Second counterexample: hedonic desires

3.1.3. Second counterexample: hedonic desires

Turn next to hedonic desires. They are the likings or dislikings of our own present conscious states such as a desire for a cold shower, an aversion to the sound of squeaking chalk, or an aversion to the touch of velvet. Parfit thinks that these desires are not based on reasons. (Parfit 2001, 26) We only have to go through the potential grounds of these desires to see what led him to this thought. First, we can safely assume that the object of these desires is not intrinsically valuable: the feeling of cold water on our skin and the sound of squeaking chalk are not things we regard as valuable in themselves. Then the question arises whether they are instrumentally valuable. This is possible but contingent. It might be that I desire to pull down the chalk on the blackboard in order to draw the attention of my class. Yet, nothing guarantees that we find similar considerations for every hedonic desire. Similarly with state-given reasons: the state of desiring these sensations can hardly be intrinsically valuable, while reference to the effects of desiring is a contingent matter. The first premise is compromised.

At this point, a defender of the premise can point out that we forgot about a crucial feature of these desires, namely that they are hedonic. That is, even if we abstract away from all possible but contingent instrumental grounds, and even if we accept that neither the state of desiring nor the object of desire is intrinsically valuable, it will still be the case that the agent has these desires for a reason: that their satisfaction is pleasurable. And pleasure, the argument continues, is a

Attila Tanyi, THESIS, Page 111

consideration that is a reason in its own right, ungrounded in desire.⁵⁵ (Quinn 1993, 243; Raz 1999a, 58; Scanlon 1998, 44; Heuer 2004, 50) But opponents are ready with an answer. Already in his ‘definition’ Parfit introduced the core idea. Hedonic desires, he said, are the likings and dislikings of our present conscious states that make these states pleasant, painful or unpleasant.

(Parfit ibid.) In other words, what makes the object of these desires pleasurable or painful is the desire itself, and the making relation here is understood as that of constitution: the fact that I can’t stand the sound of squeaking chalk just consist in the fact that I dislike it and so on. (Chang 2004, 76) As a result, pleasure is not a reason in its own right but is a reason that it itself based on desire. The premise is back again.

Parfit takes the truth of his claim for granted; yet, it is far from obvious that he is right.

After all, pleasure and desire are distinct existences; this would be hard to deny. But then the relation between them could be of different kinds and Parfit’s idea about constitution is just one among the potential candidates.⁵⁶ One historically influential approach, for instance, holds that all

55 This statement raises two questions that are not, to my knowledge, taken up by advocates of the first premise. First, it is not clear what figures as the reason when we say pleasure provides the reason, i.e. what ontological category we opt for (this is the element I marked as ‘q’ in the structure of practical reason in Chapter I). Since many who hold this view do not like the idea that reasons are mental states, they should try to say that it is the proposition or the state of affairs involving pleasure that serve as reason here. This is not an unprecented view; cf. Katz (1986), Chapter 4, pp.

105-6. The second question concerns the meta-ethical background of the idea. The issue here is whether non-naturalists can also claim that pleasure provides reason for action that is not based on desire. They would have to say that the property of value is still a non-natural property that supervenes upon the phenomenology of pleasure. This is an important question because many who like the idea that pleasure is reason-providing just by virtue of its qualia are not naturalists. Cf. also Sobel (2005), pp. 442 note 12.

56 There are other issues involved here. First, there is the question whether pleasure is a general concommittant of desire-satisfaction and frustration of its non-satisfaction. There is evidence that the answer is negative. See Kagan (1992), pp. 170; Raz (1999a), pp. 58-9; Heuer (2004), pp. 50; Katz (2006), pp. 22-3 and (forthcoming); also Schroeder (2004), pp. 27-35 who brings examples from neuropsychology in addition to the cases I mention here. I may be unaware that something I strove to achieve was realized as a direct result of my activity. Due to my lack of awareness I did not experience pleasure; yet, my desire was satisfied. To claim the opposite, one would have to claim that satisfaction of desire is intended in a psychological sense, i.e. the agent must feel that his desire was satisfied.

But this is certainly not the case. Satisfaction here is understood in the logician’s sense: the question is simply whether or not the state of affairs that form the object of the desire obtains. As to the other side of the relation, it is also possible to experience pleasure without preceding desire: pleasant surprises, for instance, are not necessarily preceded by the agent’s desiring the object of the surprise. Nor need frustration follow the non-satisfaction of desire.

For instance, if the opportunity for satisfying a desire does not arise, the agent need not feel frustration over this.

Second, there is the question whether pleasure is always among the reasons to have a desire. However, it is easy to find real life examples that confirm the opposite conclusion. Many of our desires, especially altruistic ones, are based

Attila Tanyi, THESIS, Page 112

pleasurable experiences share a phenomenological commonality, which we call ‘pleasure’. This common core and pleasure, moreover, are only contingently related, both metaphysically and causally. That is, it is possible both in some possible world as well as in our actual world that a pleasurable experience is not liked by the agent. Consequently, the value and reason-giving power of pleasure lies in this phenomenology and not in the agent’s desire. Parfit clearly rejects this idea. His claim about constitution puts him at the other end of the spectrum: he assumes a particular theory of pleasure, one we might call desire-based.⁵⁷ (cf. also Parfit 1984, 493) This theory holds that the only thing common to pleasurable experiences is that they are liked by the agent, while they are occurring. Accordingly, in its account of pleasure’s value and reason-giving power this theory attributes a central role to desire: it is the desire that confers value on pleasure and makes it reason giving.

Parfit, however, nowhere defends his choice. For a proper defense we have to turn to David Sobel. Sobel’s strategy is to show that the desire-based theory is the best available account of pleasure. His first target is the idea I described above, which I will call the phenomenological account. His rejection rests on two reasons. One is the traditional objection that no matter how carefully we introspect our mental life we cannot find the common phenomenological state the alternative account claims to exist. (Sobel 1999, 230; 2005, 444) But his real problem is not this.

(Copp and Sobel 2002, 271-2; Sobel 2005, 444-5) On the phenomenological account, he claims, it can happen that a phenomenological state occurs without the agent’s having a favourable response to it. This, we saw, is indeed the case. And, Sobel goes on, advocates of this account

on object-given intrinsic reasons other than pleasure. The desire to end famine in Africa, or the desire to help the drowning boy are all instances of this phenomenon: it is the fact that action would bring famine relief, or save the boy that gives us reason to desire it and not the fact that it would bring us pleasure or save us pain. For further examples see Sidgwick (1907), pp. 49-52; also Quinn (1993), pp. 243.

57 Alternatively, Parfit could argue for a hybrid account on which pleasure is constituted by the agent’s desire in one case but is not so constituted in another case. But this would be just another choice from the possible theories of pleasure; in fact, a very difficult and unusual choice. Hence in the text I assume that he would go for the simpler and more general desire-based account of pleasure.

Attila Tanyi, THESIS, Page 113

claim that the state gives reason to the agent to experience it even in this case. This, again, is true.

But, Sobel says, this is implausible. For this implies that an ordinary sensation, which pleasure is taken to be on this picture, is reason providing even when it is not liked by the agent. The analogy here, Sobel claims, is with sensations like pins and needles or the taste of chocolate or of strawberry ice cream. But to hold that sensations like these can provide reason for agents to experience them regardless of the agents’ response to them is something we say “only when joking”, he says.

However, even if we accept Sobel’s rejection of the first alternative, we could still opt for a theory, which attempts to strike a middle ground between the phenomenological and the desire-based account. The idea is to accept that pleasure and desire are inseparable: an experience is pleasurable (if) only if the agent is desiring it while it is occurring. But then add that this is only because desire is taken as a response to or evidence of pleasure’s value where pleasure is still regarded as a phenomenological state common to all pleasurable experiences. The result is a version of the idea, mentioned in Chapter I, that desire can co-vary with valuable states without itself conferring value on them. As Scanlon puts it, “When we have reason to bring about an experience in virtue of its being pleasant, what we have reason to bring about is a complex experiential whole that involves, say, having a certain sensation while also desiring that this sensation occur. So these cases remain ones in which the quality of the experience (considered broadly) is a reason to bring it about, rather than cases of having a reason to do something because it will fulfill some desire.” (Scanlon 2002, 340; cf. Chang 2004, 64, note 13)

But Sobel is not happy with this more concessive approach either. His refutation takes the following form. (Sobel 2005, 449; cf. Trigg 1970, 121-3) He argues that if the desire-condition is introduced the phenomenological state cannot ground the reason because it has no content other than being a state that is desired (by ‘content’ Sobel must mean phenomenological, not

Attila Tanyi, THESIS, Page 114

propositional content). His explanation is this. Reasons do not supervene on their grounds, i.e. the requisite phenomenological state can occur without being desired by the agent: hence no pleasure, hence no reason. This is a problem, he goes on, because we don’t know what, in the absence of desire, could ground the reason. All we know is that the phenomenological state counts as pleasure and provides reason when desired but does no such thing when not desired. On closer inspection, therefore, the second approach collapses into the desire-based account.

To this, however, an advocate of the second approach can say that Sobel is mistaken: there is desire-independent content here provided by the phenomenological state that this approach also invokes.⁵⁸ The reason why I think Sobel does not consider this response is because he thinks it only produces an insolvable dilemma for the responder. (cf. Sobel ibid. 451) If, in response to his claim, we say that the phenomenological state does have content other than being desired, we will be forced to endorse the phenomenological account. We will have to say that the phenomenological state provides reason even in the absence of corresponding desire. If, on the other hand, we deny that the phenomenological state has any such independent content, we will have to accept Sobel’s conclusion. The first option, as has been argued previously, is implausible, whereas the second collapses into the desire-based theory. Hence, in the absence of powerful rivals, the desire-based account is the best available theory of pleasure. The first premise is in trouble again.

58 Sobel here clearly repeats the argument of Sobel (1999), pp. 232-3, but that argument was premised on the fact that the phenomenological view of pleasure is rejected. Sobel there was considering a view according to which pleasure is not a feeling state but is merely a non-sensory dimension of experience, i.e. that it has no phenomenological content, which is inseparably related to desire. He attributed this view to Kagan (1992), pp. 172-9 and Katz (1986), Chapter 2, 47-8, Chapter 4, pp. 105, 112, rightly in the first case, wrongly in the second, in my view. Once we introduce a phenomenological element, however, we will give desire-independent content to the concerned phenomenological state thereby making ground for the response considered in the text.

Attila Tanyi, THESIS, Page 115

There are problems with both parts of Sobel’s argument.⁵⁹ Start with his rejection of the phenomenological account. His first point is that our failure of introspection shows that there is no phenomenological state common to all pleasurable experiences. But today’s science and philosophical research both have resources to remedy this failure of introspective psychology.

(Katz 2006, 1.3, 2.3.4) Without going into details, there is scientific evidence that introspection of affective, as opposed to, for example, sensory experience is especially prone to errors of omission. And, as we shall soon see, there is good reason to think that the phenomenological state we identify with pleasure is best understood as a feeling-state. Furthermore, instead of endorsing the early empiricist, Lockean view that we form the concept of pleasure solely by ostension to mental items discovered exclusively and known exhaustively by introspection, we can opt for a direct reference to pleasure, a shared but fallible ability preserved since early childhood. Finally, we can borrow Ned Block’s distinction between phenomenal consciousness and cognitive awareness. The idea would then be to identify pleasure with ‘bare’ pleasure, i.e. an immediate phenomenal experience, instead of our consciousness of pleasure. If this is accepted, failure of

59 In addition, there is a problem with the positive part of Sobel’s argument, namely his endorsement of the desire-based theory. For the kind of desire pleasure is related to must be a desire that the agent has while the experience is occurring. It is therefore a future desire considered in relation to the act of bringing about the relevant experience.

And, remember, the Model only invokes actual or hypothetical desires of the agent but future desires are not admitted as reason-giving – or, at least, this is what we have assumed so far. Scanlon (2002), pp. 339 notes this problem claiming that though this extension of the Model is ‘quite coherent’ (?), it does eliminate the connection between reason and motivation, which advocates of the Model typically want to maintain. And Parfit (2001), pp. 21 seems to suggest that for him the extension is unacceptable because he thinks that reasons that are allegedly given by future desires are actually value-based reasons. This also explains why Parfit, who is otherwise a defender of the first premise, allows for the exception of hedonic desires. Both Chang (2004), pp. 77-9 and Sobel (2005), pp. 454-5 take up this issue, but their responses are not convincing. Sobel’s brief remark is just unclear to me, whereas Chang’s footnote mentioned she only says that the same strategy can be employed here too; yet, she does not actually provide the argument. Hence, until this argument is provided, I regard this problem as not yet adequately answered. (We should also note that in her discussion of Parfit Chang takes for granted the desire-based theory of pleasure, a surprising move given her earlier rejection of Sobel’s argument on the ground that something like the second, Scanlonian approach to pleasure is right.)

Attila Tanyi, THESIS, Page 116

introspection poses no problem since by construction we don’t have direct cognitive access to phenomenal consciousness.

Sobel’s implausibility charge fares no better. First of all, in his treatment he identifies garden-variety sensations, such as pins and needles or taste of chocolate, with pleasure. But among those who adhere to the phenomenological view of pleasure many deny that the two are the same; and others, who otherwise are not advocates of the theory, also reject the analogy (for a good overview see Katz 2006, 1.1. and note 3). The more accepted view is that pleasure is a simple and undescribable feeling in momentary consciousness. (Katz 1986, Chapter 2, 47-8;

2006, 1.1) And if we make this distinction between sensation and affective experience, it is not necessarily implausible to hold that pleasure gives reasons for agents to experience it just by virtue of its intrinsic nature, i.e. just by the way it feels. Furthermore, Sobel also forgets to distinguish the non-hedonic aspects of an experience from its hedonic aspect. But as C. D. Broad has pointed out, it is possible to claim that the value of a pleasurable experience stems not solely from its hedonic aspect (which Broad takes to be a quality but we need not to), but rather from the combination of this aspect with the other non-hedonic aspects (making up, I take it, some kind of an organic experiential whole). (Broad 1930, 232-7) Hence the claim is not that e.g. the taste of chocolate understood as a sensation in our mouth is valuable without our favorable response to it. Rather, the idea is that it is valuable because it combines this sensation with the hedonic and other non-hedonic aspects of the same experience.

Finally, there is a way the phenomenological account can integrate the second reading. It can do this with the help of a distinction between immediate liking and full desire. There is evidence that ‘liking’ and ‘wanting’ are different neural core processes where the former is not necessarily conscious (though may be phenomenally conscious: see Katz 2005 referring to Block’s distinction above), but may enter into larger ones, such as the latter, that are. (Katz 2006,

Attila Tanyi, THESIS, Page 117

3.3 referring to Berridge 2004) And Leonard Katz points out that liking, unlike wanting is a pre-intentional state, which need not have an object, and which can but need extend to, pre-intentional and object-bound, desire. (Katz forthcoming; 2006, 2.3.3 and note 35; Scanlon 2002, 340 also seems to suggest this when he says that the appetitive states of human infants and animals are different from ordinary desires) Of course, this distinction between liking and wanting takes us beyond the widely-shared functionalist theory of desire since these states presumably have the same direction of fit. Nevertheless, I suggest that, for the sake of argument and further insight, we assume for the moment that the distinction can be made.

Then two further ways of understanding the phenomenological account becomes open to us. In one case we understand pleasure as an experience that contains within itself an immediate liking of itself by itself.⁶⁰ (Katz 1986, Chapter 2, 48; Chapter 4, 104 may suggest this; though cf.

Katz 2006 note 35) Here it is a question how such a self-referential liking can be objectless. But

Katz 2006 note 35) Here it is a question how such a self-referential liking can be objectless. But