Eberhard Karls Universität Tübingen
Studierendenkonferenz 2010 mit Thomas M. Scanlon
How to Avoid the Wrong Kind of Reasons Problem:
Judgment-Sensitive Attitudes and Correctness
The Wrong Kind of Reason Problem (WKR) probably is the best known and the most frequently discussed
objection to Thomas M. Scanlon’s buck-passing account of value. Most of the proposed solutions to it have been
confuted in the debates about buck-passing accounts. Andrew E. Reisner and Gerald Lang argue against a
“Brentano style” solution to the WKR in terms of correctness which is defended by Sven Danielsson and Jonas
The main aim of the essay is to show that this solution may be successfully defended against its critics.
In a first step the WKR and the proposed solution of Danielsson and Olson will be briefly reconstructed in order
to examine Reisner’s and Lang’s objections on this basis. Then Scanlon’s conception of judgment-sensitive
attitudes will receive some attention. In a last step it will be argued that a criterion of correctness for judgmentsensitive
attitudes is possible and a modification of Scanlon’s original account will be suggested. But the
proposed modification is not any longer a buck-passing account and it may be questionable if Scanlon can
defend his buck-passing account against the WKR without giving up the core idea of buck-passing.
1. Introduction: The Buck-Passing Account of Value
In the second chapter of What We Owe to Each Other Thomas M. Scanlon argues against a purely
teleological conception of value – namely that to be valuable means to be to be promoted – and offers
his own proposal, which he calls the buck-passing account of value. 1
The main idea thereby is that value has to be understood in terms of reasons:
… being valuable is not a property that provides us with reasons. Rather, to call something valuable is to
say that it has other properties that provide reasons for behaving in certain ways with regard to it. 2
The reason-providing properties are natural, non-normative properties, 3 so goodness or value become
“purely formal, higher-order properties of having some lower-order properties that provide reasons of
the relevant kind.” 4 Even if Scanlon is not explicit about what is meant by “behaving in certain ways
See Scanlon: What We Owe to Each Other, p. 96 and p. 97. The buck-passing account of value can be
classified as a special form of a fitting-attitudes theory of value. Other earlier proponents of fitting-attitudes
theories may be found in Franz Brentano, Charles D. Broad and Alfred C. Ewing.
Scanlon: What We Owe to Each Other, p. 96.
Later Scanlon accepts evaluative properties as well. See Scanlon: Reasons, Responsibility, and Reliance:
Replies to Wallace, Dworkin, and Deigh, p. 513.
Scanlon: What We Owe to Each Other, p. 97.
with regard to it” it seems to be clear that he thinks of judgment-sensitive attitudes in the first place.
One may then schematize his account as follows:
X is valuable. ↔ X has other natural (or evaluative) properties that provide reasons (of the
relevant kind) for certain judgment-sensitive attitudes.
Scanlon thus offers an intra-normative reduction: the evaluative is analysed in terms of the deontic.
And he can be understood as not only offering a conceptual analysis but as making a metaphysical
claim as well.
It is not the aim of this essay to examine if this intra-normative reduction is especially helpful or
reasonable, but Scanlon’s proposal shall be made as strong as possible and therefore it is necessary to
defend it against the most common objection, the Wrong Kind of Reasons Problem (WKR).
2. The Wrong Kind of Reasons Problem (WKR)
The main point of the WKR is that there may be reasons for a certain (judgment-sensitive) attitude 5
towards an object X which would not be called valuable. 6 The paradigmatic example of the debate is
the evil demon which will inflict a severe pain on me unless I prefer this saucer of mud. 7 Intuitively, a
saucer of mud is not preferable or valuable but the example suggests that I do have a reason to prefer
it, namely the threat of the evil demon. However, this reason seems to be of the wrong kind.
Scanlon seems to have anticipated this objection by sometimes using the careful formulation
“providing reasons of the relevant kind”, but he never explains what “reasons of the relevant kind” are
and so he offers no real solution to the problem.
There have been several proposals how to distinguish between the right kind of reasons and the wrong
kind of reasons, but none of them has been fully convincing. By modifying the example “the evil
demon always strikes back.”
3. A “Brentano style” Solution to the WKR 8
Danielsson and Olson introduce the distinction between two kinds of epistemic reasons for an attitude
(whether cognitive or conative): there are reasons which justify having the attitude, so called holdingreasons
and there are reasons which justify the correctness of the attitude, so called content-reasons. 9
The whole debate focuses on attitudes or pro-attitudes and thereby ignores Scanlons specific concept of
Justin D’Arms and Daniel Jacobson call it the Conflation Problem. See Justin D’Arms/Daniel Jacobson:
Sentiment and Value; The Moralistic Fallacy; Sensibility Theory and Projectivism.
It was introduced by Roger Crisp and it has been modified several times to show that the proposed solutions all
The following remarks refer to Danielsson/Olson: Brentano and the Buck-Passers.
The term “content-reason” and especially the analogy between correctness and truth may suggest that we are
dealing with ontological reasons, which means facts that render the representational content of a belief or attitude
true or correct (truth-makers or “correctness-makers”). However, this is not how Danielsson and Olson
Because it is a plausible normative hypothesis that we normally ought to have correct attitudes, having
a content-reason for some attitude implies a defeasible holding-reason for this attitude (but not vice
For the ascription of value only content-reasons are decisive.
X is valuable. ↔ X has other properties which provide content-reasons (reasons for the
correctness) for a certain attitude.
One may illustrate the distinction between holding reasons and content reasons referring to the
analogy to reasons for belief. In this case reasons for correctness are substituted by reasons for truth. 11
Pascal’s wager may be used as an example for a holding-reason for a belief: as I would come to
heaven if God existed and I believed in him, I have a reason to believe in him. But obviously this is
not a reason which could render my belief that God exists true and therefore no content-reason. For
this I would rather need conclusive arguments to the effect that the belief is or would be true. 12
Similar in the case of attitudes: content-reasons for an attitude would be arguments to the effect that
the attitude is or would be correct. 13 If we now turn to the example of the evil demon it may be
described as follows: Although I clearly have a reason to prefer the saucer of mud – namely the threat
of the demon –, this is only a holding-reason for preferring. There is no reason for the correctness of
the attitude in question, because a saucer of mud cannot provide content-reasons for preferring it or
render preferring correct. Of course, the subsequent question has to be what would justify the
correctness of preferring: obviously, we should not refer to the property “preferable” in our
Danielsson and Olson adopt this analogy between truth and correctness from Brentano, who thinks
that correctness is a technical philosophical notion for conative attitudes which is analogous to the
notion of truth for beliefs. Both are fundamentally epistemic notions, which means, that they cannot
satisfactorily be explained without reference to the notion of experiences of evident knowledge. 14
Thus the justification of the correctness of an attitude runs via the experience of evident knowledge.
Danielsson and Olson do not decide on a special account of correctness and they think that they do not
understand the term: content-reasons are epistemic reasons. Danielsson and Olson admit that they are not
entirely happy with the term “content-reason”. See Danielsson/Olson: Brentano and the Buck-Passers, p. 517.
Danielsson and Olson refine their proposal by analysing the notion of a holding-reason in terms of the notion
of a content-reason: „To say that there is a holding-reason to have some attitude is to say that there is a contentreason
to favour the occurence of this attitude, or possibly that there is a content reason to disfavour the nonoccurence
of this attitude.“ In the case of the evil demon that means that the holding reason to prefer the saucer
of mud can be explained in terms of a content-reason to favour preferring the saucer. See Danielsson/Olson:
Brentano and the Buck-Passers, pp. 518ff. However, this refinement is not important for the present purposes.
There may be reasons for having a belief (e.g. strategic reasons) and there may be reasons for the truth of the
belief (e.g. a conclusive argument). As in the case of attitudes we may differentiate between holding-reasons and
content- or truth-reasons for belief. Having the latter implies defeasible holding-reasons for the belief (but not
See Danielsson/Olson: Brentano and the Buck-Passers, p. 516.
See ibid., p. 517.
See ibid., p. 516.
have to, because as we can talk of truth without being precise about the nature of the truth predicate
we can as well talk of correctness without being precise about this notion. 15
Nevertheless, an unexplained notion like that of correctness is not very helpful and always calls for
There are two main lines of criticism to this interesting proposal. First, the notion of correctness is too
evasive and Danielsson and Olson fail to say anything illuminating to it. 16 And second, there seems to
be a directionality problem. 17
Concerning the first point, there seems not to be much to say in defence of Danielsson and Olson.
There is an obvious disanalogy between truth and correctness: in the case of truth, we operate with the
concepts of reason, truth and belief, all of which have some prior analysis. Even if we do not agree on
a common account of truth, we have a secure grasp of the extension of the predicate “true”. Talking
about reasons, truth and belief is somehow informative.
In the case of correctness things look different: we do not seem to have the same secure grasp of the
extension of the predicate “correct”. Correctness is only understandable schematically by the role it
plays in the proposed theory and it seems to be just what is in question when we consider the WKR.
As Reisner formulates it: “… we have a […] account of what distinguishes the right kind of reasons
from the wrong kind of reasons: namely that what distinguishes them is just that property which
distinguishes them.” 18 So the proposal becomes uninformative if we do not know what correctness
However, Olson thinks we have a fairly secure grasp of the extension of “correct”. 19 People have the
intuition that preferring is not a correct attitude towards a saucer of mud. This reply in terms of
intuitions leads to the second line of criticism, the directionality problem.
In order to decide whether a certain attitude is correct or whether a consideration really is a contentreason,
it seems that we have to judge if the object in question is in fact intrinsically valuable. That
means, “[t]he epistemology of determining whether some considerations contribute to the correctness
of the content of a pro-attitude thus works back form an initial intuition concerning what is
intrinsically valuable.” 20 We seem to have a much better grasp on the left-hand side of the buckpassing
biconditional than on the right-hand side. To understand the primitive concepts of the
See Danielsson/Olson: Brentano and the Buck-Passers, p. 516.
See Lang: The Right Kind of Solution to the Wrong Kind of Reason Problem, p. 482; Reisner: Abandoning
the Buck Passing Analysis of Final Value, pp. 385f.
See Reisner: Abandoning the Buck Passing Analysis of Final Value, p. 387; Lang: The Right Kind of Solution
to the Wrong Kind of Reason Problem, pp. 487f.
Reisner: Abandoning the Buck Passing Analysis of Final Value, p. 385.
See Olson: The Wrong Kind of Solution to the Wrong Kind of Reason Problem.
Reisner: Abandoning the Buck Passing Analysis of Final Value, p. 387.
analysans – correctness or content-reason – we have to resort to the analysandum. So a question of
directionality arises: is it really necessary or easier or more natural to start with the right-hand side of
the biconditional than it is to start with the left-hand side? If we follow the epistemological trail it
seems to be plausible to take intrinsic value as the primitive concept which can be used to analyse
reasons and correctness. So the whole buck-passing account of value seems to be in danger if we
cannot offer a plausible understanding of correctness which does not go directly back to our intuitions
5. What Are Judgment-Sensitive Attitudes?
The enterprise of finding a criterion of correctness for attitudes crucially depends on what is
understood by an “attitude”. As mentioned above, Scanlon thinks of judgment-sensitive attitudes. So
what are judgment-sensitive attitudes?
These are attitudes that an ideally rational person would come to have whenever that person judged
there to be sufficient reason for them and that would, in an ideally rational person, ‘extinguish’ when
that person judged them not to be supported by reasons of the appropriate kind. Hunger is obviously not
a judgment-sensitive attitude, but belief is, and so are fear, anger, admiration, respect and other
evaluative attitudes such as the view that fame is worth seeking. 21
In addition to this list Scanlon mentions intentions, hopes, contempt, indignation and wanting. 22
Judgment-sensitive attitudes are intentional mental states which have a sort of (propositional) content
whereby they are distinguished from mere feelings.
Especially interesting about Scanlon’s set of judgment-sensitive attitudes is that we find cognitive (e.g.
beliefs) and conative (e.g. intentions) mental states as well as some emotions in there.
23 What unifies
them is that they either directly reflect the agent’s judgment or are supposed to be governed by it. 24
We have to justify our judgment-sensitive attitudes by giving reasons for them – they are “‘up to us’ –
that is, they depend on our judgment as to whether appropriate reasons are present” as Scanlon says.
However, because of their dependence on reasons, it is not open to us to adopt, modify or reject
judgment-sensitive attitudes in any way we choose: reasons normally depend on facts outside us 26 and
these facts tie judgment-sensitive attitudes somehow to the world. 27 This gives us a first hint at how
correctness of judgment-sensitive attitudes may be construed: facts of the world bear on the
correctness of judgment-sensitive attitudes. Furthermore, it is not open to us to choose any
consideration as counting in favour of a judgment-sensitive attitude:
Scanlon: What We Owe to Each Other, p. 20.
Ibid., p. 18, p. 20 and p. 50.
Scanlon nowhere explicitly defines which judgment-sensitive attitudes are essential for buck-passing.
See Scanlon: What We Owe to Each Other, p. 22.
Ibid. Here again we find Scanlon talking about „appropriate reasons“.
Perhaps one might even speak of a mind-to-world direction of fit. However, as noticed above it is unclear how
facts could bear on the correctness of conative attitudes which have the opposite direction of fit.
A reason is a consideration that counts in favor of some judgment-sensitive attitude, and the content of
that attitude must provide some guidance in identifying the kinds of considerations that could count in
favor of it… These background judgments about the kinds of things that can count in favor of certain
attitudes are themselves open to question and piecemeal modification. 28
This gives us a second hint: not any fact of the world can bear on the correctness of judgment-sensitive
attitudes. The contents of the judgement-sensitive attitudes play a decisive role in identifying possible
reasons for them.
6. Correct Judgment-Sensitive Attitudes and the Right Kind of Reasons
If we go back to Olson’s attempt to respond to the objection that we do not have a secure grasp on the
predicate “correct” we see that his line of thought is not absurd – we all think preferring is not a
correct attitude towards a saucer of mud – but his appeal to our intuitions about value causes problems.
We have to stay on the right-hand side of the biconditional and search for solutions there. Admittedly,
the danger of circularity always stays at least in the background.
The following will be attached to Scanlon’s second hint, that the content of judgment-sensitive
attitudes plays a decisive role in identifying possible reasons for them.
As Wlodek Rabinowicz and Toni Rønnow-Rasmussen observe, “[i]t seems, then, that different types
of pro- and contra-attitudes, by their very nature, exhibit certain characteristic kinds of concerns:
different kinds for different attitudes.” 29 So it might be thought that the correctness of an attitude is
determined by its characteristic concern.
Instead of talking of “characteristic kinds of concerns” of attitudes the concept of a formal object of a
judgment-sensitive attitude will be introduced. 30
The suggestion is that mental states have formal objects which are correlates that correspond to each
kind of mental state 31 and by which mental states can be individuated – Fabrice Teroni calls this the
“Individuation requirement”. 32 In the case of emotions and in the most cases of judgment-sensitive
attitudes 33 this is not a real object of the world but an evaluative property which has to be ascribed
necessarily to the object to render the emotion or judgment-sensitive attitude intelligible – the
“Intelligibility requirement”. 34 A third requirement is the “Correctness requirement”, namely that
formal objects are needed to access whether an emotion or judgment-sensitive attitude fits its
Scanlon: What We Owe to Each Other, p. 67.
Rabinowicz/Rønnow-Rasmussen: The Strike of the Demon, p. 420.
The idea of formal objects stems from the scholastic tradition and was picked up by the modern philosophy of
emotion. Note for example Anthony Kenny and William Lyons. Fabrice Teroni offers a very illuminating
discussion of the formal objects of emotions.
See Teroni: Emotions and Formal Objects, p. 396.
See ibid., pp. 398f. and pp. 401ff. For example, the formal object of admiration would be the admirable.
An exeption may be belief: the formal object of belief is truth.
See Teroni: Emotions and Formal Objects, p. 399 and pp. 403ff. For example, we cannot make sense of
somebody who admires a painting of Caravaggio but who would not ascribe the property of being admirable to
particular object. 35 An emotion or judgment-sensitive attitude is correct iff the particular object really
exemplifies the evaluative property specified by the formal object. Thus we may specify epistemic
reasons which justify the ascription of the formal object and which justify the emotion or judgmentsensitive
It is now suggested that a correct judgment-sensitive attitude has to take up properties in its intentional
content which refer – in a sense to be specified – to the formal object and which thereby play the dual
role just mentioned.
Obviously, if we want to avoid circularity and maintain the reductive claim of the analysis we should
not mention an evaluative property like the formal object on the right-hand side of the biconditional. 37
However, we may mention descriptive properties which refer to the formal object, e.g. via a reliable
indicator account or via specific supervenience relations. 38 This idea becomes clear if we think of
“thick” evaluative concepts which necessarily include descriptive parts. These descriptive parts or
properties justify the ascription of the evaluative property specified by the formal object. Thus they
have to be taken up by the intentional content in order to speak of a correct judgment-sensitive
This suggestion works especially well for relatively specified or “thick” judgment-sensitive properties
whose formal object is a thick evaluative property but it may be also suitable for rather “thin” attitudes
like preferring. 39 At first sight, one may think one has to appeal to other evaluative properties to
capture the formal object “preferable” but these may open the possibility to name some of their
descriptive parts. In almost the same manner, one can also allow for the possibility that a judgmentsensitive
attitude like admiration can be correct towards an elegant movement as well as towards a
gracious movement because elegant and gracious can be seen as parts of the formal object, namely the
See Teroni: Emotions and Formal Objects, p. 399 and pp. 408ff. For example, admiring a painting of
Caravaggio is correct if the painting really is admirable.
This proposal is quite similar to Peter Goldie’s suggestion of justifying reasons for emotions: Non-axiological
properties provide justifying reasons which justify both the ascription of emotion-proper properties (or formal
objects) and emotional responses. See Goldie: Emotion, Feeling, and Knowledge of the World, p. 98.
If we think of belief and its formal object truth this possibility becomes obvious: we have to ascribe the
property of being true to our beliefs (otherwise we get into Moore’s paradox) but truth does not have to figure
explicitly in the content of each of our beliefs. The same may hold for admiration: if we admire the painting of
Caravaggio the admirable does not have to figure explicitly in the content of our admiration.
Clearly this does not completely rule out circularity but it seems to diminish the danger of vicious circularity.
However, one may allow explicit circularity and claim that it is not vicious and unavoidable like for example
Wiggins does but this does no longer seem to be a proper conceptual analysis. Rabinowicz/Rønnow-Rasmussen
come to the conclusion that circularity can be tolerated in philosophical analyses. See Rabinowicz/Rønnow-
Rasmussen: Buck-Passing and the Right Kind of Reasons, p. 120. They refer to Humberstone: Two Types of
Circularity; and Keefe: When Does Circularity matter?
See Teroni: Emotions and Formal Objects, pp. 410f. In the case of the Caravaggio painting we may mention
the colours or the specific use of light etc.
Rabinowicz and Rønnow-Rasmussen think that the suggestion cannot account for “thin” attitudes. See
Rabinowicz/Rønnow-Rasmussen: The Strike of the Demon, pp. 421f.
admirable. No information about the tiny difference between elegant and gracious is lost since the
difference may be taken up in the intentional content of admiration. 40
A similar apprehension is that it might not be clear in all cases which properties bear on the
correctness of a judgment-sensitive attitude as it might not be clear which descriptive parts a “thick”
value predicate contains. It seems as if this worry cannot and should not be fully removed because of
the essential contestability and the ongoing development of our value concepts: what was courageous
a long time ago may now seem to be rather careless. As Scanlon recognizes: “These background
judgments about the kinds of things that can count in favor of certain attitudes are themselves open to
question and piecemeal modification.” 41 All this seems to speak in favour of the no-priority-argument
and a reciprocal adjustment of value ascriptions and subjective reactions. 42
The aim of this essay is to defend and develop the “Brentano style” solution to the WKR suggested by
Danielsson and Olson. This proposal works with the unexplained notion of correctness which asks for
objections. After having examined two possible objections by Lang and Reisner Scanlon’s notion of
judgment-sensitive attitudes has been exhibited. Scanlon gives us two hints concerning the correctness
of judgment-sensitive attitudes: first, even if some of them are conative mental states, their correctness
depends on facts about the world 43 and second, their intentional content plays a decisive role in
identifying possible reasons for them. By introducing the concept of formal objects for judgmentsensitive
attitudes the following modified account of value is the consequence:
X is valuable. ↔ X has other natural (or evaluative) properties that provide reasons for a
certain correct judgment-sensitive attitude, which means the properties
appear in the intentional content of this attitude and refer to the formal
object of this attitude. These properties play a dual role: (i) they justify
the ascription of the formal object and (ii) they justify the judgment-
Thus the right kind of reasons are those reasons which justify a correct judgment-sensitive attitude.
With the introduction of the notion of a formal object of judgment-sensitive attitudes we can
accommodate our intuitions that not everything can be the proper object of a judgment-sensitive
attitude but since the formal object is conceptually tied to the judgment-sensitive attitude we do not
have to rely on our intuitions about value to explain this first intuition.
However, due to the application of the notion of formal objects circularity cannot be fully avoided and
so this modified account is no longer a buck-passing account of value but rather another special form
of a fitting-attitudes theory. Scanlon might accept a solution like this but then he would have to give
Reisner wants to establish an objection of this kind which he calls the “inaccuracy objection”. See Reisner:
Abandoning the Buck Passing Analysis of Final Value, pp. 389ff.
Scanlon: What We Owe to Each Other, p. 67.
See Wiggins for a fuller development of this point.
One might even think about giving up the strict division of the mental into two mutually exclusive categories.
up the core idea of buck-passing: a non-circular intra-normative reduction of the evaluative to the
deontic. But the problems with the WKR may suggest that a buck-passing account cannot be resolved
and that one has to look for other possibilities.
But of course, in the end one can never be sure if the proposed modification can resist all evil demons.
Scanlon, Thomas M.: What We Owe to Each Other. Cambridge, Mass./London: Belknap 1998.
Scanlon, Thomas M.: Reasons, Responsibility, and Reliance: Replies to Wallace, Dworkin, and Deigh. In: Ethics
112 (2002), pp. 507-528.
Crisp, Roger: Reasons and the Structure of Justification: How to Avoid Passing the Buck. In: Analysis 65 (2005),
Danielsson, Sven/Olson, Jonas: Brentano and the Buck-Passers. In: Mind 116 (2007), pp. 511-522.
D’Arms, Justin/Jacobson, Daniel: Sensibility Theory and Projectivism. In: David Copp (Ed.): The Oxford
Handbook of Ethical Theory. Oxford/New York e.a.: Oxford University Press 2006, pp. 186-218.
D’Arms, Justin/Jacobson, Daniel: Sentiment and Value. In: Ethics 110 (2000), pp. 722-748.
D’Arms, Justin/Jacobson, Daniel: The Moralistic Fallacy: On the ‘Appropriateness’ of Emotions. In: Philosophy
and Phenomenological Research 61 (2000), pp. 65-90.
Goldie, Peter: Emotion, Feeling, and Knowledge of the World. In: Robert C. Solomon (Ed.): Thinking about
Feeling. New York: Oxford University Press 2004, pp. 91-106.
Humberstone, Lloyd: Two Types of Circularity. In: Philosophy and Phenomenological Research 57 (1997), pp.
Keefe, Rosanna: When Does Circularity Matter? In: Proceedings of the Aristotelian Society 102 (2002), pp. 275-
Kenny, Anthony J.P.: Action, Emotion and Will. London: Routledge & Kegan Paul 1963, especially chapter 9.
Lang, Gerald: The Right Kind of Solution to the Wrong Kind of Reason Problem. In: Utilitas 20 (2008), pp.
Lyons, William: Emotion. Cambridge: Cambridge University Press 1980, especially chapter 6.
Olson, Jonas: The Wrong Kind of Solution to the Wrong Kind of Reason Problem. Forthcoming in Utilitas.
Rabinowicz, Wlodek/Rønnow-Rasmussen, Toni: Buck-Passing and the Right Kind of Reasons. In: The
Philosophical Quarterly 56 (2006), pp. 114-120.
Rabinowicz, Wlodek/Rønnow-Rasmussen, Toni: The Strike of the Demon: On Fitting Pro-attitudes and Value.
In: Ethics 114 (2004), pp. 391-423.
Reisner, Andrew E.: Abandoning the Buck Passing Analysis of Final Value. In: Ethic Theory and Moral
Practice 12 (2009), pp. 379-395.
Teroni, Fabrice: Emotions and Formal Objects. In: Dialectica 62 (2007), pp. 395-415.
Wiggins, David: A Sensible Subjectivism? In: David Wiggins: Needs, Value, Truth. Essays in the Philosophy of
Value. Aristotelian Society Series. Vol. 6. Oxford: Basil Blackwell 1987, pp. 185-214.