A New and Improved Whole Life Satisfaction Account of Happiness Suikkanen, J. (2011). An improved whole life satisfaction theory of happiness. International Journal of Wellbeing, 1(1), 149-166. doi:10.5502/ijw.v1i1.6 Jussi Suikkanen University of Birmingham jussiphil@gmail.com Copyright belongs to the author(s) www.internationaljournalofwellbeing.org 149 ARTICLE An improved whole life satisfaction theory of happiness Jussi Suikkanen Abstract: According to the popular Whole Life Satisfaction theories of happiness, an agent is happy when she judges that her life fulfils her ideal life-plan. Fred Feldman has recently argued that such views cannot accommodate the happiness of spontaneous or preoccupied agents who do not consider how well their lives are going. In this paper, I formulate a new Whole Life Satisfaction theory that is not vulnerable to this objection. My proposal is inspired by Michael Smith’s advice-model of desirability. According to it, an agent is happy when a more informed and rational hypothetical version of her would judge that the agent’s actual life matches the best life-plan for her. I will argue that my new Whole Life Satisfaction theory is a flexible model that can avoid many of the problems besetting previous theories of happiness. Keywords: happiness, whole life satisfaction, conditional fallacy, advice-model 1. Introduction The ordinary concept of ‘happiness’ is ambiguous.1 It is often merely used to describe how a person feels, i.e., a particular phenomenological state of the person. This is the state we are in when we feel contentment, satisfaction, euphoria, and the like. Most of us have had direct experiences of feeling happy in this way. For instance, we usually feel happy when we unexpectedly see an old friend whom we haven’t seen for a while. When the term ‘happiness’ is used in this first sense, nothing more is required to be happy than to feel happy or to be in a happy mood. Hereafter, I will use the sub-indexed term ‘happinessp’ to refer to this kind of happiness. This is because sometimes it is also called ‘psychological’ (or ‘local’) happiness. The second sense of ‘happiness’ is often said to be ‘deeper’ than the first one.2 It encompasses much more than agents’ mere momentary phenomenological states. Consider a 1 This ambiguity is recognised and elucidated in different ways by Feldman (2008, sec. 1; 2010, sec. 1.2), Foot (2001, ch. 6), Haybron (2000, sec. 2.1; 2003, sec. 1; 2008, pp. 5–6, and sec. 2.2), Kraut (1979, p. 168), Nettle (2005, pp. 17–20), Sizer (2010, pp. 133–134), Tatarkiewicz (1966, p. 1; 1976, pp. 1–6), Telfer (1980, pp. 1–2),Thomas (1968), and von Wright (1963, p. 87). Feldman (2010, pp. 131–136) argues against the consensus by claiming that we should understand ‘happiness’ as a very vague term. 2 For methodological reasons, I cannot offer a clear-cut definition of the term ‘happiness’ when it is used in this second sense. This is because I am not stipulating a new technical meaning for the term. Rather, the considerations in this section suggest that ‘happy’ is used in ordinary language to make judgments that are neither about pleasure nor about wellbeing. For reasons familiar from Wittgenstein (1958, §65–§71) and Williamson (2007, ch. 3), I doubt whether this usage could be captured with any obvious and simple definition of the term. Instead, we can investigate what the property of being happy could be such that it would make the core uses of the term in the ordinary language true. The most popular suggestions of this property are given in Section 2, and my own theory will be explained in Section 5. The bi-conditionals in these sections could also be understood as hermeneutic, non- obvious, and non-stipulative definitions of the term ‘happy’ when it is used in this second sense. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 150 case in which a person uses drugs in order to feel euphoric. In such cases, we might hesitate to say that this person becomes genuinely happy even if she feels happy as a result (Sizer, 2010, p. 137, and p. 139). Happiness in this second sense seems to thus require more than merely experiencing (perhaps fleeting) sensations of happinessp. We typically use the term ‘happiness’ in this second way when we assess a longer period of an agent’s life. For instance, when we talk about a person having been happy in the 1990s, we do not mean just that the person had many euphoric experiences during that decade.3 In fact, some people may feel miserable for much of the happy periods of their lives (Feldman, 2002, sec. 4). For instance, consider the life of an Olympic athlete who undergoes a gruelling four- year training regime in order to prepare for the next games. Whatever the outcome of the Olympics is for the athlete, it is conceivable that she later thinks of this period in her life as a happy one. After all, she was able to focus wholeheartedly on what she wanted. Hereafter, I will use the term ‘happinessg’ to refer to happiness used in this second sense. This is because sometimes this sense of happiness is called ‘global happiness’. In addition, some philosophers use the term ‘happiness’ in a third sense (Feldman, 2008, p. 219). When they talk about the happiness of an agent, all they mean is that the agent is living a good life. Furthermore, they understand a life to be a good life when it has a high level of wellbeing (or a lot of ‘prudential value’). I will use the sub-indexed expression ‘happinessw’ for this third use of the term ‘happiness’. It is also sometimes called ‘eudaimonic happiness’ or ‘prudential happiness’.4 I will assume hereafter that this third sense of ‘happiness’ is distinct from the other two senses of the term. It seems conceivable that a person’s life can be happy in the other two senses even if her life lacks the normal constituents of wellbeing.5 For instance, it is often thought that, the further we go back in history, the lower the level of general wellbeing was. Despite this, it is reasonable to think that people were not necessarily less happyp, or even less happyg. So, to summarise, it seems like being happyg requires more than just feeling happy (i.e., being happyp). Yet, it does not seem to necessarily require as much as living a life with a high level of wellbeing (i.e., being happyw). My main goal is then to consider what is required for happinessg. What could the middle ground between happinessp and happinessw be? What constitutes being happy in that sense? I will begin by introducing, in Section 2, different versions of the most popular theory of happinessg – the Whole Life Satisfaction theory.6 Section 3 will then explain Fred Feldman’s recent counter-examples to all the different versions of that theory (Feldman, 2008; Feldman, 2010, ch. 5). Section 3 also argues that these cases are based on the so-called conditional fallacy, which is a typical flaw in philosophical theories based on subjunctive conditionals. Section 4 will then look at how the conditional fallacy has been solved by Michael Smith in the context of theories of desirability, which have often been based on subjunctive conditionals (Smith, 1995, sec.1). By using Smith’s ‘advice-model’, I will then construct, in Section 5, a Whole Life Satisfaction theory with a new and improved structure. I will argue that the resulting theories 3 Philippa Foot (2001, p. 85) claims that this is what Wittgenstein meant on his death-bed, when he said that he had had a wonderful life. 4 Foot (2001, ch. 6), Murphy (2008), and Taylor (2002) are some of the philosophers who use ‘happiness’ in this sense. 5 However, Crisp (2006, ch. 4) argues that happinessw consists of instances of happinessp. 6 In different forms, these views has been defended, for example, by Benditt (1974), Brandt (1967), Kekes (1982), Kraut (1979), Montague (1966), Nozick (1989), Sumner (1996, pp. 145–146), Tatarkiewicz (1966), Telfer (1980, pp. 8–9), Thomas (1968), and von Wright (1963, pp. 98–99). Whole Life Satisfaction models are also often used in empirical research (see, for instance, Veenhoven (1984)). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 151 of happiness with the ‘advice-structure’ will be able to avoid Feldman’s objections to the Whole Life Satisfaction theories. Finally, Section 6 will discuss some of the advantages of the resulting view and respond to some of the objections to it. 2. Whole Life Satisfaction theories of happiness So, why is being happy during a longer period of time neither a matter of just feeling happy nor of merely living a life with a high level of objective wellbeing? According to the most popular theory of happinessg, this is because at the heart of happiness lies the agent’s own conception of how her life is going. The idea is that everyone has a set of goals. Even if some of these goals change at different stages of our life, they still tend to be fairly stable. In the past and the present, we have either achieved some of our goals, or we would have wanted to have done so.7 And, the rest of our goals we will want to achieve in the future. The plans that lead to satisfying as many of these goals as possible during the different stages of our lives can be said to constitute our ‘ideal life- plans’. We also typically have some conception of what has happened to us in our lives, and we have some beliefs about what will happen to us in the future. In this situation, we can make judgments about whether our lives match up with our ideal life-plan. These judgments form the essence of Whole Life Satisfaction theories. Even though it is possible to formulate many different versions of the Whole Life Satisfaction theory from the previous core idea, I will focus here on the three main alternatives. They all develop, in slightly different ways, the platitudinous idea that a person is happy when she is satisfied with her life. 2.1 Cognitive Whole Life Satisfaction theories8 A simple version of a Cognitive Whole Life Satisfaction (CWLS) theory might claim that a person, S, is happyg to degree n at time t if and only if:9 1) There is a certain life that S has lived up to t, 2) at t, S has an ideal life-plan for her life, 3) at t, S has a moderately detailed conception of how her life has transpired so far, and 4) at t, S judges that her actual life so far matches her ideal life-plan to degree n. According to this proposal, being happyg consists of being in a cognitive state that represents how well an agent’s actual life matches up to her life-plan. As a result, according to CWLS, she cannot be happy without actually believing that her life satisfies her ideal life-plan. Accounts like CWLS may seem at least somewhat plausible. I have a conception of what an ideal life for me would be like. On the basis of this conception, I am able to formulate a life-plan which might consist of having a safe childhood, getting a good education, having friends and a family, having a meaningful career, and so on. If I also believe that so far my life, by and large, fits this plan, then it seems reasonable to believe that I am happy. 7 It is worth pointing out that many of these goals are to engage in a variety of activities rather than to bring about some end-state. Because of this, the resulting life-plans might be more a matter of living in certain ways rather than accomplishing certain feats. 8 Versions of this view have been defended by Kekes (1982), Thomas (1968, pp. 104–108), and von Wright (1963, pp. 98–99). 9 The formulations of all of the theories in this article are based on Feldman (2008; 2010, ch. 5). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 152 CWLS also has another advantage (Feldman, 2008, sec. 4; Feldman, 2010, sec. 5.4). We can create many different versions of this simple view. Each one of them will classify agents as happy in slightly different ways. We can change how complete the agent’s life-plan must be, how accurate a conception she must have about her actual life, whether all aspects of her life must fit the plan or only the most important ones, whether the judgment in clause 4 must be about her whole life or a part of it, and so on. By changing these aspects of the theory, we can seek a reflective equilibrium between CWLS and our intuitive judgments about which agents are happy. 2.2 Affective Whole Life Satisfaction theories According to some philosophers, CWLS cannot be the true theory of happiness, because it would make happiness too intellectual, i.e. merely a matter of having certain beliefs.10 They argue that whether an agent feels pleased by how she lives is more important than any beliefs could ever be. Hence, according to the views based on this idea, happiness is a specific kind of positive affective state based on an agent’s conception of her life. This causal origin of the relevant feeling of contentment is then taken to distinguish happiness g from the mere feelings of happinessp. It is worth emphasising that, according to these views, the relevant affective state must at least result from the agent’s implicit and vague assessment of how well her life is going overall. However, on these views, the prior cognitive judgments that must prompt her to have the relevant positive attitudes are not themselves constituents of her happinessg. I will call views of this kind Affective Whole Life Satisfaction (AWLS) theories.11 One formulation of such a view might claim that a person, S, is happyg to degree n at time t if and only if: 1) There is a certain life that S has lived up until t, and 2) at t, S takes pleasure of degree n in the fact that she has lived that life so far. 2.3 Hybrid Whole Life Satisfaction theories Finally, some hold hybrid views, according to which happinessg requires both a cognitive judgment of how an agent’s life matches up to her life-plan and a positive affective state based on that judgment.12 A Hybrid Whole Life Satisfaction (HWLS) view might claim that a person, S, is happyg to degree n= f(p, q) at time t if and only if: 1) There is a certain life that S has lived up to t, 2) at t, S has an ideal life-plan for her life, 3) at t, S has a moderately detailed conception of how her life has transpired so far, 4) at t, S judges that her actual life matches her ideal life-plan to degree p, and 5) at t, S takes pleasure of degree q in her judgment that her life measures up to her life-plan to the degree p. Thus, according to HWLS, the degree of an agent’s happinessg (expressed by ‘n’) is a function of two variables (‘f(p, q)’): the degree to which she believes that her life matches her life-plan (‘p’), and the strength of the positive attitude that results from that belief (‘q’). And, again, we could further specify different versions of HWLS by changing its variables in various ways. 10 This objection is explained by Haybron (2007, sec. 2; 2008, pp. 14–16). 11 Different versions of this view have been defended by Tatarkiewicz (1966), Telfer (1980, pp. 8–9), and perhaps Montague (1966, p. 96). See also Haybron (2008, p. 82). 12 Views of this type have been defended by, for instance, Brandt (1967, pp. 413–414), and Sumner (1996, p. 145). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 153 Admittedly, CWLS, AWLS, and HWLS are simplified theories of happiness. In order to be plausible, they would have to be further developed in many ways. However, Fred Feldman has recently argued that all theories using this type of structure inevitably face insurmountable problems. In the next section, I will explain what these problems are and how they seriously challenge the plausibility of CWLS. I will return to AWLS and HWLS at the end of section 5. 3. Feldman’s dilemma There is both an ‘actualist’ and a ‘hypotheticalist’ version of CWLS. Feldman has argued that both of those theories lead to implausible conclusions about who is happy (Feldman, 2008, p. 230; Feldman, 2010, sec. 5.5). I formulated CWLS in the actualist way above. According to that formulation, being happy g requires that an agent has actually (a) formed a life-plan and a conception of how her life has transpired so far, and (b) compared the two. Feldman has argued that these requirements make achieving happiness in the second sense too demanding (2008, pp. 230–232; 2010, pp. 82–84).13 It seems intuitive that an agent can be happy without having formed a life-plan or a detailed conception of what has happened in her life, let alone without having made any comparisons between the two. It is reasonable to claim that anyone who is unreflective in this way could be just as happy as someone who has extensively compared her life-plan with her actual life (and found that they accord).14 There is also a risk that the actualist CWLS makes certain central varieties of happiness g unattainable for actual human beings (Feldman, 2004, pp. 231–232; 2010, p. 82). Consider a case in which the actualist CWLS is used to assess how happyg someone’s life is overall from her birth to her death. In this case, the previous theory would require that, in order for the agent to be happyg, she would have to be capable of forming a sufficiently comprehensive awareness of her whole life from its very first day to its end. Feldman seems to be right when he claims that actual human beings are probably not able to form such awareness (ibid.). Hence, one consequence of the actualist CWLS seems to be that no actual human being could live a happy life. This certainly seems to be a problematic consequence of the theory. We could try to avoid these problems by creating a Hypotheticalist Whole Life Satisfaction (HYWLS) theory (Feldman, 2008, p. 234; Feldman, 2010, p. 86). Such a view might claim that a person, S, is happyg to degree n at time t if and only if: 1) There is a certain life that S has lived up to t, and 2) if, at t, S were to form an ideal life-plan for her life, and 3) if, at t, S were to form a moderately detailed conception of how her life has transpired so far, then, 4) at t, S would judge that her actual life matches her ideal life-plan to degree n. On this view, being happyg does not require that an agent has actually formed a life-plan and a conception of her actual life and made a judgment of how the two match up. Rather, all it takes 13 See Haybron (2008, pp. 85–86) for more on this ‘problem of attitudinal scarcity’. Haybron (2007; 2008, ch. 5) also offers many other objections to actualist versions of Whole Life Satisfaction views. 14 Note that unreflective persons can experience a lot of pleasure, and their lives can score highly on objective measures of well-being. This means that these individuals can be both happyp and happyw. This will pose no theoretical problems for any plausible theories of happinessp and happinessw because self-reflection is presumably not a central constituent of either pleasure or objective well-being. Thus, the happiness of self-reflective agents is only a problem for the Whole Life Satisfaction theories of happiness g that require certain actual judgments for an agent to be happy. One alternative for a defender of the Whole Life Satisfaction theory of happiness would be to bite the bullet here and to claim that such agents are not happyg even if they can be happyp and happyw. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 154 for someone to be happyg according to HYWLS is that a certain subjunctive conditional is true of her. It requires only that, if she were to form a life-plan and a conception of her actual life, she would judge that the two match up. This allows unreflective people to be happy because the subjunctive conditional can be true even if its antecedent is false. Even if a person never self-reflected, it could be true that, if she were to do so, she would judge that her life fits her ideal life-plan. Fred Feldman (2008, pp. 234–237; 2010, pp. 87–89), however, has argued that HYWLS faces serious problems of its own, and I am inclined to agree with him about this. In order to illustrate these problems, Feldman introduces two characters to argue for this conclusion: an unreflective Timmy and a contemplative philosopher. Timmy is consistently active and cheerful. This is because he is always engaged in activities that he enjoys. What is special about Timmy is that he never thinks about his own life at all. According to Feldman, it could be true of Timmy in this situation that, if he began to consider his ideal life-plans and his actual life, he would come to judge his whole life to be worthless (Feldman, 2004, p. 235). This could be in part because he might make false judgments about his situation, in part because he might come to hold a demanding conception of an ideal life and find that his life falls short of it, and in part because he might find the whole business of self- reflection itself very depressing. As a result, according to HYWLS, Timmy cannot be happy. However, this seems to conflict with our basic intuitions about Timmy’s happiness. Most of us would want to claim that Timmy does seem to be happy just as he is – he is cheerful and likes the activities he is engaged in. This example shows that the subjunctive conditional, which HYWLS uses to measure happiness, does not seem to track our intuitions about which agents are happy. In Feldman’s second example, the philosopher is deeply engrossed in philosophical deliberation. Because of this, she does not think about her own life unless someone interrupts her. If someone did this, the philosopher would stop thinking about philosophy and start thinking about her own life. This would be annoying for the philosopher. If the philosopher had then formed a life-plan and a conception of her life in these counterfactual circumstances in which she is very annoyed, she would have concluded that her life is not going according to her plan. Instead of philosophising, she would now spend time thinking about her own life. As a result, in light of the subjunctive conditional of HYWLS, the philosopher is not actually happy. Yet, intuitively, there is no reason why she could not be happy now (in the actual life in which she thinks about philosophy and not about her life). After all, it is that life we are evaluating and not the counterfactual situation, in which her philosophical contemplation has been interrupted.15 These cases suggest that the hypothetical HYWLS has unintuitive consequences. We should notice, however, that these cases are so effective because they reveal the so-called conditional 15 The previous two examples are directly from Feldman (2008, pp. 234–237; 2010, pp. 87–89). I have not changed them, because I hope that they enable the reader to assess how well my own Whole Life Satisfaction theory can respond to Feldman’s objections as they were presented. However, it is possible to imagine more vivid cases. Consider the world’s best tennis players, Formula 1 drivers, and fighter pilots. Such activities require extreme focus. It might well be that, if these individuals reflected on their lives, they would lose focus and begin to lose matches, crash cars, and be shot down. If these individuals were to form judgments about their lives, their lives would change for the worse. As a result, in these counter-factual, self-reflective circumstances, they would judge their lives not to fulfil their plans. According to HYWLS, the truth of this counter-factual would mean that the tennis players, Formula 1 drivers, and fighter pilots are not happy even when they are pursuing their careers successfully (because they do not self-reflect). This seems implausible to me. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 155 fallacy committed by HYWLS. This fallacy is a well-known problem for philosophical accounts based on subjunctive conditionals.16 The problem with such accounts is that, when we assess a subjunctive conditional, realising the counterfactual antecedent of the conditional analysans can change the assessed state of affairs. If an agent were to form a life-plan, her life would be different from the one that we are trying to evaluate. The crux of the fallacy is that the changes made in the assessed states of affairs can impinge on the truth-value of our assessments. We cannot rule out that the changed life would be less happy, or happier, than the one we were originally assessing. In the previous cases, we wanted to assess the happiness of spontaneous and philosophising lives, but our account forced us to assess self-reflective lives instead. And, there is no guarantee that the latter lives are as happy as the former. Yet, there are ways in which we can try to reformulate a theory based on a subjunctive conditional so as to avoid the conditional fallacy. In the next section, I will explain one such attempt from the debates about desirability. Before that, it is worthwhile to briefly consider how happyp and happyw Timmy and the philosopher are, and what kind of theoretical implications this might have. Intuitively, Timmy is happyp because of all the pleasure he experiences during his spontaneous activities. The philosopher is a more difficult case with respect to happiness p. It may be that she does not experience any specific sensation of bliss whilst philosophising. However, she might still be in a state of consciousness she desires to be in. The former consideration entails that internalist theories of happinessp would not classify the philosopher as happyp, whereas the latter fact entails that externalist theories would classify her as happyp. We can also stipulate that both Timmy and the philosopher are happyw by the lights of the standard theories of happinessw. According to success theories of happinessw, a person is happyw when her core desires are successfully satisfied. Timmy’s desires change constantly because he finds different enjoyable activities appealing at different times. However, he still always succeeds in satisfying these desires. The philosopher only desires to continue her philosophical investigations, and she succeeds in doing so. For these reasons, both Timmy and the philosopher are happyw according to success theories. Objective list theories of happinessw claim that happinessw consists of having certain goods that are necessary preconditions for being able to pursue any reasonable conception of a good life. These goods may include wealth, health, knowledge, important relationships, autonomy, and so on. We can stipulate that both Timmy and the philosopher have a sufficient amount of these goods.17 Therefore, according to the objective list theories too, they both count as happyw. These considerations suggest one potential way in which the Whole Life Satisfaction theories of happiness might deal with Feldman’s objection. It could be claimed that, when we consider Timmy’s and the philosopher’s happiness, our intuitions are really about either happinessp or happinessw. We have just confused these intuitions for our beliefs about how happyg these agents are. This would mean that HYWLS’s judgment that Timmy and the philosopher are unhappyg could still be correct without conflicting with our intuitive 16 My presentation of the fallacy follows Wright (1992, p. 118). See also Shope (1978). 17 According to some objective list theories, happinessp and/or happinessg also belong to the constituents of happinessw. This might make it true that, depending on our views on happinessp and happinessg, Timmy and the philosopher might not be completely happyw even if they are considerably happyw because they have all the other relevant goods. This makes no difference to what follows below. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 156 judgments about happiness, which in this case happen to be about the other two senses of happiness. I am not convinced that this proposal could deal with the objection. It seems that we can intentionally concentrate on the ordinary sense of happiness, which is neither just the amount of pleasurable experiences nor solely the objective level of well-being. When we focus on happiness in the ordinary sense, many of us would probably want to believe that Timmy and the Philosopher are happy individuals in this very sense of happiness. This would entail that HYWLS still conflicts with our intuitions. If this were true, then we should continue to attempt to formulate a theory of happinessg that better fits our considered judgments about happiness in this sense. I formulate such a theory in the next section. 4. Desirability, the conditional fallacy, and the advice-model Many philosophers have tried to account for desirability in terms of subjunctive conditionals (Rawls, 1971, ch. 7; Brandt, 1979, ch. 1).18 They are motivated by the thought that, even if we usually desire what is intuitively desirable, sometimes we may also desire what is not desirable. Even if I desire to drink the liquid from the glass in front of me, that liquid may not actually be desirable because, unbeknownst to me, it is petrol rather than gin.19 Similarly, drinking gin might not be desirable even if I happen to desire to drink it, because I have a more fundamental desire to work tomorrow. A natural way to avoid such problems is to give an account of desirability in terms of what an agent would desire in a counterfactual situation in which she were fully rational. This leads to theories based on subjunctive conditionals, according to which something is desirable if and only if it is what an agent would desire if she had all the relevant information and her desires were made coherent. If I knew that there was petrol in the glass, I would not want to drink it. This is why drinking the liquid from the glass in front of me is not desirable now. Likewise, if my desires were made coherent, my desire to drink gin might disappear because of my more fundamental desires. We can then generate counterexamples to the previous subjunctive account of desirability using the idea behind the conditional fallacy introduced above (Smith, 1995, p. 111). Imagine that you have just lost a game of squash. Is it desirable that you attempt to shake hands with your opponent? Let us assume that, when you lose, you become so angry that, if you do attempt to shake hands with your opponent, you will hit her. Yet, if you were rational, you would not do so. You would know how much it hurts to be hit and the process of making your desires coherent would excise your desire to hit the opponent. As a result, you would remain cool, calm, and collected. According to the view of desirability based on the subjunctive conditional, it is desirable for you to shake hands with your opponent in the actual situation. However, given that you will hit your opponent in that case, it is hard to see what would be desirable about the action you are contemplating. This again is an instance of the conditional fallacy. Making the antecedent of the subjunctive conditional true when we assess the whole conditional changes the circumstances that we are supposed to be assessing. As a result, we cannot bring our conclusions about the desirability of the given action in the counterfactual case to the real circumstances that we are assessing. 18 Similar theories about reasons have been defended by Darwall (1984, pp. 80–82), Falk (1990, pp. 36–39), Korsgaard (1986, p. 15), and Williams (1981). 19 These examples are from Williams (1981). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 157 Michael Smith has formulated a theory of desirability which can avoid the previous conditional fallacy (Smith, 1995, sec. 1; see also Railton, 1986, p. 174).20 Smith makes a distinction between the evaluated state of affairs of the actual agent, and the evaluating world in which the agent’s hypothetical fully-rational self is located. We then determine whether some course of action is desirable in the actual circumstances by considering whether the rational version of the agent (living in the evaluating world) would want the actual agent to act in that way in the evaluated circumstances. So, in the previous case, A considers in the actual world whether to attempt to shake hands with her opponent. She has a counterpart, A+, in the evaluating world with an informed and coherent version of her desires. Instead of asking what A+ would want herself to do, we ask what would A+ want A to do in the real world. Presumably, A+ would not want A to attempt to shake hands with her opponent. She knows that A would only hit her opponent. This means that Smith’s advice-model gives the right conclusion. It is not desirable for A to attempt to shake hands with her opponent because, when thinking about this from the evaluating world, the idealised version of A would not want A to do so in the evaluated circumstances. It is clear why this theory avoids the conditional fallacy. That we have two separate perspectives – the evaluated and the evaluating world – guarantees that making the antecedent of the subjunctive conditional true in the evaluating world cannot affect the evaluated real circumstances. Because of this, Smith’s view is immune to the conditional fallacy. The changes we make to the evaluating world cannot affect the circumstances of the real world, the world that is being evaluated. This is why the truth-value of our assessments cannot be distorted by these changes. 5. The ‘Advice Version’ of Whole Life Satisfaction theories of happiness The same move can be made to save HYWLS from Feldman’s counterexamples.21 The resulting ‘advice version’ of the Hypothetical Whole Life Satisfaction (ADHYWLS) view could claim that a person, S, is happyg to degree n at time t if and only if: 1) There is a certain life that S has lived up to t in the evaluated circumstances, and 2) if, at t, an idealised version of S, S+, were to form an ideal life-plan for S’s life, and 3) if, at t, S+ were to form a moderately detailed conception of how S’s life in the evaluated circumstances has transpired so far, then, 4) at t, S+ would judge that S’s actual life matches S+’s ideal life-plan for S to degree n. The basic idea of this proposal is the same as the one behind Smith’s theory of desirability. Consider the following example. We are trying to assess how happy Ann is. She lives in the evaluated (actual) circumstances. We can then imagine an idealised version of Ann, Ann+, looking at Ann’s life from the hypothetical evaluating circumstances. Ann+ shares Ann’s likes and dislikes. She also knows what Ann wants and values during different stages of her life. On the basis of this information, Ann+ can formulate an ideal life-plan for Ann. If Ann’s actual life went according to this plan, she would be engaged in activities that she likes and her desires would be satisfied. 20 Robert Johnson (1999) has argued that Smith’s proposal cannot both avoid the conditional fallacy and be an internalist view of desirability in the desired sense. That is, the resulting view will no longer have room for a direct connection between what is desirable and what we are motivated to pursue. However, Johnson does not contest that Smith’s view does manage to avoid the conditional fallacy. 21 Feldman (2008, p. 236; 2010, p. 88) comes close to formulating something like this proposal on the basis of Valerie Tiberius’s unpublished work (see Tiberius (2003)). I will address the objections, which he might be presenting to this view, below (Section 6.5). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 158 Ann+ will also be informed about what has happened in Ann’s life so far. Ann+ will therefore be able to make an informed judgment about the degree to which Ann’s life matches up with her life-plan for Ann. ADHYWLS then claims that Ann is happy in the actual world at a given time if and only if Ann+ would judge that Ann’s life at that point meets the standards set by the ideal life-plan which Ann+ has for Ann. This proposal avoids Feldman’s objections to both the actualist CWLS and the hypotheticalist HYWLS. ADHYWLS does not require that the agent whose happiness is assessed actually form any life-plans or beliefs about how her life matches those plans. Rather, ADHYWLS is based only on a purely hypothetical judgment which an imaginary version of the given agent could make. So, the objections to the actualist CWLS do not apply to this view. The objections to the hypotheticalist HYWLS do not apply to this view either. Consider first unreflective Timmy. According to ADHYWLS, in assessing his happiness, we do not need to consider what Timmy himself would make of his life. Rather, we first imagine a more rational version of Timmy who knows Timmy’s likes and dislikes (and his desires, values, and the like). Because of this knowledge, the idealised Timmy (Timmy+) can come to believe that the best life- plan for Timmy in the real world just is to live spontaneously. Any self-reflection from his actual counterpart would merely be counterproductive. For this reason, the life-plan which Timmy+ would have for Timmy would not contain any such reflection. According to the ideal life-plan then, Timmy should not think about his life. Rather, to follow this plan, Timmy should simply continue to take part in the activities he enjoys. Timmy+ can therefore come to judge that Timmy’s actual life meets the standards of the ideal life-plan he has for Timmy. As a result, this view can provide the right intuitive conclusion according to which Timmy is happyg in real life. The same can be said about the contemplative philosopher. We can imagine a hypothetical, idealised version of her. This philosopher+ knows that the actual philosopher likes being immersed in philosophising and hates thinking about her own life. As a result, the philosopher+ can formulate an ideal life-plan for the philosopher that includes much philosophising and little self-reflection. The philosopher+ can then judge that the actual philosopher’s life really does satisfy the ideal life-plan that she has for the actual version of herself. This means that ADHYWLS gives the right intuitive conclusion about the philosopher’s happiness. Thus, unlike the previous Whole Life Satisfaction views, ADHYWLS is immune to Feldman’s counterexamples. Those counterexamples were based on the conditional fallacy, which many accounts based on subjunctive conditionals commit. The problem with these views is that making the antecedent of the relevant subjunctive conditionals true interferes with the circumstances we are trying to assess. Yet, if we make Smith’s distinction between the evaluated and the evaluating perspectives, this interference cannot happen. The antecedent of the counterfactual is made true only in the hypothetical (evaluating) circumstances from which the unchanged actual circumstances are then evaluated. This is why assessing the subjunctive conditionals of ADHYWLS does not distort the truth-value of the assessed happiness claims. At this point, it is worth noting that AWLS and HWLS – the affective and hybrid versions of the Whole Life Satisfaction theories – are also vulnerable to Feldman’s counterexamples. According to AWLS, Timmy cannot be happy because he cannot be feeling satisfaction on the basis of having a sense of how well his life is going. As an unreflective person, Timmy simply has no conception of this. Likewise, Timmy cannot be happy according to HWLS, because that would require that he had formulated a life-plan, come to a conception of his life, made a judgment of An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 159 how the two match up, and felt satisfaction as a result. He has, of course, done none of these things. Furthermore, the hypotheticalist versions of AWLS and HWLS would also commit the conditional fallacy. In the counterfactual situations which would be specified by those views, Timmy would not feel satisfaction about his new life in which he is made to self-reflect. As a result, even the actual Timmy would not count as happy. Finally, it’s not clear whether the defenders of these views could avoid these problems by adopting the advice-model introduced above. Both of these views are motivated by the thought that one constituent of happinessg must be certain positive attitudes: feelings of satisfaction and the like. If we tried to use the advice-model to revise AWLS or HWLS, the result would be that one requirement for being happyg is that an idealised version of a person experienced satisfaction from comparing her actual life to her ideal life-plan. However, it would be strange if such a hypothetical feeling of satisfaction were required for actual happiness. So, it does not make sense to use the advice-model to develop AWLS or HWLS in order to avoid Feldman’s counter-examples. Rather, I will try to explain in the next section how ADHYWLS itself can be used to explain why feeling satisfied is often an important part of happinessg (Section 6.2). 6. Some advantages of ADHYWLS and responses to objections In this concluding section, I will first sketch some advantages of ADHYWLS as a theory of happinessg. I acknowledge that these advantages do not amount to a full argument for this view.22 I will then conclude by responding to some potential objections to ADHYWLS. 6.1 Flexibility The first advantage of ADHYWLS is that it offers a flexible structure for many different theories of happinessg. The first form of flexibility is temporal. ADHYWLS is an account of what is required for being happy during a certain period of time. Let us assume that this period, t, is shorter than the agent’s whole life but longer than a mere fleeting moment. In this situation, one version of ADHYWLS would claim that A is happyg during t if and only if A+ would judge that A’s life during t matches up with the corresponding period of A+’s ideal life-plan for A. Other versions might claim that, for A to be happyg during t, it is required that A’s life matches up with A+’s life-plan for her not only during t but also during some period before t (and maybe even for some time after t too).23 A second source of flexibility comes from the fact that ADHYWLS can accommodate different kinds of idealisation of agents, and different ways in which they could construct the relevant life-plans. According to the simplest option, A+ would merely be aware of the complete sets of A’s wants, likes, desires, intentions, goals, and values (hereafter ‘conative attitudes’) during different periods of her life.24 A+ would then devise a plan that would maximise the number of these attitudes satisfied throughout A’s life. 22 So, I make no claims that this proposal is, for instance, definitely superior to Haybron’s emotional state conception of happiness, defended in Haybron (2008). 23 If t is agent’s whole life, then presumably A+ must compare A’s whole life to the ideal life-plan. Feldman (2010, pp. 91–97) explores the advantages and difficulties of the different temporal formulations of the Whole Life Satisfaction views. Haybron (2001, p. 510) emphasises the forward looking aspect of happiness. 24 Kraut (1979, pp. 178–179) and Haybron (2007, sec. 3.1) emphasise that it is important that the agent’s moral and evaluative beliefs are also included in the input for the ideal life-plans. One advantage of the advice-model is that the idealised version of the agent can, by balancing them, find a coherent combination of the ethical perspectives that the An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 160 Alternatively, A+ could devise a life-plan that would lead only to the satisfaction of A’s strongest, most deeply held and integrated conative attitudes. These two views would disagree about whether satisfying merely trivial desires could make an agent happyg. It is also possible to further idealise the ideal life-plans devised by the advisors. For instance, following Smith, one could stipulate A+ to have an informed and coherent version of A’s set of conative attitudes. We could then consider what kind of a life-plan A+ would form for A on the basis of that idealised set. The resulting theory would claim that a happyg life needs to be a desirable life in Smith’s sense. In this case, an agent could not become happierg by spending nights at the pub, if we assume that her ideal self would not plan such nights for her. The idealised self would know how bad the agent’s hangovers are, and how her desire to go to the pub conflicts with her more fundamental desires to work on the following days.25 We could even stipulate that the advisor is informed about all the objective evaluative facts regarding what constitutes a good life. She would then form the ideal-plan for the agent, not only on the basis of the agent’s conative attitudes, but also on the basis of which lives would be objectively good for the agent. According to this version of ADHYWLS, an agent could not live a happyg life unless her life had a high level of objective well-being. It would not be possible to live a happy life, according to this view, that consisted of mainly counting blades of grass.26 Finally, some views of happinessg claim that happy lives must have a certain narrative ‘shape’ (Annas, 2004, pp. 47–48).27 According to these views, all happy lives must develop in overlapping and interrelated stages. Such lives must include overarching projects and obstacles to achieving them, development of skills, helpers and enemies, instances of overcoming difficulties and the like. The corresponding version of ADHYWLS would insist that the life- plans formulated by the agents’ ideal selves also need to have a certain story-like structure. Alternative accounts accept that people who live episodic lives with no narrative structure can be just as happyg as those who lead narratively-connected lives (Feldman 2004, ch. 6; Feldman, 2010, pp. 175–177; Strawson, 2008, sec. 10). According to their version of ADHYWLS, the ideal life-plans for some people could consist of various unconnected episodes lived in completely different ways. These views would allow people who do not experience their lives as narratives to be just as happy as anyone else. There thus seem to be many possible versions of ADHYWLS. This is an advantage. The next stage in developing ADHYWLS would be to compare which of the possible versions of the view mentioned above best fits our intuitions about who is happy.28 In this way, we could reach actual agent can take towards her life. This solves many of the problems related to the arbitrariness of actual Life Satisfaction assessments (see Haybron, 2008, ch. 5, sec. 5). 25 This would admittedly make happinessg more objective, and less dependent on the agent’s own attitudes. 26 This would weaken the Personal Authority that individuals are often thought to have over the question of their happiness, i.e., this would make happiness less transparent for individuals (Haybron, 2008, p. 13; but see Feldman (2010, ch. 11) for an objection). Someone might also add that a happyg life must be a morally virtuous life. In this case, an ideal life-plan for anyone would necessarily plot a virtuous life. On this view, evil people could not be happy (Foot, 2001, p. 92). Such views would be influenced by Aristotle’s discussions of ‘eudaimonia’ (Kraut, 1979, pp. 174–176, and sec. 4). Kraut (1979) argues convincingly that the ideal-life plans relevant for happiness should not include objective facts about the virtuous life. 27 Antti Kauppinen (forthcoming) argues that a narrative shape is a requirement for having a ‘meaningful’ life. 28 There is also another sense in which the view is theoretically neutral. ADHYWLS is formulated as a bi-conditional according to which the property of being happyg and the property of living a life that matches up to one’s ideal life- plan are necessarily co-instantiated. This bi-conditional itself is neutral about whether the theory should be understood as a conceptual analysis of ‘happiness’, as a theory of the constitution of the property of happiness g, as a theory of what makes one happy, as a theory of what is merely a reliable indicator of happiness, and so on. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 161 a reflective equilibrium between our considered judgments about the happiness of different agents and our philosophical theory of happinessg. 6.2 Happinessg, happinessp, and happinessw Could ADHYWLS provide a plausible account of the relations between the different forms of happiness? Above, I claimed that feeling happy (i.e., happinessp) does not always seem to be necessary or sufficient for actually being happyg. Yet, in considering the motivation for AWLS and HWLS above, I admitted that feeling happy does often seem to make people happierg. Can ADHYWLS accommodate both of these claims? First, it is easy to see why feeling happy is not always necessary for being happy g according to ADHYWLS. An ascetic person may not desire, like, or value feeling happy, euphoric, satisfied, and so on. If an ideal version of this person formulated a life-plan for her, this plan would not include many hedonic experiences. As a result, such a person would not be happy g on the basis of feeling happy (Haybron, 2008, p. 84). Of course, most ordinary people do desire, like, and value hedonic experiences. So, the ideal life-plans that their advisors would create for them would include having these experiences often. Therefore, these individuals would be happierg the more they experienced feelings of happiness, pleasure, satisfaction, and the like. This is why, according to ADHYWLS, feeling happy can generally make most people happyg, at least to some degree. It is worth emphasising that people can also desire experiences of happiness with a specific causal origin. I may desire the feeling of happiness that I get from smelling flowers but not that which is induced by drugs. This is why, according to ADHYWLS, I would not be made happierg by taking drugs that made me feel euphoric. I also claimed earlier that it is possible to be happyg without living a life with a high level of wellbeing (see also Haybron, 2003, sec. 4.3). This will be true only if we assume an objectivist view of wellbeing.29 This view claims that there are universal constituents of wellbeing that can be listed.30 They might include, for instance, wealth, knowledge, health, pleasure, autonomy, friendship, parenting, food, shelter, meaningful work, virtue, perfection of human nature, and so on. In this situation, some individuals might care little about the items on the previous list. They might, for instance, prefer making art, collecting stamps, or philosophising to having a sufficient amount of all the listed items. As a result, the ideal life-plans that their advisors would create for them might not include many of the universal constituents of wellbeing. These individuals could be happyg even if their lives were lacking in wellbeing (Haybron, 2008, p. 81). Of course, the vast majority probably would prefer to have most of the constituents of objective wellbeing present in their lives (Foot, 2001, p. 88). These individuals would be happy g according to ADHYWLS only if their lives had a high level of objective wellbeing.31 This is why the wellbeing of these individuals is important for their happinessg, and why their happinessg is important for their wellbeing (Haybron, 2008, p. 53). 29 This will also depend on the version of ADHYWLS in question. As noted above, according to some versions of the view, ideal life-plans take into account objective evaluative facts about good lives. This would make it trivially true that a happy life must contain a high level of wellbeing. 30 Parfit (1984, pp. 493–502) famously distinguishes such theories from hedonistic and desire-satisfaction views of wellbeing. 31 ADHYWLS would thus explain why happinessg has prudential value for most people, i.e., why happiness in this sense is important for most of us (Haybron, 2003, sec. 4.3). An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 162 Finally, some views of wellbeing claim that an agent’s wellbeing consists of the satisfaction of her desires. According to these views, living a life with a high level of wellbeing might not require more than happinessg as described by ADHYWLS.32 6.3 The reason-giving and explanatory roles of happinessg Daniel Haybron has recently argued that our everyday concept of happiness serves several functions in ordinary discourse (Haybron, 2003, sec. 3.5; Haybron, 2008, ch. 3, sec. 4). Because of this, any plausible theory of happiness should be able to explain why the concept of happiness can serve these practical roles. Can ADHYWLS do so? Well, why do we talk about happiness? First, we often appeal to considerations about happiness in making important decisions. We consider the fact that a certain career will make us happy to be a reason to pursue that career. Second, we often utilise happiness in assessing how well someone’s life is going. In doing so, we assume that highly successful people are happy (other things being equal), and that people who are unhappy have failed to achieve their central goals. Third, we use happiness to predict the behaviour of others. For example, if someone is unhappy, we know that they are unlikely to be good company at a dinner party. Finally, we refer to happiness in explaining the actions of others. For instance, a person’s unhappiness can be used to explain why she has made major changes in her life. Could happinessg be used for these purposes in ordinary discourse if ADHYWLS were the correct account of this form of happiness? First, if it were, we would have in most cases reasons to pursue our happinessg. After all, our happiness would then be a matter of our life matching a life-plan that a more knowledgeable version of ourselves had created for us on the basis of our desires, likes, goals, values, and the like. Being happy would therefore imply that our most important ends were achieved. Hence, if we have reasons to achieve our goals, then, according to ADHYWLS, we also have reasons to pursue happinessg. Equally clearly, happinessg as understood by ADHYWLS could be used to measure how well someone is doing in their lives. Happiness would then simply consist of an agent’s actual life measuring up to a life-plan that is ideal for her. This understanding of happinessg could also be used for prediction and explanation. We can assume that other people share many of our ends such as enjoyment, human interaction, meaningful work, and so on. If we know that they are unhappyg, we also know that their lives do not match up to their ideal life-plans (which involve the satisfaction of these shared goals). This tells us many useful things about the lives and psychological states of these agents. This further information could be used to explain, for instance, why unhappy agents seek to change their lives. We could also use it to predict the behaviour of such agents. 6.4 Happinessg and the Experience Machine Laura Sizer has argued that Life Satisfaction theories lead to implausible consequences in Nozick’s Experience Machine case (Sizer, 2010, p. 140; Nozick, 1974, pp. 42–45). Imagine a virtual reality machine that would give you experiences such that you could not distinguish them from real experiences. Because of these experiences, it would seem to you that your life went according to your ideal life-plan. It would be as if things went well with your friends, that 32 Haybron (2007) argues that there need not even be a match between Whole Life Satisfaction and subjective notions of wellbeing. This is because Whole Life Satisfaction views take into account satisfaction of a wider range of conative attitudes than the subjectivist views of wellbeing. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 163 you had a successful career, and so on. Yet, none of this would actually be happening. You would furthermore forget that you were in the machine. Sizer claims that, because Life Satisfaction views measure happiness in terms of whether an agent judges her life to measure up to her life-plan, these views would entail that agents would be just as happyg in the Experience Machine. After all, the agent would make the very same judgments both in the Experience Machine and in real life. However, many people think that being happyg cannot be merely a matter of an agent believing that her life matches up to her life-plan. Rather, according to them, the agent’s life must actually match up to her life-plan as well. Fortunately ADHYWLS can also avoid this problem. The agent’s idealised version, A+, would know whether or not the actual agent, A, is in the Experience Machine. This would be a relevant factor in A+’s assessment of whether A’s life matches up to A+’s ideal life-plan for her. If the plan which A+ forms for A on the basis of A’s goals includes real achievements and relationships, then A+ would not think that A’s life in the Experience Machine matches up to her life-plan. This entails that, according to ADHYWLS, A would not be happyg in the Experience Machine. Yet, someone else might value the relevant experiences themselves more than anything more genuine. Such a person’s life-plan might be satisfied in the Experience Machine. ADHYWLS would classify this kind of agent as happyg. However, as explained above, it wouldn’t classify everyone in the Experience Machine as happyg. So, ADHYWLS is not vulnerable to Sizer’s objection. 6.5 Feldman’s final objections In this last sub-section, I want to address three remaining objections that Feldman presents to all forms of Whole Life Satisfaction theories. The first problem, the so-called instability (or ‘lability’) problem, is that people’s actual judgments about how their lives match up to their life-plans change more frequently than their state of happiness could ever change (Feldman, 2008, sec. 3; Feldman 2010, sec. 5.3; see also Haybron, 2007, sec. 3.3; Haybron, 2008, pp. 86–91). For instance, it has been demonstrated that trivial changes to the environment in which an agent makes life-satisfaction judgments affect those judgments (Nettle, 2005, pp. 34–36). If a subject finds a ‘lost’ dollar in the experimental setting, she is likely to indicate greater satisfaction with her life than she would have done otherwise. Because of this, our actual judgments about life-satisfaction cannot be reliable indicators of happinessg. ADHYWLS does not suffer from such problems. It relies on hypothetical judgments of an idealised version of the agent. Because these hypothetical judgments are based on being accurately informed about the agent’s real life and her stable conative attitudes, these judgments would not change from one moment to another. In fact, we do not need to think that they are made from any real temporal perspective at all. Likewise, these judgments could not be affected by trivial changes in the agent’s actual circumstances. So, ADHYWLS is able to avoid these problems with how fickle people’s actual judgments are. At the end of his article, Feldman presents two further counter-examples to Whole Life Satisfaction theories (Feldman, 2008, pp. 236–237; Feldman, 2010, p. 87, fn. 36, and pp. 88–89). The first of these concerns an agent who is badly confused about the actual circumstances of her life. For instance, we can consider Jim Carrey’s character Truman from the film The Truman Show. Unbeknownst to him, Truman is in fact living in the centre of a constructed reality An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 164 television show. Thus, although he does not know it, Truman’s relationships and projects are not real because all of the other participants in them are merely acting. Despite this, many people would probably judge that Truman is living a happy life within the show. However, were Truman to make an informed judgment about his life, he might come to judge it to be less than satisfactory. So, Feldman claims that even the hypothetical versions of the Whole Life Satisfaction theory could not accommodate the happiness of uninformed people like Truman. ADHYWLS can deal with this case in the same way as it deals with the previous cases. In some cases, the idealised version of the agent can come to judge that most of the agent’s fundamental goals (including the one of having many pleasurable experiences) will be achieved only if the agent does not have an accurate idea of her circumstances.33 As a result, the ideal life-plan for this person would not require her to be informed. This would mean that this person’s life of ignorance would match up to her ideal life-plan, and thus her ignorance would be one constituent of her happinessg. The same can be said about Feldman’s second case. In it, Tammy is miserable because she has an inaccurate picture of how well her life is actually going. If Tammy were to become more informed about her life, she might judge that her life actually matches up quite well with her life-plan. Feldman claims that intuitively Tammy is unhappy even if the hypothetical versions of the Whole Life Satisfaction views would lead to the opposite conclusion.34 This case might show that our intuitions about happiness are in part confused because of the different senses of ‘happiness’ (see Footnote 14, and the end of Section 3 above). Perhaps we do not want to say that Tammy is happy because we are thinking of phenomenological happinessp here instead of the more global happinessg. In the latter sense of the term, it might not be so implausible to say that Tammy is at least not completely unhappy because, unbeknownst to her, ‘she is not a sinner’, ‘she is well liked by others’, and her life ‘matches up with her important ideals’ (Feldman, 2008, p. 237). Of course, it might also be fair to say that she would probably be much happierg if she knew all of this because she would then be able to enjoy these achievements. However, ADHYWLS can accommodate this thought too. Tammy’s goal might be to live an informed life and to experience enjoyment and satisfaction. If this were true, then Tammy+ would not judge that Tammy’s life completely matches up to the ideal life-plan that she has for her. This would mean that ADHYWLS would come to the same conclusion as Feldman about this case: that Tammy is not happy. As a result of this, I conclude that ADHYWLS is able to deal with all of Feldman’s objections. It is therefore an improved version of the most popular view of happiness and so deserves further investigation. 33 These cases might be rare. It could be claimed that there is an even more ideal life-plan for agents like Truman that includes becoming more informed, starting a new life on the basis of this new information, and thus becoming ‘genuinely happy’. This objection relies on the intuition that Truman is not initially genuinely happy (contrary to what Feldman’s example states). This thought would not be a problem for my view. We would simply need to add one requirement for ideal life-plans: that they must be informed. However, if we do believe that Truman can be genuinely happy, then the above view just explains how my view could accommodate that intuition too. Of course, we could also say that, even if Truman is happy, he is not living a desirable life. We could then use the informed life- plans to account for desirability, and the non-informed plans to account for happiness. 34 Here is one case which AWLS (and possibly even HWLS) seems to give the right results. That view would claim that Tammy is unhappy because of the displeasure she experiences as a result of making her (false) judgment about how her life is going. An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 165 Author Jussi Suikkanen University of Birmingham jussiphil@gmail.com References Annas, J. (2004). Happiness as achievement. Daedalys, 133, 44–51. http://dx.doi.org/10.1162/001152604323049389 Benditt, T. (1974). Happiness. Philosophical Studies, 25, 1–20. http://dx.doi.org/10.1007/BF00364711 Brandt, R. B. (1967). Happiness. In P. Edwards (Ed.), The Encyclopedia of Philosophy (pp. 413–414). New York: The MacMillan Company and Free Press. Brandt, R. B. (1979). A theory of the good and the right. Oxford: Oxford University Press. Crisp, R. (2004). Reasons and the good. Oxford: Oxford University Press. Darwall, S. (1984). Impartial reason. Ithaca, NY: Cornell University Press. Falk, F. D. (1990). Ought and motivation. In his Ought, reasons, and motivation (pp. 21–41). Ithaca, NY: Cornell University Press. Feldman, F. (2002). The good life: A defence of attitudinal hedonism. Philosophy and Phenomenological Research, 65, 604–628. http://dx.doi.org/10.1111/j.1933-1592.2002.tb00223.x Feldman, F. (2006). Pleasure and the good life: Concerning the nature, varieties, and Plausibility of Hedonism. Oxford: Clarendon Press. Feldman, F. (2008). Whole life satisfaction concepts of happiness. Theoria, 74, 219–238. http://dx.doi.org/10.1111/j.1755-2567.2008.00019.x Feldman, F. (2010). What is this thing called happiness? Oxford: Oxford University Press. http://dx.doi.org/10.1093/acprof:oso/9780199571178.001.0001 Foot, P. (2001). Natural goodness. Oxford: Oxford University Press. http://dx.doi.org/10.1093/0198235089.001.0001 Haybron, D. (2000). Two philosophical problems in the study of happiness. The Journal of Happiness Research, 1, 207–225. http://dx.doi.org/10.1023/A:1010075527517 Haybron, D. (2001). Happiness and pleasure. Philosophy and Phenomenological Research, 62, 501–528. http://dx.doi.org/10.1111/j.1933-1592.2001.tb00072.x Haybron, D. (2003). What do we want from a theory of happiness? Metaphilosophy, 34, 305–329. http://dx.doi.org/10.1111/1467-9973.00275 Haybron, D. (2007). Life satisfaction, ethical reflection, and the science of happiness. Journal of Happiness Studies, 8, 99–138. http://dx.doi.org/10.1007/s10902-006-9006-5 Haybron, D. (2008). The pursuit of unhappiness. Oxford: Oxford University Press. Johnson, R. (1999). Internal reasons and the conditional fallacy. Philosophical Quarterly, 49, 53–71. http://dx.doi.org/10.1111/1467-9213.00128 Kauppinen, A. (forthcoming). Meaningfulness and time. Philosophy and Phenomenological Research. Kekes, J. (1982). Happiness. Mind, 91, 358–376. Korsgaard, C. (1986). Scepticism about practical reason. Journal of Philosophy, 83, 738–753. http://dx.doi.org/10.2307/2026464 Kraut, R. (1979). Two Conceptions of happiness. The Philosophical Review, 88, 167–197. http://dx.doi.org/10.2307/2184505 Montague, R. (1966). Happiness. Proceedings of the Aristotelian Society, 67, 87–102. Murphy, J. (2004). The unhappy immoralist. The Journal of Social Philosophy, 35, 11–13. http://dx.doi.org/10.1111/j.1467-9833.2004.t01-1-00207.x Nettle, D. (2005). Happiness: The science behind your smile. Oxford: Oxford University Press. Nozick, R. (1974). Anarchy, state, and utopia. Cambridge, Ma.: Harvard University Press. Nozick, R. (1989). The examined life. New York: Simon and Schuster. Parfit, D. (1984). Reasons and persons. Oxford: Oxford University Press. Railton, P. (1986). Moral realism. Philosophical Review, 95, 163–207. http://dx.doi.org/10.2307/2185589 http://dx.doi.org/10.1162/001152604323049389 http://dx.doi.org/10.1007/BF00364711 http://dx.doi.org/10.1111/j.1933-1592.2002.tb00223.x http://dx.doi.org/10.1111/j.1755-2567.2008.00019.x http://dx.doi.org/10.1093/acprof:oso/9780199571178.001.0001 http://dx.doi.org/10.1093/0198235089.001.0001 http://dx.doi.org/10.1023/A:1010075527517 http://dx.doi.org/10.1111/j.1933-1592.2001.tb00072.x http://dx.doi.org/10.1111/1467-9973.00275 http://dx.doi.org/10.1007/s10902-006-9006-5 http://dx.doi.org/10.1111/1467-9213.00128 http://dx.doi.org/10.2307/2026464 http://dx.doi.org/10.2307/2184505 http://dx.doi.org/10.1111/j.1467-9833.2004.t01-1-00207.x http://dx.doi.org/10.2307/2185589 An improved whole life satisfaction theory of happiness Suikkanen www.internationaljournalofwellbeing.org 166 Rawls, J. (1971). A theory of justice. Cambridge, Ma.: Harvard University Press. Shope, R. (1978). The conditional fallacy in contemporary philosophy. The Journal of Philosophy, 75, 397– 413. http://dx.doi.org/10.2307/2025564 Smith, M. (1995). Internal reasons. Philosophy and Phenomenological Research, 55, 109–131. http://dx.doi.org/10.2307/2108311 Strawson, G. (2008). Against narrativity. In his Real materialism (pp. 189–207). Oxford: Oxford University Press. Sumner, W. (1996). Welfare, happiness, and ethics. Oxford: Clarendon Press. Tatarkiewicz, W. (1966). Happiness and time. Philosophy and Phenomenological Research, 27, 1–10. http://dx.doi.org/10.2307/2106134 Tatarkiewicz, W. (1976). Analysis of happiness. Trans. by E. Rothert and D. Zielinsk. The Hague: M. Nijhoff. Taylor, R. (2002). Virtue ethics: An introduction. Amherst, NY: Prometheus Books. Telfer, E. (1980). Happiness. New York: St Martin’s Press. Thomas, D. A. L. (1968). Happiness. Philosophical Quarterly, 18, 97–113. http://dx.doi.org/10.2307/2217508 Tiberius, V. (2003). ‚How’s it going?‛: Judgments of overall life-satisfaction and philosophical theories of well-being. Presented at Minnesota Interdisciplinary Workshop on Well-Being. http://www.tc.umn.edu/~tiberius/workshop_papers/Tiberius.pdf Veenhoven, R. (1984). Conditions of happiness. Dordrecht: D. Reidel. von Wright, G. (1963). Varieties of goodness. London: Routledge and Kegan Paul. Williams, B. (1980). Internal reasons. In his Moral luck (pp. 101–113). Cambridge: Cambridge University Press. Williamson, T. (2007). The philosophy of philosophy. Oxford: Blackwell. http://dx.doi.org/10.1002/9780470696675 Wittgenstein, L. (1958). Philosophical investigations. Oxford: Basil Blackwell. Wright, C. (1992). Truth and objectivity. Cambridge, Ma.: Harvard University Press. http://dx.doi.org/10.2307/2025564 http://dx.doi.org/10.2307/2108311 http://dx.doi.org/10.2307/2106134 http://dx.doi.org/10.2307/2217508 http://www.tc.umn.edu/~tiberius/workshop_papers/Tiberius.pdf http://dx.doi.org/10.1002/9780470696675