The Buck-Passing Account of Value: Lessons from Crisp
Forthcoming in Philosophical Studies
S. MATTHEW LIAO
Faculty of Philosophy, Oxford
University, Littlegate House, 16/17 St. Ebbes St., Oxford OX1 1PT, UK; e-mail:
October 1, 2009
The Buck-Passing Account of Value: Lessons from Crisp
T. M. Scanlon’s buck-passing account of value (BPA) has been subjected to a barrage of criticisms. Recently, to be helpful to BPA, Roger Crisp has suggested that a number of these criticisms can be met if one makes some revisions to BPA. In this paper, I argue that if advocates of the buck-passing account accepted these revisions, they would effectively be giving up the buck-passing account as it is typically understood, that is, as an account concerned with the conceptual priority of reasons or the right vis-à-vis value or the good. I conclude by addressing some of the broader implications of my arguments for the current debate about the buck-passing account of value.
Keywords: buck-passing account of value; Roger Crisp; reasons and value; T. M. Scanlon
The Buck-Passing Account of Value: Lessons from Crisp
1. Scanlon’s buck-passing account of value
According to Scanlon’s buck-passing account of value (BPA), being good or valuable “is not a property that itself provides a reason to respond to a thing in certain ways. Rather, to be good or valuable is to have other properties that constitute such reasons.” As Scanlon explains, since the claim that some property constitutes a reason is a normative claim, BPA “takes goodness and value to be non-natural properties, namely the purely formal, higher-order properties of having some lower-order properties that provide reasons of the relevant kind.” In other words, on BPA, the normative ‘buck’ is passed from goodness to other properties that provide reasons, that is, the reasons are provided not by goodness itself but by other properties that make something good.
BPA is part of a long philosophical tradition going back to philosophers such
as Franz Brentano, G. E. Moore, W. D. Ross, C. D. Broad, H. W. B. Joseph, and
A. C. Ewing, who are concerned with the relation between reasons and value, or,
more common in those days, the relation between the right or ought and the
good. According to one view in this tradition, the
good or value is conceptually prior
to the right or reasons. By conceptually ‘prior,’ it is meant that
reasons are to be analyzed in terms of value or the good. For example, in Principia Ethica (but not subsequently),
To motivate BPA, Scanlon argues that reflection on particular cases suggests that reasons are provided not by goodness but by the lower-order properties. In particular, Scanlon observes that
the fact that a resort is pleasant is a reason to visit it or to recommend it to a friend . . . These natural properties provide a complete explanation of the reasons we have for reacting in these ways to things that are good or valuable. It is not clear what further work could be done by special reason-providing properties of goodness and value, and even less clear how these properties could provide reasons.
several writers have pointed out, one of the main attractions of BPA is that it
has the potential to reduce the evaluative realm to the normative realm,
thereby “demystifying value” – to borrow a phrase from Wlodek Rabinowicz and
Tonni Rønnow-Rasmussen – and making the fundamental ethical category to be that
of reasons rather than values. For example, in discussing the merits of BPA,
2. Criticisms of BPA
While Scanlon’s BPA has attracted the support of philosophers such as Derek Parfit, Philip Stratton-Lake, and others, it has also been subjected to many criticisms. Roger Crisp has helpfully identified and grouped together at least three difficulties that BPA faces, as a result of the fact that Scanlon tends to regard lower-order properties as natural properties.
As noted earlier, using the example of a pleasant resort, Scanlon suggests that pleasantness is a natural property and that pleasantness provides a full explanation of the reason we have for visiting the resort. Crisp points out that treating lower-order properties as natural properties faces, firstly, the problem of misrepresenting the phenomenology of evaluative experience. For example, when one sees Leonardo da Vinci’s Mona Lisa, one might regard it as a good or beautiful painting that is to be admired. But Crisp argues that “the reason for admiration lies not in the natural properties – these could be understood by someone with no aesthetic sense – but in the beauty.”
Secondly, Crisp argues that treating lower-order properties as natural properties faces the problem of not being able to discriminate between different kinds of values. As Crisp observes, a painting may be beautiful because of its sublimity, delicacy, profundity, boldness, imagination, vitality, grace, honesty, and so on. But, according to Crisp, some evaluative properties, such as grace and delicacy, are such that they invite very similar responses. Crisp believes that this makes it difficult if not impossible to provide a buck-passing account that distinguishes the evaluative properties in terms of different responses. Because of this difficulty, Crisp then considers and rejects the possibility of analyzing such evaluative properties in terms of sets of natural properties. For example, Crisp points out that it is unlikely that an account of grace in purely naturalistic terms would be satisfactory, because it could not be explained why this particular set of natural properties was being grouped together without reference to the notion of grace itself.
Finally, Crisp argues that BPA also faces the following problem. Either BPA should be read as a biconditional account or it should not be. A biconditional account of BPA entails that whenever something has natural properties such that there is reason to respond to it, that thing is valuable. But, as Crisp points out, a biconditional account of BPA implies that if I say to someone, ‘I promise to φ,’ the natural fact that I have said these words would seem to give me a reason to φ. And because φ-ing has this natural, reason-giving property, on a biconditional account of BPA, it is therefore valuable. But, so argues Crisp, this conclusion seems odd because one might think that there are reasons to keep a promise which are purely deontological and which do not justify an inference to value. As one can see, this difficulty is a special case of the well-known “Wrong Kind of Reasons” (WKR) Problem, which has been discussed by many other writers.
Crisp notes that in personal communication, Scanlon has said that he does not view BPA as a biconditional account. However, Crisp points out that a non-biconditional account of BPA faces the problem of explaining which reason-giving properties give rise to goodness and which do not. For example, Crisp argues that a non-biconditional account of BPA would be unable to distinguish between the kind of choice which one’s having made a promise gives one a (deontological) reason to make, and that which a resort’s being pleasant gives one a reason to make.
3. Crisp’s proposed solutions on BPA’s behalf
Intending to be helpful to BPA, Crisp has suggested some ways for BPA to meet these criticisms, even though he himself does not endorse BPA. Two distinct proposals can be discerned in Crisp’s writings. According to the first proposal, one should distinguish between a negative and positive component of BPA:
Negative component: Being good is not itself a reason-providing property.
Positive component: Being good is merely the higher-order property of having lower-order properties that provide reasons to respond in particular ways.
Crisp argues that a buck-passer should keep the negative component but give up the positive component, because
What is essential to the buck-passing account, I suggest, is the negative component. In other words, as long as the normative buck is passed, it is open to Scanlon to adopt a different positive account of goodness.
As Crisp explains the rationale for this proposal:
[Scanlon’s] argument really concerns the negative component. What puzzles Scanlon is how goodness could provide reasons, not how it might consist in anything other than the higher-order property of having lower-order reason-providing properties.
Although Crisp does not say this explicitly, the idea here could be that if a buck-passer keeps only the negative component, then a buck-passer would not need to be committed to any substantive view about whether natural properties constitute lower-order properties and to whether BPA should be read as a biconditional account or not. If such a buck-passing account is tenable, then this version of the buck-passing account would avert the problems that Crisp has raised against BPA.
Crisp makes explicit that he is not thereby endorsing the buck-passing account of value. As he says, “Whether this view of the relation of goodness and reasons is the most plausible is of course another matter. All I am suggesting in this paper is that, if one is to pass the buck, one do so negatively rather than positively.”
According to the second proposal, Crisp suggests that the lower-order properties (e.g. pleasantness) that Scanlon has tended to regard as natural properties should instead be construed as non-natural, evaluative properties. Crisp notes that Wallace has questioned whether pleasantness (as in the case of Scanlon’s pleasant resort) is a non-evaluative property, and that in response to Wallace, Scanlon seems to accept that the lower-order properties could be evaluative properties. Scanlon says,
I agree with Wallace’s reinterpretation of my “buck-passing” account of value. My thesis was that goodness is not itself a property that provides reasons, not that the underlying properties that do this are always natural properties, and I should not have written in a way that suggested this. He is quite right that more specific evaluative properties often play this role.
While this quote shows that Scanlon appears to endorse the idea that lower-order properties are sometimes evaluative properties, Crisp argues that in order to respond to the problems raised above, “it must be the case that the lower-order properties grounding goodness are always evaluative” (my italics). According to Crisp, the reason is that “if any natural properties were allowed to ground goodness on their own, the question would again arise about the difference between those natural properties that do and those that do not result in goodness.” Crisp calls a buck-passing account according to which the lower-order properties are always evaluative properties the Revised Buck-passing Account (RBPA).
To illustrate RBPA, consider again the case of a pleasant resort. On RBPA, the resort’s having a certain evaluative property, that is, being pleasant, makes it good, and it is this evaluative property (being pleasant) that provides one with a reason to visit it. Crisp argues that RBPA can respond to all three of the problems raised above. For example, the phenomenology of value is preserved, because beauty – which Crisp regards as the reason for admiring da Vinci’s Mona Lisa – can be the lower-order evaluative property that provides the reason to admire this painting. Also, as Crisp argues, since lower-order properties are evaluative properties, the capacity to make fine-grained distinctions between what appear to be different evaluative properties such as grace and delicacy is maintained by the retention of those very properties themselves. Finally, so Crisp argues, the difference between those ways of responding that are characteristic of goodness and those that are not can be made on RBPA, because “good-related choices . . . will at least include those objects characterized in terms of specific evaluative properties which provide reasons for those choices.” If Crisp is right, RBPA appears to answer at least some of the criticisms that have been leveled against BPA.
4. Assessing Crisp’s proposals
Crisp’s proposals on behalf of BPA can be read as
(a) An adequate buck-passing account needs to have a negative component but it does not need to have a positive component; or
(b) An adequate buck-passing account needs to have a negative component and a positive component, but instead of the positive component that Scanlon has provided, one should adopt instead the idea that lower properties that provide reasons to respond in particular ways are always evaluative properties.
In proposing (a), Crisp seems to be reacting to an aspect of Scanlon's redundancy argument, according to which reasons are not provided by goodness. Crisp seems correct in thinking that only the negative component is necessary for a buck-passing account if all one wants from a buck-passing account is to argue that the value property of being good is not reason-providing. But (a) cannot be right because an adequate buck-passing account cannot just have the negative component and no positive component. There are two distinctive features of an adequate buck-passing account. The first is a claim about conceptual priority of reasons or the right vis-à-vis value or the good. The second is a claim about the reason-providing relation, according to which it is not value properties but rather the properties upon which value supervenes that provide reasons to respond or act in certain ways. Given the second feature of the buck-passing account, an adequate buck-passing account should therefore have a positive story of the reason-providing relation. That is, it should have a positive story about the types of properties that are reason-providing. (Note that having a positive story about the types of properties that are reason-providing does not entail that there are actually properties that are reason-providing, since it may be the case that no token of the types of properties that are reason-providing actually exists). However, the negative component holds only that goodness is not itself reason-providing, but it does not have a positive story of the reason-providing relation. In fact, having just the negative component is compatible with the view, according to which there are no types of properties that are reason-providing, since the claim that goodness is not itself reason-providing is compatible with the claim that no types of properties are reason-providing. But the claim that no types of properties are reason-providing is incompatible with the buck-passing account, since on the buck-passing account, some types of properties are reason-providing. So, an adequate buck-passing account cannot just have the negative component and no positive component.
To assess (b), and in particular, RBPA, as a means of helping BPA, it is useful to distinguish two distinct issues. The first is concerned with whether the evaluative such as the good or value can be analyzed in terms of the normative such as the right or ought or reasons. The second is concerned with the relation between general, thin evaluative concepts such as the good and specific, thick evaluative concepts such as grace, delicacy and beauty. As I mentioned earlier, BPA as it is typically understood is concerned with the first issue. In recent decades, philosophers including John McDowell, Bernard Williams, Jonathan Dancy, and Hilary Putnam have been interested in the second issue. For example, McDowell and others have argued that important, thick concepts such as ‘courageous’ cannot be reduced to thin concepts such as ‘good’ and ‘bad.’ An argument that McDowell offers to this effect is the following: The general form of a reductive analysis of a thick concept is ‘X has a descriptive component that fully determines the concept’s extension (e.g. X has descriptive properties A, B, and C), and an evaluative component that says that X is good/bad/right/wrong for having these properties.’ According to McDowell, the reductive analysis implies that one can always identify the extension of a thick term using only its descriptive part and without any knowledge of the evaluations a speaker uses it to make. However, so argues McDowell, this is impossible for virtue-terms such as ‘courageous,’ whose extension depends crucially on evaluations. Therefore, McDowell concludes that the reductive view is false and that important thick concepts are irreducible.
As we shall see, there are good reasons to believe that a number of those currently engaged in the debate regarding the buck-passing account of value have in mind the second issue rather than the first issue. So let us consider RBPA as a means of helping BPA as applied to each issue in turn.
The plausibility of RBPA as applied to the first issue concerning the conceptual priority of reasons vis-à-vis value depends on whether the buck-passing account as it is typically understood can allow lower-order properties that provide reasons to respond in particular ways always to be evaluative properties. I believe that it cannot. On the buck-passing account as it is typically understood, lower-order properties are reason-providing. If they are also always evaluative properties, as Crisp has proposed, then it seems that reasons would be analyzed in terms of value, because what would make these lower-order properties reason-providing would be the values that constitute the evaluative properties. But, as we have noted, on a buck-passing account as it is typically understood, reasons are not analyzed in terms of value; instead value is analyzed in terms of reasons. So, a buck-passing account as it is typically understood cannot hold that the lower-order properties are always evaluative properties without ceasing to be a buck-passing account as it is typically understood. In other words, applied to the issue concerning the conceptual priority of reasons or the right vis-à-vis value or the good, Crisp’s RBPA would deprive a buck-passing account as it is typically understood of much of its attractions, e.g., its potential to reduce the evaluative realm to the normative realm. Note that I am not arguing that it cannot be the case that lower-order properties are always evaluative properties. I am arguing instead that on a buck-passing account as it is typically understood, it cannot be the case that lower-order properties are always evaluative properties.
But perhaps Crisp did not intend for RBPA to be a buck-passing account as it is typically understood, that is, as an account that is concerned with the conceptual priority of the right vis-à-vis the good. Instead, perhaps Crisp intended for RBPA to be concerned with the second issue regarding the relation between general, thin evaluative concepts and specific, thick evaluative concepts. There are reasons to believe that a number of those currently engaged in the debate regarding the buck-passing account of value have in mind this issue rather than the first issue. For example, Wallace says that “If there is an issue of priority, it would seem to concern . . . the relation between general and specific concepts,” where general concepts can be understood as thin concepts and where specific concepts can be understood as thick concepts. Similarly, the Scanlon who endorses Wallace’s reinterpretation of BPA seems also to have this issue in mind when Scanlon says that ‘[Wallace] is quite right that more specific evaluative properties often play this role’ (my italics).
In his discussion of RBPA, Crisp explains how RBPA as a means of helping BPA would construe the relation between general, thin evaluative concepts and specific, thick evaluative concepts:
In the case of the resort, then, on RBPA the thought is that its having a certain specific evaluative property—being pleasant—makes it good, as well as providing one with a reason to visit it, and that its being good does not consist merely in its having that lower-order reason-providing property. Goodness is a self-standing property, but it is not reason-giving (my italics).
On this interpretation of RBPA, what it is for something to be good (thin evaluative concept) is for it to have reason-providing properties of certain kinds, namely, thick evaluative properties. In other words, the general, thin evaluative property of goodness passes the normative buck onto more specific, thick evaluative properties such as pleasantness, such that the thick evaluative properties such as pleasantness are reason-providing, but the thin evaluative property of goodness is not reason-providing.
One worry regarding interpreting RBPA this way as a means of helping BPA is that Crisp has argued against it. As Crisp points out, RBPA so interpreted fails because general, thin evaluative properties can be reason-providing. To show this, Crisp asks us to consider an analogy between the structure of causal explanation and that of normative justification. For example, suppose that something is on fire, and you ask for a full explanation of the fire. A replies that “C threw a burning object into it,” while B replies that “C threw a lighted taper into it.” According to Crisp, on RBPA’s view of explanation, the explanatory buck will pass from the more general property to the more specific, and you will accept B’s explanation, but deny A’s. But, so argues Crisp, these two explanations are quite compatible; it is just that one is more specific than the other. As he says, “The more general explanation is not otiose, since it would help someone who did not know what a ‘taper’ was.” Similarly, Crisp argues that in the case of visiting a resort, the more general justification that the resort is good may not be otiose if we are trying to justify our visit to, say, someone who knows nothing of pleasure. Suppose that we grant Crisp that his criticism of this interpretation of RBPA is correct. For our purpose, it would seem odd to adopt this interpretation of RBPA as a means of helping BPA to answer the criticisms that Crisp has identified, knowing that RBPA so interpreted has its own share of problems.
A second worry against interpreting RBPA this way as a means of helping BPA is that it seems that those who interpret RBPA this way have to accept a further, substantive thesis, namely, the thesis that specific, thick evaluative concepts cannot be analyzed in terms of general, thin evaluative concepts. The reason is that if thick evaluative concepts can be analyzed in terms of thin evaluative concepts, but if the negative component of BPA about goodness is true, that is, goodness does not itself provide reasons, then it seems that thick evaluative properties cannot provide reasons as such. However, on RBPA, thick evaluative properties are reason-providing. So, it seems that those who interpret RBPA this way have to accept that thick evaluative concepts cannot be analyzed in terms of thin evaluative concepts.
A final worry against this way of interpreting RBPA is that if a buck-passer were to move to this interpretation of RBPA, again, one of the main merits of the original BPA would be lost, in particular, its potential to reduce the evaluative realm to the normative realm. In other words, if RBPA is not intended to be a buck-passing account as it is typically understood, then it is of course no objection to RBPA that it is not a buck-passing account as it is typically understood. However, BPA is supposed to be a buck-passing account as it is typically understood. That is, part of the point of BPA is to analyze value in terms of reasons and thereby to make the fundamental ethical category to be that of reasons rather than value. Indeed, as I noted earlier, historically, most philosophers in the BPA tradition are concerned with the conceptual priority of the right vis-à-vis the good. If BPA is suppose to be a buck-passing account as it is typically understood, and if RBPA is not a buck-passing account as it is typically understood, then RBPA does not really help BPA to address the criticisms that have been leveled against BPA. RBPA appears to solve the problems for BPA by changing the topic.
5. General implications
If my arguments are correct, Crisp’s proposed solutions on behalf of BPA are unsuccessful. But they are unsuccessful for interesting reasons, and these reasons have broader relevance for the current debate about the buck-passing account of value. I conclude by briefly detailing them.
First, advocates of BPA should not rely on the idea that BPA needs only to have a negative component and/or the idea that lower-order properties are always evaluative properties to address the criticisms that Crisp has identified against BPA. The reason is that they would effectively be giving up the buck-passing account as it is typically understood, and thereby giving up one of the main merits that the original BPA has, in particular, its potential to reduce the evaluative realm to the normative realm.
Secondly, it is worth remembering that Crisp proposed the ideas discussed above in order to be helpful to BPA, even though he himself does not endorse BPA. Since Crisp’s proposed solutions appear not to help BPA, this leaves advocates of BPA with the task of addressing the criticisms of BPA that Crisp has identified in other ways.
Thirdly, and more generally, it is important to distinguish between the issue concerning the conceptual priority of the right or reasons vis-à-vis the good or value, and the issue concerning the relation between general, thin evaluative properties and specific, thick evaluative properties. As we have seen, a number of those currently engaged in the debate regarding the buck-passing account of value such as Crisp and even Scanlon appear to run these two issues together when discussing the buck-passing account of value. To make progress in this debate, it seems worthwhile to keep the two issues distinct.
I would like to thank Roger Crisp, Andrew Reisner, Guy
Brentano, F. (1955). Vom Ursprung sittlicher Erkenntnis (originally published in 1889). Hamburg: Meiner.
Broad, C. D. (1930). Five Types of Ethical Theory London: Routledge & Kegan Paul.
Crisp, R. (2005). Value, Reasons and the Structure of Justification: How to Avoid Passing the Buck. Analysis, 65, 80-85.
Crisp, R. (2008). Goodness and Reasons: Accentuating the Negative. Mind, 117(466), 257-265.
Dancy, J. (1996).
In Defense of Thick Concepts. In P. French, T. Uehling & H. Wettstein
Dancy, J. (2005).
Should We Pass the Buck? In T. Rønnow-Rasmussen & M. J. Zimmerman (Eds.), Recent Work on Intrinsic Value (pp.
Elstein, D., & Hurka, T. (forthcoming). From Thick To Thin: Two Moral Reduction Plans. Canadian Journal of Philosophy.
Heuer, U. (2006). Explaining Reasons: Where Does the Buck Stop? Journal of Ethics and Social Philosophy, 1(3).
Lang, G. (2008). The Right Kind of Solution to the Wrong Kind of Reason Problem. Utilitas, 20, 472-489.
McDowell, J. (1978). Are Moral Requirements Hypothetical Imperatives? Proceedings of the Aristotelian Society, 52, 13-29.
McDowell, J. (1979). Virtue and Reason. The Monist, 62, 331-350.
(1981). Non-Cognitivism and Rule-Following. In S. H. Holtzman & C. M. Leich
(Eds.), Wittgenstein: To Follow a Rule
Moore, G. E.
(1903). Principia Ethica.
Moore, G. E.
Olson, J. (2004). Buck-passing and the Wrong Kind of Reasons. Philosophical Quarterly, 54(215), 295-300.
Olson, J. (2006). G. E. Moore On Goodness and Reasons. Australasian Journal of Philosophy, 84(4), 525-534.
(2001). Rationality and Reasons. In D. Egonsson, B. Petersson, J. Josefson
& T. Rønnow-Rasmussen (Eds.), Exploring
Practical Philosophy: From Action to Values (pp. 17-41).
(2002). The Entanglement of Fact and Value. In The Collapse of the Fact/Value Dichotomy and Other Essays (pp.
Rabinowicz, W., & Rønnow-Rasmussen, T. (2004). The Strike of the Demon: On Fitting Pro-attitudes and Value. Ethics, 114, 391-423.
Reisner, A. (forthcoming). The Possibility of Pragmatic Reasons for Belief and the Wrong Kind of Reason Problem. Philosophical Studies.
(1998). What We Owe To Each Other.
Scanlon, T. (2002). Reasons, Responsibility, and Reliance: Replies to Wallace, Dworkin, and Deigh. Ethics, 112, 507-528.
P., & Hooker, B. (2006). Scanlon versus
Suikkanen, J. (2005). Reasons and Value - In Defence of the Buck-Passing Account. Ethical Theory and Moral Practice, 7, 513-535.
(2006). Resisting the Buck-passing Account of Value.
Wallace, R. J. (2002). Scanlon's Contractualism. Ethics, 112, 429-470.
Williams, B. (1985). Ethics and the Limits of Philosophy. London: Fontana.
 Scanlon (1998, pp. 95-100, p. 97).
 Scanlon (1998, p. 97).
 See, e.g., Dancy (2005, p. 33); Rabinowicz and Rønnow-Rasmussen (2004) for historical outlines of the buck-passing tradition.
 Some writers explore these matters in terms of metaphysical priority. See, e.g., Väyrynen (2006).
 Whereas I discuss the reasons/value relation in terms of ‘analysis,’ some writers discuss this relation in terms of ‘explanation.’ See, e.g. Wallace (2002, p. 447); Heuer (2006).
 Brentano (1955, originally published in 1889).
 Broad (1930, p. 283).
 Scanlon (1998, pp. 96-97).
 Wallace (2002, p. 447).
 Roger Crisp calls this the redundancy argument (2005, p. 81). Scanlon has another argument for BPA, what Crisp calls the argument from pluralism, according to which many things can be said to be good, and the grounds for these judgments vary widely. Philip Stratton-Lake and Brad Hooker (2006) have argued that this argument is problematic.
 Scanlon (1998, p. 97).
 Rabinowicz and Rønnow-Rasmussen (2004, p. 400).
 Olson (2004, p. 295).
 Lang (2008, p. 472).
 Lang (2008, p. 473).
 Parfit (2001); Stratton-Lake and Hooker (2006). For some criticisms of BPA in addition to the ones that we shall consider here, see, e.g., Rabinowicz and Rønnow-Rasmussen (2004); Dancy (2005); Väyrynen (2006). For some responses to these criticisms, see, e.g., Olson (2004); Suikkanen (2005).
 Crisp (2008); Crisp (2005).
 Crisp (2005, p. 82).
 Crisp (2005, p. 82).
 Crisp (2005, p. 82).
 This problem raised by Crisp about promising also draws on Dancy's worry that BPA threatens to resolve the debate between consequentialism and deontology in favor of consequentialism. See Dancy (2005, pp. 40-42).
 See, e.g., Rabinowicz and Rønnow-Rasmussen (2004).; Dancy (2005); Väyrynen (2006); Olson (2004); Reisner (forthcoming).
 Crisp (2005, p. 83); Crisp (2008, pp. 260-62).
 Crisp (2008, p. 263). See also Väyrynen (2006) for this distinction.
 Crisp (2008, p. 264).
 Crisp (2008, p. 264).
 Crisp (2008, p. 264).
 Crisp (2008, pp. 262-63); Crisp (2005, p. 84).
 Wallace (2002).
 Scanlon (2002, p. 513).
 Crisp (2008, p. 262).
 Crisp (2008, pp. 262-63).
 Crisp (2005, p. 84).
 Crisp (2005, p. 84).
 I thank an anonymous reviewer for this point.
 See, e.g., Williams (1985, esp. pp. 128-30) for the distinction between thick and thin moral terms.
 See, e.g., McDowell (1978); McDowell (1979); McDowell (1981); Williams (1985); Dancy (1996); Putnam (2002, pp. 34-40).
 See, e.g., McDowell (1981, p. 144).
 McDowell (1981, p. 144); See Elstein and Hurka (forthcoming) for a good discussion of this argument by McDowell and why they believe that it fails.
 See also Williams (1985, pp. 130, 141-42).
 Wallace (2002, p. 447).
 Crisp (2008, p. 264).
 Crisp (2005, pp. 84-85).
 Crisp (2005, p. 85).
 See also Stratton-Lake and Hooker (2006) for this point.