In his recent book Pleasure and the Good life, Fred Feldman endorses two propositions that together entail a theory of happiness. The first is what Feldman calls ‘eudaimonism’: that the good life is the happy life. He restates eudaimonism as the proposition ‘that the value in itself of a life for the person who lives it is directly proportional to the net amount of happiness that the person enjoys throughout her life'’.Footnote ¹ In the context of the discussion in the book, Feldman regards eudaimonism as platitudinous but unhelpful for understanding the good life because he finds the concept of happiness obscure. Hence, he sets eudaimonism aside and argues instead for an understanding of the good life he calls ‘intrinsic attitudinal hedonism’.Footnote ²

My interest in this article is in the theory of happiness entailed by eudaimonism and intrinsic attitudinal hedonism: namely, that a person, S, is happy at time, t, just in case S's intrinsic attitudinal pleasure at t is greater than S's intrinsic attitudinal displeasure at t. This understanding of happiness is ‘Feldman's Theory of Happiness’ (or FTH).Footnote ³

I find the idea that motivates FTH initially compelling, but FTH itself fails to capture this idea. FTH presupposes a distinction between intrinsic and extrinsic attitudinal pleasure and displeasure, and the theory excludes extrinsic attitudinal pleasure and displeasure from the aggregation that determines happiness. This is directly contrary to the underlying hedonist idea that drives the attempt to understand happiness as the aggregation of the attitudinal pleasure and displeasure a person has at a time. One might think that Feldman has provided a strong reason for his departure from the underlying idea, but I shall argue that he has not.

The primary technical term Feldman employs to state FTH is ‘attitudinal pleasure’. This term is supposed to denote an attitude expressed by certain familiar and ordinary sentences. Feldman says that ‘attitudinal pleasures are always directed onto objects, just as beliefs and hopes and fears are directed onto objects.’Footnote ⁴ He himself understands the object to be a state of affairs. He says that a ‘person takes attitudinal pleasure in some state of affairs if he enjoys it, is pleased about it, is glad that it is happening’.Footnote ⁵ One might have thought that the object is a proposition,Footnote ⁶ but nothing in my argument against FTH depends on the ontological distinction, if any, between a proposition and a state of affairs. I assume only that Feldman is correct about this much: that English speakers attribute attitudinal pleasure and displeasure to subjects by using sentences of forms: ‘S is pleased/glad that P’ and ‘S is displeased/sad that P’.Footnote ⁷

Once it is assumed that these attitudes are expressed by such sentences, it is crucial to have a firm grip on when, in a given situation, a natural use of a sentence of the form ‘S is pleased/glad that P’ is true. I do not question the truth conditions themselves. I assume that the sentence is true if, and only if, S stands in the relation of attitudinal pleasure to the state of affairs (or proposition or whatever) expressed by P. My interest here is not in the truth-conditions, which I grant, but in when these sentences are false but convey useful information.

Feldman himself does not address this issue directly, but there is a clear indication of his view in what he says about other matters. A case in point is his argument to show that attitudinal pleasure is not factive. The question of the factivity of attitudinal pleasure is obviously interesting in its own right, but I do not reconstruct Feldman's argument against factivity to answer this question. My aim, in reconstructing the argument, is to show how the natural use of sentences of the form ‘S is pleased/glad that P’ in various circumstances leads Feldman to see attitudinal pleasure where there may be none. It seems to me that this tendency is a large part of what motivates his belief in attitudinal pleasure extrinsic to happiness.

The question of the factivity of attitudinal pleasure is the question of whether it is possible to stand in the relation of attitudinal pleasure to a given state of affairs even though that state of affairs is not true. One might well think, at least initially, that this indeed is not possible. It sounds decidedly odd to say of a given person that he is pleased that P but that not P, for example, that S is pleased that it is raining but that it is not raining. Feldman, nevertheless, argues that the attitudinal pleasure is not factive. He argues that the subject of the attitude need only stand in the relation of belief to the object of the attitude. To make his point, he describes a case in which he mistakenly thinks that he will soon meet G. E. Moore:

In this example, as the circumstances are set out, Feldman assumes that he is pleased about something and that the problem is to identify the state of affairs he is pleased about. Feldman notes, correctly I think, that it is more natural for him to utter

1. (1) ‘I am pleased that I am going to meet Moore.’

than it is to utter

1. (2) ‘I am pleased that I think I am going to meet Moore.’

Hence, since he is pleased about something, Feldman decides that the more natural utterance is true and that he stands in the relation of attitudinal pleasure to the state of affairs.

To understand Feldman's argument, it is important first of all to set aside a possible source of confusion in the use of the phrase ‘going to’. Some philosophers of language have plausibly maintained that sometimes the future tense is used to express present tendencies. So, to borrow an example from Michael Dummett,Footnote ⁹ one might say ‘Feldman was going to meet Moore, but the meeting will not now take place’. To ensure that Feldman's example does not trade on such uses, it must be restated without this phrase. The claim, in the prior quotation, in need of reconsideration is the following:

In respect to this claim, there are two points to consider. The first is about what Feldman could natural report about himself. The other is about what some third party could naturally report about Feldman. Given the example, it would sound natural enough for Feldman himself to use the following sentence

1. (3) ‘I am pleased that I shall meet Moore soon.’

It would sound much less natural for a third party to use the following sentence

1. (4) ‘Feldman is pleased that he will meet Moore soon, but Feldman will not meet Moore soon.’

The proper conclusion to draw from the oddity of such a report is not immediately clear, but certainly one possibility is that the proposition (3) expresses is false. Feldman, in the example, may think he has identified what he is pleased about, but perhaps he is mistaken.

The pressure not to draw this conclusion stems from two sources. Feldman has stipulated that in the example there must be some state of affairs in which he takes pleasure. In the example, he is smiling and showing others signs of being pleased. Feldman assumes that these signs really do indicate that he is pleased and hence, given his conception of happiness in terms of attitudinal pleasure, that there must be some state of affairs in which he takes pleasure. Moreover, in the example, it would not be natural for Feldman to use

1. (5) ‘I am pleased that I believe that I shall meet Moore soon.’

to report the state of affairs in which he takes pleasure. Hence, Feldman decides that attitudinal pleasure is not factive. He concludes that (3) expresses the true proposition that he takes attitudinal pleasure in the false state of affairs that he will meet Moore soon.

Whether or not Feldman is correct, he seems to assume that sentences of the form ‘S is pleased/glad that P’ and ‘S is displeased/sad that P’ are true just in case their use would be natural in the circumstance. This assumption is important to notice because it seems to motivate his attempt to deny that happiness at a moment is simply a matter of aggregating the amounts of attitudinal pleasure and displeasure the person takes at that moment. The problem is not with the notion of ‘amounts’. Feldman himself says that he believes ‘it makes sense to speak about ‘amounts’ of attitudinal pleasure’.Footnote ¹⁰ Hence, on the assumption that positive (negative) numbers can be assigned to measure the amount of attitudinal pleasure (displeasure) a person takes at a moment in a given state of affairs, one would naturally expect total happiness to be understood as the sum of attitudinal pleasure minus the sum of attitudinal displeasure.Footnote ¹¹

Feldman, however, maintains that this conception of happiness is subject to counterexample because it fails to incorporate a distinction within attitudinal pleasure. The distinction is between what he calls intrinsic attitudinal pleasure and extrinsic attitudinal pleasure. This distinction is not the distinction between intrinsic and extrinsic value. The person takes pleasure in some state of affairs, but the person is supposed to take this attitudinal pleasure in the state of affairs only ‘because he thinks it is related to some other state of affairs, and he takes pleasure in that other state of affairs’:

Although extrinsic attitudinal pleasure is attitudinal pleasure, only intrinsic attitudinal pleasure is supposed to be relevant to the aggregation that constitutes happiness. Feldman supposes that extrinsic attitudinal pleasure is not included in the aggregation of total happiness because otherwise certain intuitive judgments about happiness would be false.

Feldman's beer-and-peanuts example, in the passage above, is supposed to establish the need to distinguish between intrinsic and extrinsic attitudinal pleasure. The person who sees the waiter heading to the table and who has announced that he is pleased to see the waiter coming is supposed to be happy that he will soon be eating peanuts and drinking beer. Hence, it is supposed to be a mistake to include the attitudinal pleasure he takes in seeing the waiter come to the table in the aggregation of attitudinal pleasure that determines his happiness.

I do not find this example persuasive. The distinction between intrinsic and extrinsic attitudinal pleasure, as Feldman describes it, depends solely on what the person ‘thinks’ about the connection between two states of affairs. I find this defining feature somewhat difficult to grasp, but whatever it is exactly, it is hard to see how the presence or absence of this feature matters. The real question is whether the person in fact takes attitudinal pleasure in the state of affairs. If the person takes pleasure in the state of affairs, it is wrong to exclude it from the calculation of his happiness. If, alternatively, he does not take pleasure in the state of affairs because he is not pleased to see the waiter coming, then the announcement Feldman takes to show that there is attitudinal pleasure taken in this state of affairs shows no such thing. In this case, the proposition expressed by the sentence used in the announcement is false.

When I say that the proposition expressed by the sentence used in the announcement is false if the person is not pleased to see the waiter coming, I do not mean to deny either that the utterance is natural or that it conveys useful information. It would be perfectly natural for the person to utter the following sentence in conversation with his companions:

1. (6) ‘I sure am pleased to see the waiter heading to our table.’

There is no question that this would be an ordinary thing to say. The only question is whether such an utterance should be understood to show that the subject stands in the relation of attitudinal pleasure to the state of affairs that the waiter is heading for the table. As far as I can tell, Feldman assumes that such an utterance is true and concludes that the attitudinal pleasure is extrinsic for only one reason: namely, that if the person were asked why he said he was pleased to see the waiter heading to the table, he might well say something like

1. (7) ‘Seeing the waiter coming convinced me that soon I will be eating peanuts and drinking beer, and I am happy about that.’

In itself, however, this report is not evidence for the truth of (6). It might be that (6) is false and that the person utters the sentence to indicate that seeing the waiter caused him to believe something about which he is happy: that soon he will eat peanuts and drink beer.

The other case Feldman cites in this passage he describes as posing a ‘compositional’ problem, not the more familiar ‘instrumental case’. I am not sure what the agent ‘thinks’ such that his attitudinal pleasure is extrinsic, but the idea would seem to be that the agent takes attitudinal pleasure in a certain conjunctive state of affairs only because he ‘thinks’ that certain states of affairs in which he takes attitudinal pleasure jointly entail this conjunctive state of affairs. In the beer-and-peanuts example Feldman sets out, the conversation would naturally go like this. The waiter, making his rounds, says ‘How is everything here?’. The patron replies

1. (8) ‘The food was sure a long time in coming, but now I am pleased that I am eating peanuts and drinking beer.’

This would be natural to say to the waiter. The question is whether it is true. Since Feldman himself stipulates that the pleasure is not in an ‘interesting combination of tastes’,Footnote ¹³ one would think that the ‘I am pleased that . . .’ conjunct in (8) expresses a false proposition. Feldman himself thinks otherwise, and, as far as I can tell, the only reason is that, in the example, the person might naturally utter (8) even though pleasure in the conjunctive state of affairs is not part of his happiness. If the person were asked why he said he is pleased that he is eating peanuts and drinking beer, he would (were he to indulge the question) say something like

1. (9) ‘I like peanuts. I like beer. I do not enjoy the tastes simultaneously. I like them one after another. Now I am happy that I am tasting these peanuts. Now I am happy that I am drinking this beer.’

Yet, in this case, as in the previous one, (9) is not evidence for the truth of (8). The person might utter (8) with the expectation that he will be understood to mean that (9) is true.

Feldman does not consider this possibility. He assumes that natural uses of sentences of the form ‘S is pleased/glad that P’ express truths. He does not consider the possibility that such sentences might be used to convey helpful information even though strictly speaking, in the circumstance of use, they express false propositions. He rejects the straightforward aggregation conception of happiness in favor of FTH. In FTH, rather than aggregate all attitudinal pleasure and displeasure, only intrinsic attitudinal pleasure and displeasure is aggregated.

I have argued that this rejection is not warranted. The underlying idea, the idea driving this whole approach to understanding happiness in the first place, is that happiness is an aggregation of attitudinal pleasure and displeasure. I find this idea initially plausible, but I have not argued for it here. It is common ground among friends of the attitudinal hedonist conception of happiness. The question I have considered in this article is whether this common ground is overturned by the examples Feldman provides and hence has to be refined in terms of a distinction between intrinsic and extrinsic attitudinal pleasure and displeasure. I believe the answer is ‘no’. Feldman has not demonstrated a need to distinguish extrinsic from intrinsic attitudinal pleasure and displeasure so that happiness may be understood, along the lines of FTH, as an aggregation of intrinsic attitudinal pleasure and displeasure only.Footnote ¹⁴