don’t want to be benefitted—they want performance. So the analysis gets the right result once more.