INTRODUCTION
According to Scanlon's buck-passing view, goodness is a formal higher-order property constituted by the reasons that count in favour of substantive first-order properties. A persistent criticism of Scanlon, inspired by Roger Crisp's work, is that buck-passing misrepresents the relationship between reasons and goodness.Footnote 1 This misrepresentation of properties argument can be characterized in two objections: the wrong kind of reasons objection; and what I shall call the reasons and good objection. I will here defend the buck-passing view from both arguments. I begin with a discussion of the wrong kind of reasons objection. In section 1, I argue that the state-given reasons/object-given reasons solution is unsuccessful. The reasons for its failure point to a more promising solution: Rabinowicz and Rønnow-Rasmussen's dual role of reasons argument. In section 2, I argue that the dual role of reasons argument succeeds even where its authors claim it fails. In section 3, I consider the reasons and the good objection. I argue that Crisp's objection conflates the distinction between general and formal properties. However, Crisp's objection to the buck-passing nature of formal properties points to a problem in Scanlon's account that is caused by an ambiguity in Scanlon's argument regarding the buck-stopping nature of reasons. This ambiguity suggests that the formal notion of a reason can be buck-stopping, despite Scanlon's claim that goodness is buck-passing in virtue of its formality. In section 4, I adjust Scanlon's argument in order to remedy this ambiguity. Following a suggestion by R. Jay Wallace, I argue that Crisp's objection can be met by augmenting Scanlon's buck-passing account of goodness with a buck-passing account of reasons. Whilst this augmented buck-passing account may provide a response to Crisp's reasons and the good objection, I end in section 5 by pointing to two outstanding problems with buck-passing that require resolution if any buck-passing argument is to succeed.
1. WRONG KINDS OF REASONS: AGAINST THE STATE-GIVEN/OBJECT-GIVEN REASONS SOLUTION
The wrong kind of reasons criticism suggests that the buck-passing relationship creates strange paradoxes in the relationships between what we value and the reasons we have. Roger Crisp has given a well-known example of this argument that has become a focus of the wrong kind of reasons problem. ‘Imagine that an evil demon will inflict severe pain on me unless I prefer this saucer of mud; that makes the saucer of mud well worth preferring. But it would not be plausible to claim that the saucer of mud's existence is, in itself, valuable.’Footnote 2 Philip Stratton-Lake has recently employed a distinction between state-given and object-given reasons in order to resolve this apparent paradox.Footnote 3 I will outline why this defence is unsuccessful, as the source of its failure points towards a more promising solution.
Stratton-Lake introduces the distinction between state-given and object-given reasons in defence of buck-passing because it distinguishes two sources of reasons in separate properties: the property of a desire for a saucer of mud that will save me from pain (P) and its associated state-given reason; and the property of an object the desire for which will shield me from pain (P') and its associated object-given reason. Rabinowicz and Rønnow-Rasmussen object that this distinction does not solve the wrong kind of reasons problem because the properties P and P' correspond such that there is no significant distinction between state-given and object-given reasons. To insist otherwise is to be profligate with properties. Stratton-Lake responds that the distinction between the desire as the ground of the reason in P and the object desired as the ground of the reason in P' is such that the ‘corresponding property [P'] simply reaffirms the presence of the state-given reason. It does not provide a further object-given reason. It merely tells us that the property of the desire is a reason. So a profligate ontology does not undermine the intuitive view about the reason-giving fact or property.’Footnote 4 In other words the distinction between properties P and P' and their associated reason is sufficient to distinguish a reason in favour of a desire from a reason in favour of an object. Thus the apparent wrong kind of reason is a reason in favour of a desire for an object and not the object itself, and therefore not a reason of the wrong kind.
However, this does not seem to meet Rabinowicz and Rønnow-Rasmussen's objection. For Rabinowicz and Rønnow-Rasmussen, the state-given reason is constituted by the property of a desire (a desire for a saucer of mud will save me from pain – P), whereas the object-given reason is constituted by the property of an object (an object the desire for which will shield me from pain – P'). The wrong kind of reasons problem still obtains as although there are two different properties that constitute reasons (P and P'), these properties correspond such that the saucer of mud is included as the intentional object of the state-given reason and is of value therefore on the buck-passing account. There may be two distinct properties, but Rabinowicz and Rønnow-Rasmussen's correspondence point is that the state-given reason must include the properties of the saucer of mud as the intentional object of the attitude (the preference for the saucer of mud that will allow me to avoid pain), and the object-given reason is constituted by the properties of the saucer of mud in the intentional relation of being an object of a evaluative attitude (the saucer of mud the preference for which will allow me to avoid pain). Stratton-Lake's response does not succeed because whilst he claims that it is ‘A's desire that will shield A from pain’, he must mean that it is in fact ‘A's desire for the saucer of mud that will shield A from pain’ and not simply the phenomenon of ‘A desiring’. Therefore, the wrong kind of reasons criticism persists because the properties of the saucer of mud are included in the state-given reason, and consequently the saucer of mud seems of value on buck-passing terms. The correspondence between the properties of the saucer of mud in the state-given and object-given reasons prevents this distinction from rebutting the wrong kind of reasons criticism. The solution from the distinction between state-given and object-given reasons is unsuccessful, and the wrong kind of reason objection drawn from Crisp remains.
2. WRONG KIND OF REASONS: IN DEFENCE OF THE DUAL ROLE OF REASONS SOLUTION
We have just seen that a distinction between state-given and object-given reasons fails to refute the wrong kind of reasons objection because there is no significant distinction between two properties providing different reasons. Rather than posit two properties that have reasons that count in favour of them, Rabinowicz and Rønnow-Rasmussen's alternative solution is a dualist view of reasons that count in favour of a single property. Therefore we retain the strategy of the object-given and subject-given reasons argument of making a distinction between two kinds of reasons, but avoid the mistake of positing two different properties as the source of these reasons. However, whilst Rabinowicz and Rønnow-Rasmussen argue that this is the most promising solution available, they believe that this too fails to provide a complete response. I believe that Rabinowicz and Rønnow-Rasmussen are unduly pessimistic.
According to Rabinowicz and Rønnow-Rasmussen's dual view of reasons, a property can provide a justification for an agent having a reason in favour of an object, and feature in the intentional content of a reason as the basis for the reason. For example: a person's honesty provides a justification for admiring the person, and is the basis for the admiration of the person. This seems to foil the wrong kind of reason objection of the evil demon who will inflict severe pain on me unless I prefer a saucer of mud, because preferring the saucer of mud provides a justification for having a reason to prefer the evil demon. However, the saucer of mud is not the basis for the reason; the basis for the reason is the evil demon's threat. The properties of the saucer of mud do not therefore feature in the reasons in a way that the wrong kind of reason objection suggests.
Rabinowicz and Rønnow-Rasmussen's sticking point with this dual role of right reasons is the case of an evil demon who will punish you unless you admire him on the basis of his determination to punish him. In this case, the property of the determination to punish features in the reasons in the right way in both roles: it justifies admiring the demon, and it is the basis of the reason to admire. Rabinowicz and Rønnow-Rasmussen claim that their dual view of right reasons cannot escape this wrong kind of reason variation.
In defence of the dual role of reasons view, we should first point out that this objection to Scanlon's buck-passing account might not obtain, given his exclusionary view of reasons. For Scanlon, certain reasons are excluded from consideration by other reasons in a deliberation. ‘The reason-giving force of C not only competes with that of D; it urges that D lacks force altogether (at least in the given context). Often, our judgment that a certain consideration is a reason builds in a recognition of restrictions of this kind at the outset.’Footnote 5 The reason to admire the evil demon on the basis of his determination to punish you is excluded by other reasons, such as the reasons against admiring undeserved punishment. But the exclusionary view of reasons may not provide a complete rebuttal, because as a further variation, Rabinowicz and Rønnow-Rasmussen suggest that a masochistic person may indeed admire the demon because of his determination to punish.
This variation of the wrong kind of reasons objection may be read in three ways: first, that the person finds pleasure in punishment; second, that the person finds pleasure in pain; and third, that the person admires the demon because of his determination. I will argue that each of the three wrong kind of reasons problems can be resolved, and therefore that Rabinowicz and Rønnow-Rasmussen's solution is more effective than they claim. We will deal with each variation on the wrong kind of reason objection in turn.
a. Punishment
In the case of the person who finds pleasure in punishment, we may question whether the masochist has understood properly the reasons that count in favour of punishment, and indeed whether they have understood the very notion of punishment itself. If punishment is (at least in part) retributive, then the reasons for punishment must necessarily include retribution of a legal/moral transgression. If the reasons in favour of the determined evil demon are the masochist's masochism and not the masochist's moral/legal transgression, then these reasons cannot be in favour of punishment because a necessary condition of punishment (retribution of transgression) does not exist. Therefore, the wrong kind of reason objection does not apply because punishment provides neither the ground nor the justification for any reasons of the masochist. Masochists qua masochists are not punished; masochists qua transgressors are punished. This objection relies on the masochists’ masochism and not their transgression, and therefore it does not produce a wrong kind of reasons objection to the dual role of reasons view.
b. Pain
However, the masochist may believe there to be reasons counting in favour of the pain caused by the evil demon's action. If this is the case, then it seems we are no longer dealing with the wrong kinds of reason objection, because the determination to inflict pain is precisely what is of value to the masochist. The wrong kind of reasons objection only holds if there is a reason in favour of an object that is not of value. In this version of the masochism objection, the pain caused by the determined demon is precisely what is of value to the masochist. The wrong kind of reason objector may reply that this merely begs the question, because whilst the masochist has reasons counting in favour of admiring an evil demon on account of the demon's determination to inflict painful punishment, such a determined evil demon is not of value. But in the case of the true masochist, such a determined evil demon is of value, at least instrumentally, to the production of the enjoyment of punishment. The wrong kind of reason objector must claim that there is an account of the source of value that explains why the masochist is, in this case, wrong. This might indeed be forthcoming, but it will not be established on the basis of this variation of the wrong kind of reason complaint against buck-passing, as there is no wrong kind of reason present here. The wrong kind of reason objection relies on a reason counting in favour of a property that is not of value, but on this reading pain is not shown to be non-valuable. Therefore the dual role of reasons solution is not defeated by the pain variation.
c. Determination
Can the attitude of determination be good in itself? Rabinowicz and Rønnow-Rasmussen's most serious worry with their own solution seems to be the case of a person who admires the evil demon's determination. However, we may question whether determination is a property that can have reasons that count in favour of it in the way supposed by this case. Determination is a quality of an approach to a task. The reasons that count in favour of determination will be partly constituted by the task, the approach to which determination is a quality of. Our attitudes towards determination to complete a task depend, in part, on the nature of the task. Without a specification of the task determination is shown towards, we cannot specify fully the reason to admire it, and therefore whether it is indeed admirable. Therefore, the wrong kind of reason objection would not apply in this instance, because there is not a fully specified reason to be of either the right or wrong kind. Consequently, determination cannot be of value in itself. When fully specified, determination may be of non-moral value (e.g. the pursuit of intellectual excellence or physical beauty); moral value (e.g. the pursuit of the living together in unity on contractualist grounds); or immoral disvalue (e.g. the pursuit of underserved painful punishment). The reasons that feature in a contractualist buck-passing account of goodness are constrained by reasonable rejection. As determination requires a specification of its objective, determination could only be good if the reasons in its favour are allowed by contractualist justification. If the determination is towards an objective that is not disallowed on contractualist grounds, then its reasons are right reasons. Therefore, it does not generate the wrong kind of reasons problem in Rabinowicz and Rønnow-Rasmussen's dualist view of right reasons as they suggest. Rabinowicz and Rønnow-Rasmussen's criticism of their own dual role of reasons succeeds where they claim it fails. The dual role of reasons does indeed refute the seemingly persistent problem of the wrong kind of reasons objection.
3. REASONS AND THE GOOD
We identified two objections to Scanlon's buck-passing arguments drawn from Roger Crisp: the wrong kind of reasons objection, and the reasons and the good objection. Crisp states the reasons and the good objection in reference to Scanlon's example of a pleasant beach, where Scanlon holds that a beach's goodness provides no reason in its favour, rather its being pleasant provides a reason which constitutes value as a formal higher-order property. Crisp criticizes this account as misrepresenting the phenomenology of value. To clarify his objection, Crisp provides an example with the use of an analogy between the structure of causal explanation and of normative justification. He imagines a case where Arun, Belinda and Cara are standing around a burning tank of petrol. We ask for a full explanation of the scene: Arun replies that Cara threw a burning object into the tank, while Belinda replies that Cara threw a lighted taper into the tank. Crisp argues that on a buck-passing view, we should reject Arun's explanation, and accept Belinda's.Footnote 6 However, Crisp argues that the two explanations are equally valid, but one is simply more specific than the other. Crisp argues that explanations of value possess a similar structure, and that buck-passing rejects explanations of value based on general reasons, because general reasons are not buck-stopping. However, it seems to me that Scanlon's buck-passing argument does not claim that general reasons cannot stop the buck. The relevant distinction for buck-passing is between formal and substantive, and not specific and general. The difference between general/specific and formal/substantive seems conflated in Crisp's objection.
a. Substantive reasons and the general/specific distinction
According to Crisp's reasons and the good objection, Scanlon's buck-passing argument holds that general reasons, such as Arun's, cannot stop the buck because only first-order substantive reasons counting in favour of objects are buck-stopping. Crisp argues this because he reads Scanlon's buck-passing argument as describing value as the general property of considerations counting in favour of objects. But on the burning tank of petrol example, a buck-passing account would accept both Belinda and Arun's explanations because both are substantive. Both explanations are substantive because they refer to the particular properties that caused the combustion of the tank of petrol. It is certainly the case that there are differing degrees of generality in the explanations, but they both include reference to the substantive cause of the fire. That buck-passing allows for both general and specific descriptions of substantive reasons is clear from an example of value employed by Scanlon. He argues that the pleasantness of a beach is a substantive reason that counts in its favour, and therefore stops the buck. The pleasantness of a beach is a general yet substantive description of certain of its more specific substantive properties, including its warmth, cleanliness, location, and so on. From this example we can see that Scanlon's buck-passing argument does not claim that the buck is passed from general to specific properties. Rather, the buck is passed from formal to substantive properties. Crisp's reasons and the good objection therefore seems to conflate the distinction between formal and general, and misrepresent the distinction in buck-passing between general/specific reasons and formal/substantive. Both general and specific reasons can be substantive, as we have shown with the burning tank of petrol and the pleasant beach example above. But only substantive reasons and the objects they count in favour of can be buckstopping.
This response may be charged with begging the question, however, as it is the formality of the property of goodness, and therefore its inability to stop the buck, that is precisely at issue. Crisp may happily accept that Scanlon's buck-passing argument assigns buck-stopping substance to both general and specific reasons, whilst continuing to object to the basic buck-passing claim that goodness and value cannot stop the buck in virtue of their formality. What is the nature of the formality of goodness and value such that it cannot stop the buck? We will see that there is an ambiguity in Scanlon's original account of the buck-passing nature of formal properties. On Scanlon's original presentation it is ambiguous whether the formal property of being a reason can be buck-stopping. If the formal property of being a reason can stop the buck, then the argument that goodness is buck-passing in virtue of its formality seems threatened. Once this ambiguity is resolved, we will be able to respond to Crisp's reasons and the good objection.
b. Formality and Buck-Stopping
Crisp's reasons and the good objection disputes that goodness and value are formal properties, and that as such they cannot stop the buck. What is the nature of this formality? On Scanlon's original presentation, if we ask someone why she visited the beach, she may answer in substantive terms: ‘because it was warm and clean’, or ‘because it was pleasant’. As described in the previous section, these answers are specific and general substantive answers. Substantive reasons refer to the substantive properties of the object (in more specific or more general terms). On Scanlon's original formulation, goodness and value are formal because they simply describe the relationship between objects and the reasons that count in their favour. The substantive content of the normative reasons counting in favour of objects stops the buck; the formal property of goodness and value merely refers to the relationship between an object and a consideration counting in its favour. On Scanlon's view of reasons as primitive, a reason is a consideration that counts in favour of an object. Therefore, reasons must include the substantive consideration and the object of which it counts in favour. The formality of goodness and value must therefore omit this content, and therefore their formality consists in their reference to the relationship of counting in favour of considerations and objects. In other words, when the explanation refers to considerations and objects it is substantive; when an explanation refers to the relationship of counting in favour, it is formal. Thus, Scanlon's explanation of why goodness and value cannot stop the buck is that they lack reference to any substantive content of reasons or the objects they count in favour of, and refer only to the relationship of counting in favour. However, as R. Jay Wallace has pointed out, there are two problems with this account of the formality of goodness and value.Footnote 7 These problems concern the nature of the properties and the nature of the reasons that are meant to be the substantive buck-stopping properties.
First, on the original presentation of the buck-passing argument, Scanlon took the view that the properties that had reasons counting in their favour were natural properties. He gave the examples of the pleasantness of the beach, and the light that is cast on the illness of cancer by scientific research.Footnote 8 However, Wallace argues that these properties include evaluative elements. The pleasantness of the beach and the knowledge instrumental to health have reasons in their favour because of the value conferred on these objects. If these first-order properties can include evaluative elements, then Scanlon's original distinction between the buck-passing formality of goodness and value, and the buck-stopping substantiveness of natural objects and the reasons that count in their favour, cannot hold. In response to Wallace's criticism regarding natural properties, Scanlon has accepted Wallace's point and revised his view in such a way that first-order properties may be evaluative.Footnote 9
A second problem with Scanlon's original account of the buck-passing nature of goodness and value concerns the relationship between reasons and values. Wallace points out that on Scanlon's account ‘what is at issue is the explanatory priority of reasons vis-à-vis values, or (as we might put it) of the normative vis-à-vis the evaluative’.Footnote 10 The formality of goodness and value in part consists in the buck-stopping nature of reasons (and the objects in which they count in favour). However, it is unclear why a reason is buck-stopping whilst the evaluative properties of goodness and value are not. When asked ‘why did you visit the beach?’, two people may respond ‘because I find beaches relaxing’, or ‘because I had reason to visit the beach’. It seems as though both answers should stop the buck, according to Scanlon's original argument, because normative reasons (and the objects in which the count in favour) are buck-stopping whilst goodness and value are not. This is the sense in which Wallace describes the normative as having priority over the evaluative. But whilst the explanation ‘I visited the beach because I had reason to’ is not tautologous, it certainly seems formal according to Scanlon's criteria of formality we established earlier. This explanation refers to the relationship of a consideration counting in favour of an object, and not to the substance of the consideration itself. If it is the formality of goodness and value that prevent them from stopping the buck, then it cannot be that the buck is passed simply to reasons (and the properties in which they count in favour), because the formal property of reasons is incapable of stopping the buck on the very same grounds as the formal property of goodness and value.
In order to defend the argument that goodness and value are buck-passing in virtue of their formality, we need to clarify the formal/substantive distinction, and how it relates to reasons and values. The clarification presented here is prompted by Wallace's remark that ‘a buck-passing account can with equal plausibility be formulated for the case of reasons for action’.Footnote 11 I describe this as an augmented buck-passing view of reasons and value.
4. AN AUGMENTED BUCK-PASSING ACCOUNT OF REASONS AND VALUE
On the augmented buck-passing view I propose, the buck is passed from both the formal evaluative property of goodness and the formal normative property of a reason, to the substantive properties of objects and the substantive considerations and evaluative attitudes that count in their favour. This augmented buck-passing account of reasons and value employs the distinction between reasons as intentional-action-related, and values as evaluative-attitude-related. This distinction allows us to distinguish the kinds of responses required from agents in favour of objects. On this augmented buck-passing view, objects can have both substantive evaluative attitudes and substantive intentional actions counting in their favour. We should not, as Scanlon does, describe this solely as a buck-passing argument about value. We should instead re-describe the buck-passing view as concerned with both the formal properties of reasons as considerations counting in favour of objects, in addition to value as a formal property of evaluative attitudes counting in favour of objects. Buck-passing requires this redescription and augmentation, because on Scanlon's original view, it is claimed that goodness cannot stop the buck in virtue of its formality, whilst on this same view the buck is stopped by the formal notion of a reason counting in favour. Therefore, for the buck-passing thesis that formal properties cannot be buck-stopping to be preserved, we must add the view that the formal property of a reason counting in favour must be buck-passing in addition to the formal property of goodness and value. The augmented buck-passing argument adds the distinction between reasons as intentional-action-related and value as evaluative-attitude-related, to provide a more consistent buck-passing view: the buck is passed from both the formal properties of a reason and goodness, to the substantive properties of considerations and evaluative attitudes that count in favour of objects. We replace Scanlon's view that the normative is prior to the evaluative, with the view that the substantive is prior to the formal. Formality is therefore uniformly buck-passing on the augmented buck-passing view, whilst it was only partially buck-passing on Scanlon's original view.
We may explicate my version of this argument with the example of a piece of music. A piece of music has certain properties related to, inter alia, structure, expression, harmony, melody, and rhythm. These properties at once provide specific reasons, and require appropriate evaluative responses. We have reason to listen to the music carefully, to listen often, to study it and learn about it, to promote its qualities to others, and to listen to it instead of inferior pieces of music. These are the specific reasons, attached to its first-order substantive properties. The music will also require certain evaluative responses from us. We will admire it, respect it, find it beautiful, and treasure it. These are the specific evaluations that are attached to its empirical properties. On my augmented buck-passing account, the buck is passed from both the higher-order properties of a reason and value onto the substantive lower-order properties of the object and specific reasons and evaluations that count in favour of them.
Scanlon presented his buck-passing account of value partly in terms of the promising response it seemed to offer to G. E. Moore's open-question argument. The buck-passing view attempted to avoid the problem of claiming that the good provided a reason. On Scanlon's original formulation, and on our augmented formulation, this problem is avoided, because we do not claim that the property of good provides any (further) property that counts in favour of the object under consideration. However, on our re-characterization of the buck-passing argument, we have now introduced an extra step by claiming that the normative status of being a reason is as formal and second-order as the evaluative status of being of value. It is redundant to claim that the fact that there is a reason to φ provides a (further) reason to φ. The buck should be passed from the formal properties of reasons and value, to the substantive considerations and evaluations that count in favour of objects, rather than from the formal notion of value to the equally formal notion of a reason, as suggested by Wallace's notion of a buck-passing account of reasons for action.Footnote 12
On this view, substantive properties ground the reasons and evaluations that together constitute our intentional actions. The formal properties of a reason and value are descriptive of the relationship between these substantive properties and the first-order substantive properties of the objects in which they count in favour. Scanlon seems to accept that the buck is passed from the formal properties of value and reason, to the substantive first-order properties of the object. The formal second-order properties do not add anything evaluative or normative to the reasons that there are. The first-order substantive properties of the beach give us reasons because of their particular properties. These particular, concrete properties are the locus of our reasons. On the augmented buck-passing view, the concept of value is formal, just as in the original buck-passing account, but so is the concept of a reason. The buck is not passed from the evaluative notion of value and goodness to the normative notion of rightness or wrongness. In this way, we may retain the buck-passing move of rejecting value as a source of reasons, but extend this rejection to the notion of a reason as a buck-stopper. Scanlon's original buck-passing account stated that the property of value does not provide reasons. Instead, reasons are provided by lower-order substantive properties of objects, and the property of value is a formal, higher-order property that is constituted by reasons counting in favour of something. On our augmented buck-passing account, both the property of being of value and the property of being a reason are higher-order formal properties. The buck is passed from them onto lower-order evaluative and normative properties.
Can this augmented buck-passing account respond adequately to Crisp's reasons and the good criticism? Crisp's objection centred on the plausibility of the buck-passing move from the formal properties of goodness and value to the substantive properties of reasons and the objects in which they count in favour.
On Scanlon's original account, goodness and value were incapable of stopping the buck because of their formality, and so the buck was passed to substantive properties of objects and the reasons that count in their favour. However, we saw that in this argument goodness and value could not be buck-passing in virtue of their formality, because the buck could be stopped by the formal notion of a reason. Formality was buck-stopping, and Crisp's objection seemed vindicated. However, in order to clarify the relationship between reasons and values, and formal and substantive properties, we presented an augmented buck-passing account of reasons and values in which substantive considerations and evaluative attitudes count in favour of first-order properties, and the notions both of a reason and of goodness and value are formal higher-order ones. This establishes the plausibility of the formal/substantive distinction required for buck-passing. We saw that in Scanlon's original argument, the formal/substantive distinction was not applied properly to the relationship between reasons and value. However, applied correctly, the formal/substantive distinction is able to support the essence of the buck-passing argument that formal properties cannot stop the buck.
On the augmented buck-passing view, the formality of the properties of reasons and value refers to the relationship of counting in favour between substantive considerations, evaluative attitudes and substantive properties of objects. These properties are formal because they refer neither to the substantive properties of the objects nor to the content of the evaluative attitudes or reasons. The explanation ‘I went to the beach because there was a reason and an evaluation counting in favour’ is incomplete, because it does not refer to the substantive properties that provide the content for the reasons and attitudes. It is not tautologous, but it is formal because it omits reference to the substantive properties of the considerations, attitudes and objects. Rather, it refers only to the relationship of counting in favour that exists between reasons, evaluative attitudes and the substantive properties of objects.
Crisp is certainly correct that both general and specific explanations can be buck-stopping. But both general and specific explanations are substantive in virtue of their reference to the substantive properties of objects and the substantive content of reasons and evaluative attitudes in their favour. Crisp's criticism seems based on a conflation of the distinction between formal and general. This conflation was justified given Scanlon's own conflation in his notion of a reason on his original view. However, on the augmented buck-passing argument presented here, the distinction between the formal and general is clarified. It is shown that a buck-passing account must maintain consistently the buck-passing nature of formal properties and the buck-stopping nature of substantive properties, and apply them across both reasons and values. Once the relationship between substantive and formal properties is clarified in the augmented buck-passing account, we can see more readily why the buck must be passed from both the formal properties of reasons and values to the substantive properties of considerations, evaluations and the objects in which they count in favour. Indeed, on each of the examples used by Crisp, the buck is stopped by a substantive property (whether general or specific). This indicates that an explanation of the considerations (and evaluative attitudes) in favour of objects must include reference to substantive properties. A purely formal explanation omits this necessary element. A purely formal answer to the question ‘why did you visit the beach’ such as ‘I visited the beach because there was a reason and an evaluative attitude counting in its favour’, absent reference to any substantive properties, refers to the counting in favour of relationship between considerations, evaluative attitudes and properties of objects. The augmented buck-passing claim is that a reason and an evaluative attitude in favour of an object cannot be provided by any property that does not include any substantive content to the reasons and evaluative attitudes in an object's favour. Thus, to say that Yarmouth is a good resort is not to say nothing. It says that there is a consideration and an evaluative attitude that counts in its favour. But to say that Yarmouth is a good resort is not to say anything that give us a specific or general substantive reason or evaluation in its favour. The formal properties of a reason and goodness are without substantive content; they simply refer to the relationship of counting in favour between considerations, evaluative attitudes and objects. Without substantive content there is no basis for any particular reason or attitude. Without substantive reasons and attitudes, there is no basis to why and how persons are to relate to particular objects positively. Without an account of why and how persons are to relate to particular objects positively, we must pass the buck of relating to objects positively to properties that can provide this account. The purely formal properties of a reason and goodness cannot stop the buck in virtue of their formality. They cannot stop the buck because they omit the buck-stopping substantive properties of reasons, evaluative attitudes and the objects they count in favour of. On each of Crisp's examples employed in his objections, the buck is stopped by substantive properties.
The goal of this augmented buck-passing account has been to respond to Crisp's reasons and the good objection, by clarifying the nature of the relationship between formal and substantive properties, general and specific properties, the normative and the evaluative, and practical reasons and evaluative attitudes. However, whilst the augmented buck-passing argument may clarify certain of these relationships and provide a response to the important criticism levelled by Crisp, we should note two remaining problems for buck-passers.
5. PROPERTIES AND THE AUGMENTED BUCK-PASSING ACCOUNT
In the preceding sections, I have defended Scanlon's buck-passing account of value. We saw that the buck-passing account of value faced two main criticisms: the wrong kind of reasons criticism; and the reasons and the good criticism. I presented arguments to show that these criticisms could be met, and augmented the buck-passing account to improve its internal coherence. Whilst I have argued that the augmented buck-passing account is an improvement on Scanlon's original presentation, it seems to me that two significant problems remain. These problems do not necessarily threaten the validity of the buck-passing account, but they are important omissions that require further work in order for the augmented buck-passing account to carry its weight in argument. I describe these as the constitution problem, and the determination problem.
a. The Constitution Problem
We recall that my augmented buck-passing account of value described a relationship between first-order substantive properties of objects and the reasons and evaluative attitudes that count in their favour. But what is the relationship between these reasons and evaluative attitudes and the first-order properties? How are we to explain the relationship between substantive first-order properties of objects and the reasons and evaluative attitudes that are related to them? Scanlon states that the properties of objects constitute these reasons and evaluative attitudes, and he provides the case of the beach that is pleasant. The beach has certain properties such as temperature, cleanliness and being sunny, and on the buck-passing view these properties provide the evaluative attitudes and the reasons that then constitute the properties of right and good. But how does temperature and sunshine constitute a reason and an evaluative attitude? Scanlon does not provide an explanation in his buck-passing account, but we do get intimations in other authors that the relationship may be one of supervenience.Footnote 13 Supervenience may be a plausible account of the relationship between the properties of objects and the reasons and evaluative attitudes that are constituted by them, but this argument requires amplification and interrogation. I do not suggest the constitution problem is insurmountable, but it is an important omission that supporters of the buck-passing account should address in order to enhance the argument.
b. The Determination of Reasons and Evaluative Attitudes
Scanlon presents his buck-passing account of value as part of his wider theory of what we owe to each other. Our duties to others are prescribed by the value of right and wrong, and the value of right and wrong is grounded on certain properties. These properties are principally our capacity to evaluate and act on reasons, and our commitment to live in a relationship with others where our actions are justified on terms no one with similar commitments can reasonably reject. Scanlon wants to explain why the reasons against wronging have a special importance and priority in our practical relationships with other persons, and the buck-passing account of value plays an important part in this explanation. But it seems to me that the buck-passing account of value, as currently presented, is ill-equipped to provide Scanlon with the resources he needs to establish the importance and priority of what we owe to each other. The buck-passing account, in itself, cannot tell us which properties have reasons and attitudes that count in favour of them, and which reasons and attitudes are constituted by these properties.Footnote 14
The buck-passing account itself could be informed by an objective or subjective account of properties and their reasons and values. An objective account would follow if we could establish that certain attitudes and reasons attach necessarily to certain objects and that the objects have a necessary particular nature. Similarly, a subjectivist account may follow if we allow that the determination of the reasons and attitudes and even the objects is determined subjectively (these are just two examples). The augmented buck-passing account is not itself an account of the objectivity or subjectivity of reasons and values or of the nature of properties, and so can only be used in addition to such an account.
A buck-passing account of value must also be augmented with an account of how we determine the reasons and the evaluative attitudes that are associated with the properties of objects. As Pekka Väyrynen points out, ‘no sufficiently definite account may be available of what distinguishes positive responses (“pro-attitudes”) from the negative ones (“con-attitudes”), and both from responses that are neither’.Footnote 15 Even if we are able to find a way to determine properties, let us say on subjective grounds that a particular beach is sunny, how do we determine the appropriate reasons and evaluative attitudes that are constituted by these properties and which are essential to the buck-passing account? Should we visit the beach to exploit its qualities, or stay away to protect it; should we admire its qualities, or should we enjoy them? Just as the buck-passing account needs an explanation of how we determine which properties are the appropriate objects for a buck-passing account, so we need an explanation of the determination of the appropriate reasons and attitudes that are derived from the properties of these objects.
The omission of an account of the constitution of reasons and evaluations by first-order substantive properties of objects, combined with the omission of an account of the determination of the properties and the reasons and values constituted by them, hinders the augmented buck-passing account's successful employment in defence of Scanlon's What We Owe to Each Other, and in moral theories more widely. My purpose has not been to suggest that these accounts are impossible, but rather to suggest areas on which defenders of buck-passing should concentrate.
CONCLUSION
I have sought to defend buck-passing from two important criticisms drawn from Roger Crisp's work. The wrong kind of reasons criticism has issued in many attempted defences and criticisms of Scanlon's buck-passing argument. I have argued that the most successful defence of buck-passing is found in Rabinowicz and Rønnow-Rasmussen's dual role of right reasons argument. I argue that this response is successful even where its authors claim it fails. In response to Crisp's reasons and the good objection, I argued for a clearer distinction between formal and general properties, and presented my augmented buck-passing account in defence of buck-passing. Inspired by Wallace's reference to the plausibility of a buck-passing account of reasons in addition to goodness and value, the augmented buck-passing account distinguishes between practical reasons as intentional-action-related, and values as evaluative-attitude-related. It seeks to avoid the notion of independent reason-giving properties of reasons as well as the notion of goodness and value. I have suggested that whilst this augmented buck-passing argument can respond to Crisp's reasons and the good criticism, the augmented buck-passing account of reasons and value contains significant omissions that must be addressed. My aim in this article has been to identify the weakness in the augmented buck-passing account as it stands currently, in order to point out where more work should be done by its supporters. These problems are particularly important for those who wish to promote Scanlon's contractualism.
I have alluded to the important role of the buck-passing argument in Scanlon's account of what we owe to each other. The buck-passing argument is important to establishing the nature of the relationship between reasons, values and properties of objects. These phenomena all feature centrally in Scanlon's account of the justification of principles of right and wrong. However, it seems that Scanlon wishes to explain why the reasons against wronging have a particular importance and priority in our relations with other persons, and the buck-passing account must therefore include an explanation of the properties that feature in this relationship and why these properties have a particular importance and priority. This explanation will get to the heart of the status of the moral reasons and the status of Scanlon's contractualist account of morality. The nature of these properties will determine the scope of Scanlon's contractualism, and the final status of the reasons against wronging. The work on addressing these omissions is therefore important to the success of the wider buck-passing account of reasons and values, and crucial to Scanlon's contractualism in particular.
