Brentano and the Buck-Passers
Sven Danielsson and Jonas Olson
According to T. M. Scanlon s  buck-passing analysis of value, x is good means that x
has properties that provide reasons to take up positive attitudes vis-Ä…-vis x. Some
authors have claimed that this idea can be traced back to Franz Brentano, who said
in 1889 that the judgement that x is good is the judgement that a positive attitude to
x is correct ( richtig ). The most discussed problem in the recent literature on buck-
passing is known as the  wrong kind of reason problem (the WKR problem): it
seems quite possible that there is sometimes reason to favour an object although that
object is not good and possibly very evil. The problem is to delineate exactly what
distinguishes reasons of the right kind from reasons of the wrong kind. In this paper
we offer a Brentano-style solution. We also note that one version of the WKR prob-
lem was put forward by G. E. Moore in his review of the English translation of Bren-
tano s Vom Ursprung sittlicher Erkenntnis. Before getting to how our Brentano-style
approach might offer a way out for Brentano and the buck-passers, we briefly con-
sider and reject an interesting attempt to solve the WKR problem recently proposed
by John Skorupski.
1. The buck-passing account and the wrong kind of reason
problem
According to  fitting-attitudes analyses of value, x is good means that a
positive attitude (a  pro-attitude ) to x is fitting ( appropriate is another
term that is also used in this context). Different versions of this view
have been defended by several authors in the last ten years or so. The
renewed interest in this format of value analysis is largely due to T. M.
Scanlon s (1998)  buck-passing account. But the idea is not quite new.
A. C. Ewing and Roderick Chisholm are often mentioned among its
progenitors, and some authors suggest that it can be traced back to
Franz Brentano, who said in his Vom Ursprung sittlicher Erkenntnis
from 1889 that the judgement that x is good is the judgement that a pos-
itive attitude to x is (would be) correct ( richtig ).1
Some authors, for instance Ewing and Chisholm, take the buck-pass-
ing format of analysis to be a reduction of good to ought. But that is not
how Brentano thought of it. He meant rather that ought should also be
1
See Dancy 2000 and Rabinowicz and RÅ‚nnow-Rasmussen 2004 for historical outlines of the
buck-passing tradition.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
doi:10.1093/mind/fzm511
512 Sven Danielsson and Jonas Olson
explained in terms of correctness, and actually via better than (Bren-
tano 1955, żż13 17). We shall return to this point in section 4 below.
Brentano s approach does seem to involve a form of buck-passing, how-
ever, in the sense that the evaluative notion of goodness is explained in
terms of the normative notion of correctness.
Another trend through the last decades is the idea that the basic nor-
mative notion is neither good nor ought but that of a reason. Combining
these two ideas and using the verb  to favour as a notational variant of
 having a pro-attitude we get the buck-passers thesis that x is good
should be taken to mean that x has properties that provide reasons to
favour x. The most discussed problem in the recent literature on buck-
passing is known as  the wrong kind of reason problem (henceforth the
WKR problem): it seems quite possible that there is sometimes reason
to favour an object although that object is not good and possibly very
evil. The standard example is a scenario in which a mighty demon
threatens to punish us if we do not favour him (or some other object
which is clearly not good). In such a scenario there would seem to be
strong reasons to favour the demon but we hesitate strongly to say that
he would be good; the reasons to favour the demon are of the wrong
kind from the point of view of the buck-passing analysis. The problem
is to delineate exactly what distinguishes reasons of the right kind from
reasons of the wrong kind. In this paper we offer a Brentano-style
solution.
A version of the WKR problem was presented already by G. E. Moore
in his review of the English translation of Brentano s Vom Ursprung Sit-
tlicher Erkenntnis, though Moore spoke not of an ugly demon but of
beautiful objects. Beautiful objects, he argued, are such that we ought
to favour them, but that does not mean that they are valuable it is our
appreciation of them that is valuable (1903, pp. 116; 120f.). As far as we
know, Brentano never responded to Moore s objection but this paper
offers a response on his behalf. Before getting to how our Brentano-
style approach might offer a way out for Brentano and the buck-pas-
sers, we briefly consider and reject an interesting attempt to solve the
WKR problem recently proposed by John Skorupski.
2. Attempts at a solution
A scenario structurally analogous to that of the demon s threat, but
applied to beliefs rather than to conative attitudes, is found in Pascal s
wager (see e.g. Pascal 1910, Sect. III). Pascal argued that we have reasons
to believe in God because in the event that God exists the consequences
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007
Brentano and the Buck-Passers 513
of believing in him would be heaven and the consequences of not
believing in him would be hell. In the event that God does not exist the
costs of believing in him would be negligible in comparison to the
expected gain, provided that the probability of God s actual existence is
greater than zero. In the demon scenario and in Pascal s wager, the rea-
sons to have the pro-attitude (of favouring the demon) and the cogni-
tive attitude (of believing that God exists) respectively, are provided by
the desirable consequences of having the relevant attitudes. Intuitively,
one is inclined to think that this is the heart of the WKR problem: rea-
sons of the right kind are not provided by the consequences of taking
up the relevant attitude. Solutions that draw on this intuition have been
offered by Olson (2004) and Stratton-Lake (2005). However, as Rabino-
wicz and RÅ‚nnow-Rasmussen (2006) and Lang (MS) have shown, the
problem cannot be so easily dealt with.
John Skorupski (2007; forthcoming) treats the problem differently.
Skorupski is a proponent of the idea that the basic normative notion is
that of a reason and he distinguishes between three kinds of reasons:
reasons for actions ( practical reasons ); reasons for beliefs ( epistemic
reasons ); and reasons for feelings and emotions ( evaluative reasons ).
According to Skorupski, you might have a reason to bring it about that
you intend to do some action, F, although you have no reason to do F.
Likewise, you might have a reason to bring it about that you believe
that p or feel a although you have no reason to believe that p or feel a.
To bring it about that you intend to do F, or believe that p, or feel a is to
do things that cause you to intend, or believe, or feel these things. A
reason to bring any of these things about is thus a reason for an action
(or a series of actions). A reason to bring it about that you believe that
p, or feel a, is consequently a practical reason. Skorupski s solution to
the WKR problem for buck-passing is that in the demon scenario, there
simply is no (evaluative) reason to favour the demon; there is only a
(practical) reason to bring it about that we favour him (2007, pp. 10
12).2 Concerning Pascal s wager, Skorupski s view implies that there is
no (epistemic) reason to believe in God; there is only a (practical) rea-
son to bring it about that we believe in God.
This is a neat solution but a problem is that it is difficult to find an
independent rationale for why we should distinguish in the relevant
way between reasons for an attitude and reasons to bring it about that
we have that attitude. In Skorupski s view, this distinction is a basic one
2
Skorupski is by no means alone in taking this view. Similar views are defended by Gibbard
(1990, p. 37), Parfit (2001, p. 27), and Persson (2007). Gibbard and Parfit are not explicitly con-
cerned with the WKR problem for buck-passing, however.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
514 Sven Danielsson and Jonas Olson
so he might maintain that apart from an appeal to intuition there is no
independent rationale to be had. But in the demon scenario, the intui-
tion that we do have reasons to favour the demon seems at least as
strong as any contrary intuition. After all, the reward (of being spared
from punishment) is obviously tied to actually favouring the demon. It
is not easy to see why this merely gives us reasons to bring it about that
we favour the demon, but no reasons to favour the demon. This makes
Skorupski s distinction between reasons for an attitude and reasons to
bring it about that one has this attitude look like an ad hoc manoeuvre
to rescue the buck-passing account from the WKR predicament.3
However, in the case of beliefs there might be something to be said
for the distinction after all. It is an intuitively plausible idea although
not one we subscribe to in this paper that reasons to believe that p
can be provided only by evidence that p is true. By contrast, reasons to
bring it about that one believes that p can be provided by other consid-
erations, such as the desirable consequences of believing that p. In the
case of Pascal s wager, this idea explains why there is no reason to
believe that God exists, although there is a reason to bring it about that
we believe that God exists.4
While this idea provides a rationale for Skorupski s distinction in the
case of beliefs it does not work equally well in the case of pro-attitudes.
Obviously, we cannot say that a reason of the right kind to take up a
pro-attitude can only be provided by evidence that (the propositional
content of ) that pro-attitude is true. Nor can we say that reasons of the
right kind must be provided by good-making features of the object
since that would make the buck-passing analysis circular. To return to
the demon scenario, this means that we are left without an explanation
as to why there is a reason to bring it about that we favour the demon,
although there is no reason to favour him.
3. A Brentano-style approach
Another way out is to distinguish not between reasons for attitudes and
reasons to bring it about that one has these attitudes, but between two
kinds of reasons for an attitude (whether cognitive or conative). One
thing is a reason for having the attitude, let us call this a holding-reason;
3
Rabinowicz and RÅ‚nnow-Rasmussen (2004, p. 412) press this criticism against approaches
similar to Skorupski s.
4
In the next section, we suggest that this actually distorts Pascal s argument. This is why we say
with some hesitation that in the case of beliefs there might be something to be said for Skorupski s
distinction.
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007
Brentano and the Buck-Passers 515
quite another thing is a reason for the correctness of the attitude, let us
call this a content-reason.5 It is a plausible normative hypothesis that we
ought normally (but not always) to have correct attitudes. In other
words, a content-reason for some attitude, a, implies a defeasible hold-
ing-reason for a. But crucially, that there is a holding-reason for some
attitude does not entail that there is a content-reason for that attitude.
In other words, that there is a reason to have an attitude does not entail
that the attitude is correct.
No doubt Pascal s wager offers, or at least is meant to offer, a holding-
reason a reason for having the attitude of believing that God exists.
Other arguments, other proposed reasons for the belief that God exists,
are of a different kind. The classical cosmological arguments, for
instance, are obviously not, at least not in the first round, arguments to
the effect that we should believe that God exists. They are arguments to
the effect that the belief that God exists is true; they are aimed at estab-
lishing content-reasons. The occurrence or non-occurrence of the
belief does not figure in these arguments. It is undoubtedly plausible to
claim that if it is true that God exists, then we ought to have such a
belief. But that claim is based on the further assumption that at least on
such issues we ought to have true beliefs. As we have just suggested, it is
a plausible normative hypothesis that a content-reason to believe that p
implies a defeasible holding-reason to believe that p, or at least a defasi-
ble holding-reason not to believe that not-p.6
In the case of beliefs the distinction between holding-reasons and
content-reasons is thus fairly unproblematic (at least as long as we do
not hold that to say that something is true means that we ought to
believe it). It is one thing to say that we ought to have a certain belief
and quite another to say that it is true (or, speaking of merely possible
beliefs, would be true if it actually occurred). Once again, however,
5
A different distinction that is frequently cited in the literature on the WKR problem is that be-
tween object-given and state-given reasons (Parfit 2001): object-given reasons for an attitude are
provided by properties of the object of that attitude; state-given reasons are provided by properties
of the attitude. Christian Piller (2006) distinguishes between attitude-related and content-related
reasons: a reason is attitude-related just in case it cannot be specified without reference to the atti-
tude for which it is a reason; otherwise it is content-related (cf. Olson 2004, pp. 299f.). Our dis-
tinction is different from both Parfit s and Piller s distinctions. First, the distinction between
content-reasons and holding-reasons is not drawn in terms of the properties that provide reasons.
Second, there might be content-reasons to favour an object that are not specifiable without refer-
ence to the attitudes for which they are (content-)reasons (see Rabinowicz and RÅ‚nnow-Rasmus-
sen 2006 for relevant examples). And it is not the case that there is a holding-reason for some
attitude just in case the reason is not specifiable without reference to the attitude for which it is a
(holding-)reason, for as we will explain below, it is a plausible normative assumption that content-
reasons imply defeasible holding-reasons.
6
Note that the implication is normative, not conceptual.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
516 Sven Danielsson and Jonas Olson
when we turn to conative attitudes, the situation is not quite as
straightforward. It is one thing to say, for instance, that there is reason
to favour something and another thing to say that this attitude of
favouring is what?
Here we can turn to Brentano. It is one thing to say that we ought to
have a certain conative attitude, and another thing to say that it is (or
would be) correct. Correctness is a technical philosophical notion and
the point of it is that it is, as Brentano says, analogous to the notion of
truth.7 In Brentano, truth is a predicate which is primarily applicable to
judgements ( Urteile ) conceived as psychological entities. And analo-
gously, correctness is a property of conative attitudes ( Gemütsbewegun-
gen ,  Lieben und Hassen ). Brentano s views on the relations between
truth and correctness are not very easy to spell out in their exact details.
His views on this matter may also have varied a bit over the years, as did
his theory of truth, but he seems to have thought that both truth and
correctness are fundamentally epistemic notions in the sense that they
cannot be satisfactorily explained without reference to the notion of
experiences of evident knowledge (see e.g. 1973, pp. 127 30).
Setting aside scholarly concerns about how to interpret Brentano on
this point, the notion of correctness may be a useful tool in analysing
norms and values. We may well find it useful to talk about truth with-
out being precise about the nature of the truth predicate, that is, with-
out committing to the correspondence theory or to minimalism, or to
any other theory. Analogously, we may find it useful to talk about cor-
rectness of attitudes without being precise about the nature of this
notion, that is, without committing to the idea that to say that some
attitude is correct just is to concur with some possible expression of the
attitude  which would be the  minimalist or  expressivist
alternative  or to say something with more metaphysically robust
implications. In this paper we remain non-committal between such
rival accounts of correctness.
Let us now move on to translate this Brentanian approach to reason-
language. To cite content-reasons for a belief is to present arguments to
the effect that the belief is (would be) true. To cite holding-reasons for a
belief is to present arguments to the effect that we ought to have the
belief. To cite holding-reasons for a conative attitude is to present argu-
ments to the effect that we ought to have that conative attitude. To cite
7
See Brentano 1973, pp. 130 32; cf. Brentano 1955, żż23; 27.
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007
Brentano and the Buck-Passers 517
content-reasons for a conative attitude is to present arguments to the
effect that that conative attitude is (would be) correct.8
That the demon threatens to punish us unless we favour him is a
strong holding-reason to favour him. But it is a content-reason to
favour him no more than Pacal s wager presents a content-reason for
beliefs in the existence of God. Brentano could have given a similar
response to Moore s challenge that although we ought to favour beauti-
ful objects this does nothing to show that these objects themselves are
valuable. He could have argued that this is so because there are holding-
reasons, but no content-reasons, to favour beautiful objects.
It should be noted that if we argue in this way, then the WKR prob-
lem can be formulated as a problem not just for the buck-passing
analysis of good. No matter what we take the proper analysis of good to
be, we can imagine a demon who threatens to punish us unless we think
that he is good. Surely we then have strong reason to think that the
demon is good, and we need an explanation why we have not thereby
arrived at an argument for his goodness. The distinction between hold-
ing-reasons and content-reasons provides the explanation. Hence, if we
assume that to take x to be good just is to hold an ordinary belief, so to
speak, then there is no special WKR problem with regard to evaluative
judgements. A special problem with regard to evaluative judgements
arises only if we think that to take x to be good essentially involves cer-
tain conative attitudes. For Brentano it was evident that this is so, and
his solution was to introduce the notion of correctness, which is a pred-
icate not of actions, things, or states of affairs as are the more familiar
evaluative and normative predicates but only of conative attitudes,
just as truth is for him a predicate of cognitive attitudes.
This Brentano-style approach, unlike Skorupski s, accounts for the
intuition that we do have reasons to favour the demon who threatens to
punish us if we do not comply with his command. However, since these
reasons are holding-reasons they do not imply that the demon is good.
There simply are no content-reasons to favour the demon. This distinc-
tion between different kinds of reasons for attitudes is thus preferable
to Skorupski s distinction between reasons for attitudes and reasons to
bring it about that one has these attitudes.
One might get the impression that a holding-reason to believe that p,
or feel a, just is a reason to bring it about that one believes that p, or
8
We are not entirely happy with the term  content-reason but have not found a better alterna-
tive. One of the reasons against it in the present context is that Brentano, in contrast to his prede-
cessor Bernard Bolzano and presumably the majority of modern analytic philosophers, did not
assume any abstract  contents of judgements.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
518 Sven Danielsson and Jonas Olson
feels a. Our distinction between holding-reasons and content-reasons
would then be equivalent to Skorupski s distinction.9 But this impres-
sion is based on a misunderstanding. Whereas Skorupski s distinction
concerns reasons for attitudes and reasons for actions, holding-reasons
and content-reasons are both reasons for attitudes.10 Another weakness
of Skorupski s approach now transpires: it seems to distort the inten-
tion behind Pascal s wager. Recall that Skorupski s view implies that
Pascal s wager does not provide a reason to believe that God exists, but
merely a reason to bring it about that one believes that God exists. But
in our view, this gives the wrong picture. Pascal thought that the intel-
lect could not settle the issue whether God exists, so he refrained from
offering an argument to the effect that God exists. What he did offer
was a decision-theoretical argument to the effect that there is reason to
believe that God exists. He presumably also thought that there is reason
to bring it about that one believes that God exists (by taking the holy
water, having masses said, etc.) but that is because the high expected
utility of believing that God exists provides a reason to believe that God
does exist. In our terminology, Pascal did not argue that there is con-
tent-reason to believe that God exists but he did argue that there is
holding-reason to believe that God exists.
4. Refining the approach
At this juncture a critic might complain that we have not accomplished
very much. The buck-passer starts out with the notion of good and the
notion of a reason and he aims to analyse the former in terms of the lat-
ter. The critic might complain that all we have done is to introduce two
notions of a reason holding-reasons and content-reasons and so
we are still left with no less than two problematic notions at our
hands.11 But as was mentioned in the opening section, Brentano seems
to have thought that the notion of ought should be analysed in terms of
correctness. In the idiom of this paper, this means that the notion of a
holding-reason should be analysed in terms of the notion of a content-
reason. To say that there is a holding-reason to have some attitude is to
9
Skorupski has made this suggestion in personal correspondence.
10
An anonymous referee suggested that Skorupski could retort that in the demon scenario, the
practical reason to bring it about that we favour the demon is a reason to want to favour the de-
mon. If Skorupski makes this move, his practical reasons and evaluative reasons will both be rea-
sons for attitudes. But now, the puzzle from section 2 reappears: why do we have reason merely to
want to favour the demon, but no reason to favour him? Again, the reward (of being spared from
punishment) is tied to actually favouring the demon; not to wanting to do so.
11
Rabinowicz and RÅ‚nnow-Rasmussen (2004, p. 423) level this critique against positions like ours.
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007
Brentano and the Buck-Passers 519
say that there is a content-reason to favour the occurrence of this atti-
tude, or possibly that there is a content-reason to disfavour the non-
occurrence of this attitude.12 And just as we ought, in most cases, to
have true beliefs or at least to avoid having false beliefs, we ought in
most cases to have correct conative attitudes or at least avoid having
incorrect conative attitudes. This means, as we have already seen, that a
content-reason for some cognitive or conative attitude implies a defea-
sible holding-reason for that attitude.
To return once again to the demon scenario, to say that there is a
holding-reason to favour the evil demon is to say that there is a con-
tent-reason to favour favourable attitudes towards the evil demon, or
possibly a content-reason to disfavour the non-occurrence of such atti-
tudes. This implies that favouring the demon is good, which is the right
implication; favouring the demon is instrumentally good since it
shields us from punishment.
But now a difficulty that has to do with the recursiveness of Bren-
tano s evaluative theory arises.13 According to Brentano, it is correct to
favour correct attitudes. Now suppose, plausibly, that it is correct to
disfavour a demon that imposes malicious threats. It follows from this
plausible substantive assumption and from what we have said above
that there is a defeasible holding-reason to disfavour the demon, which
is to say that there is a defeasible content-reason to favour disfavouring
the demon. Consequently, there is a content-reason to favour favouring
the demon as well as a content-reason to favour disfavouring the
demon. This might appear to be an unacceptable tension, but it is not.
After all, cases in which there are reasons to favour or disfavour confl-
icting attitudes seem about as commonplace as cases in which there are
reasons for conflicting attitudes to other kinds of objects. This is of
course not to say that two conflicting attitudes can be correct all things
considered at one and the same time. The demon scenario may well be
an instance of an abnormal case in which we ought not to have correct
attitudes. If a powerful and credible demon threatens to unleash world
disaster if we do not favour him it is highly plausible to hold that the
content-reason to favour favouring him outweighs the content-reason
to favour disfavouring him, which is to say that the holding-reason to
favour him outweighs the holding-reason to disfavour him.
12
A. C. Ewing distinguished between the ought of fittingness and the moral ought and suggested
that the latter might be analysable in terms of the former in a way which resembles our suggestion
(1939, p. 14).
13
Thanks to an anonymous referee for raising this difficulty.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
520 Sven Danielsson and Jonas Olson
To return to Moore s challenge to Brentano, to say that there is a
holding-reason to appreciate beautiful objects is to say that there is a
content-reason to favour appreciation of beautiful objects, or possibly
to disfavour the non-occurrence of such attitudes. It is not to say that
beautiful objects are valuable in themselves. In other words, in stating
his objection Moore failed to take notice of Brentano s distinction
between ought and correctness.14
The version of buck-passing we end up with is thus one that analyses
goodness in terms of content-reasons: x is good means that x has proper-
ties that provide content-reasons to favour x. This analysis gives the right
results in the cases just mentioned: favourable attitudes towards the
demon are instrumentally good since they shield us from pain; appreci-
ating beautiful objects (on account of their beauty) is, ex hypothesi,
good in itself. Clearly, what provides the content-reasons to favour x
will be roughly coincidental with what are often called  the good-mak-
ing properties of x . Since there seems to be no reason to assume that
the notion of a reason is immune to the kinds of metaphysical and epis-
temological objections that are traditionally raised against a Moorean
notion of intrinsic goodness15 one may justifiably ask what the attraction
of the view presented in this essay is supposed to be. One feature of the
Brentano-style approach that we find particularly attractive is that a
kind of internalism will be included in the bargain. This will be a kind
of internalism that establishes a necessary link between values and atti-
tudes: necessarily, to claim that an object is valuable is to claim that a
pro-attitude towards that object is (would be) correct.16
Department of Philosophy sven danielsson and jonas olson
Uppsala University
Box 627, 751 26 Uppsala
Sweden
Brasenose College
Oxford OX1 4AJ
UK
14
That Moore misinterpreted Brentano on this point and possibly on some others is argued in
a commendable article by Gabriel Franks (1976).
15
See Olson forthcoming.
16
Earlier versions of this paper have been presented at seminars in Oxford and Uppsala and at
the Open Sessions of the Mind Association and Aristotelian Society Joint Session in Southampton,
July 2006. We also thank the participants of these meetings for valuable discussions. We also thank
Krister Bykvist, Wlodek Rabinowicz, John Skorupski, and an anonymous referee of Mind for their
helpful comments.
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007
Brentano and the Buck-Passers 521
References
Brentano, Franz 1955: Vom Ursprung sittlicher Erkenntnis. Hamburg:
Meiner. Originally published 1889.
  1973: The Foundation and Construction of Ethics, trans. and ed.
E. H. Schneewind, London: Routledge and Kegan Paul.
Chisholm, Roderick M. 1986: Brentano and Intrinsic Value. Cambridge:
Cambridge University Press.
Dancy, Jonathan 2000:  Should we pass the buck? in O Hear 2000.
Egonsson, Dan, Jonas Josefsson, Björn Petersson, and Toni RÅ‚nnow-
Rasmussen (eds) 2001: Exploring Practical Philosophy: From Actions
to Values. Aldershot: Ashgate.
Egonsson, Dan, Jonas Josefsson, Björn Petersson, and Toni RÅ‚nnow-
Rasmussen (eds) [online resource]: Hommage á Wlodek: Philosoph-
ical Essays Dedicated to Wlodek Rabinowicz. Lund: Lund University.
Available at .
Ewing, A. C. 1939:  A Suggested Non-Naturalistic Analysis of Good .
Mind, 39, pp. 1 22.
Franks, Gabriel 1976.  Was G. E. Moore Mistaken about Brentano? in
McAlister 1976. Originally published in The New Scholasticism, 43
(1969).
Gibbard, Allan 1990: Wise Choices, Apt Feelings. Cambridge, Mass.:
Harvard University Press.
Lang, Gerald MS:  The Right Kind of Solution to the Wrong Kind of
Reason Problem .
McAlister, Linda L. (ed.) 1976: The Philosophy of Brentano. London:
Duckworth.
Moore, G. E. 1903:  Review of Franz Brentano: The Origin of Our
Knowledge of Right and Wrong . International Journal of Ethics, 14,
pp. 115 23.
O Hear, Anthony (ed.) 2000: Philosophy, the True, the Good, and the
Beautiful. Cambridge: Cambridge University Press.
Olsaretti, Serena (ed.) 2006: Preferences and Well-Being. Cambridge:
Cambridge University Press.
Olson, Jonas 2004:  Buck-Passing and the Wrong Kind of Reasons . Phil-
osophical Quarterly, 54, pp. 295-300.
  forthcoming:  Reasons and the New Non-Naturalism .
Parfit, Derek 2001:  Rationality and Reasons . In Egonsson, Josefsson,
Petersson, and RÅ‚nnow-Rasmussen 2001.
Pascal, Blaise 1910: Pascal s Pensees, trans. and ed. W. F. Trotter. Toronto:
J.M. Dent & Sons, Ltd.
Mind, Vol. 116 . 463 . July 2007 © Danielsson and Olson 2007
522 Sven Danielsson and Jonas Olson
Persson, Ingmar 2007:  Primary and Secondary Reasons . In Egonsson,
Josefsson, Petersson, and RÅ‚nnow-Rasmussen [online resource].
Piller, Christian 2006:  Attitude-Related and Content-Related Reasons
for Preferences , in Olsaretti 2006.
Rabinowicz, Wlodek and Toni RÅ‚nnow-Rasmussen 2004:  The Strike of
the Demon: On Fitting Pro-Attitudes and Value . Ethics 114, pp. 391
423.
  2006:  Buck-Passing and the Right Kind of Reasons . Philosophical
Quarterly, 56, pp. 114 20.
Scanlon, T. M. 1998: What We Owe to Each Other. Cambridge, MA:
Harvard University Press.
Skorupski, John 2007:  Buck-Passing about Goodness , in Egonsson,
Josefsson, Petersson, and RÅ‚nnow-Rasmussen [online resource].
  forthcoming:  The Unity and Diversity of Reasons .
Stratton-Lake, Philip 2005:  How to Deal with Evil Demons: Comment
on Rabinowicz and RÅ‚nnow-Rasmussen . Ethics, 115, pp. 788 98.
© Danielsson and Olson 2007
Mind, Vol. 116 . 463 . July 2007