Hard Choices and Ultimate Ends

Abstract

Graduate Paper from the 2023 Joint Session

I

Hard Choices and Their Challenge to Practical Rationality. ‘Classic’ hard choices, transformative experiences, and aspirational choices have been of interest to philosophers of practical rationality (Chang 2013, 2020, 2021; Paul 2014; Callard 2018; Pettigrew 2019).

I propose a unified framework to analyse these kinds of choices. On my novel view, ‘classic’ hard choices, transformative experiences, and aspirational choices, usually discussed separately, are all characterized as hard choices. I call a choice ‘hard’ when (i) the individual has no all-things-considered reason to choose one option over another, and (ii) the objects of choice are ultimate ends.

Construing hard choices in this way illuminates how we should appraise the practical rationality of hard choices. For one thing, it helps explain the intuition underlying the common assumption that, when faced with hard choices, it is impermissible to choose arbitrarily. I show that this is because the distinctive features of hard choices I identify make plumping, but not picking, impermissible. For another, it sheds light on moral and agential considerations that constrain permissible choice procedures when faced with hard choices.

II

Hard Choices. I shall first present ‘classic’ hard choices, transformative experiences, and aspirational choices, and show their relevant similarities.

• (1) ‘Classic’ Hard Choices. In some choices, no option is evaluated as better or worse than, or equal to, the others by the choosing individual. Philosophers disagree on what value relation between options (or property thereof), if any, holds in these cases.¹ Regardless, in ‘classic’ hard choices an individual does not have an all-things-considered reason to choose one option over all others.

• (2) Transformative Experiences. Transformative experiences are ‘life-changing experiences that shape who we are and what we care about’ or affect our understanding of the world (Pettigrew 2019; Paul 2020, p. 16). An experience is transformative when its subjective evaluation will change after having embarked on it and, as a result, an individual’s ex ante and ex post preferences differ. Since there is no principled way to adjudicate on the basis of what preferences (ex ante or ex post) one should choose, the choosing individual has no all-things-considered reason to embark on the experience or not.

• (3) Aspirational Choices. Aspirational choices are ‘purposive value acquisition’ (Callard 2018, p. 8). They are choices through which one comes to value something that one does not yet value at the moment of choice (say, becoming a musician, a wine connoisseur, and so on). Take the aspiring wine connoisseur who finds the taste of wine repulsive but aspires to value wine-drinking for its own sake. The values of her current self, who does not appreciate wine, and the values that she wants her future self to acquire pull in opposite directions. Since she has no way of adjudicating between these two opposing values, she has no all-things-considered reason to transform herself or not.

Choices of types (1), (2) and (3) share two features. First, the individual has no all-things-considered reason to choose one option over another. Second, it is impermissible to choose arbitrarily between options. Or so many have argued (for example, Reuter and Messerli 2017; Andersson and Herlitz 2022).² From now on, I will refer to choices that satisfy these two characteristics as ‘hard choices’. I take choices (1), (2) and (3) to be hard choices in this sense.

There are at least two ways of choosing arbitrarily: picking or plumping. Picking amounts to choosing arbitrarily when a person has sufficient reason to choose either of the options, but not one option over all others (Chang 2015, p. 217). Plumping equates to choosing arbitrarily by acting outside the scope of rationality, without responding to reasons, without giving thought to the choice or consideration to the values conveyed by the objects of choice, but rather by relying solely on impulses or passing desires.³ To appreciate the distinction between these two kinds of arbitrary choices, consider ‘Sophie’s choice’ (Styron 1979), where Sophie is forced to decide which of her children to send to death. Imagine that she decided to flip a coin, making an arbitrary choice. She could do so to express her rejection to be associated with the outcome of the coin toss as something that she has ‘chosen’ in a substantive sense. The choice of tossing a coin, in this case, would not amount to plumping. Rather, she would be picking, displaying that she has sufficient reason to choose that each of her children is saved, but not one over the other, and thus illustrating her refusal to favour one over the other. It seems to be a shared presumption that, since Sophie lacks an all-things-considered reason to choose one option rather than the other, it is impermissible for her both to plump and to pick. I will dispute this assumption below.

Why is there the presumption that it is impermissible to choose arbitrarily when faced with hard choices? After all, the lack of an all-things-considered reason in favour of an option does not warrant the rational impermissibility of choosing arbitrarily by picking or plumping: think of ‘Buridan’s ass’ (Weintraub 2012; Chislenko 2016). Buridan’s ass does not have an all-things-considered reason to choose one rather than the other of two identical haystacks, and starves to death. Yet there appear to be weighty practical considerations, such as the donkey’s survival, that provide the donkey with sufficient reason to pick (or plump for) either of the haystacks.

Some have suggested that the value relation between the options is the culprit. Chang (2017) charges parity.⁴ Others have pointed to the ‘importance’ (Reuter and Messerli 2017) or ‘weightiness’ (Ullmann-Margalit 2006) of the decision, or the fact that it is about one’s ‘rational identity’—‘a description of who [an individual] would be were [she] perfectly rational’ (Chang 2009, p. 262). These answers, however, need further argumentation to be satisfactory. On the one hand, ‘weightiness’ or ‘importance’ are vague concepts if not further defined. Choosing what to eat may be important for one’s health. But it nonetheless seems permissible to choose arbitrarily between meals if there are practical considerations, such as time constraints. On the other hand, it is not immediately clear what decisions are about one’s (rational) identity and why (rational) identity makes choosing arbitrarily impermissible. In the next section, I will provide further argumentation that clarifies why it is impermissible to choose arbitrarily when faced with hard choices. Doing so will show that it is only impermissible to choose arbitrarily by plumping, but not by picking.

III

Ultimate Ends: Picking versus Plumping. Sophie’s choice is a paradigmatic example of a type of hard choice. Other examples are choices between jobs, lifestyles, and ideologies. What distinguishes hard choices from, say, a choice between incomparable or never-experienced dinners is that the former are between what I call ‘ultimate ends’, while the latter are not. I will argue that it is permissible to choose arbitrarily by plumping when one faces incomparable dinner plans but not when one faces ultimate ends.

In the following, I will assume that one can either choose directly between ultimate ends or make choices that are significant for their pursuit. And so, using the term ‘choice’ quite liberally,⁵ I will refer to both cases by the shorthand expression, ‘choosing between ultimate ends’.

Ultimate ends encompass various things, from jobs to lifestyles, and can be direct objects of choice or promoted by choosing. Ultimate ends are pursued for their own sake. They are, in Frankfurt’s (1992) and Schmidtz’s (1994) words, ‘final’. They are also enduring—they extend over time and are accomplished by repeatedly promoting them over the time during which an individual upholds them.

Taken together, ultimate ends constitute what Korsgaard calls an individual’s practical identity—a description of ‘who [an individual is], under which [she values herself]’ (Korsgaard 1996, p. 101). Adopting a new ultimate end shapes a person’s practical identity, and disposing of it comes at the cost of one’s practical identity (for instance, one is no longer a minimalist, or a doctor). Since disposing of an ultimate end may change one’s practical identity, doing so is hard—it is costly in terms of time, effort, and self-perception (Nguyen 2020, pp. 34–5).

Lastly, by virtue of their role in an individual’s practical identity, ultimate ends have greater normative significance than non-ultimate ends. On the one hand, ultimate ends constrain what non-ultimate ends an individual can pursue—they veto (that is, rule out as impermissible) the adoption of non-ultimate ends that conflict with an individual’s ultimate ends. For example, if upholding a minimalist lifestyle counts as an ultimate end for a person, and impressing one’s neighbours is a non-ultimate end, then being a minimalist vetoes collecting lavish cars to impress one’s neighbours. On the other hand, ultimate ends are not subordinate to each other. That is, while non-ultimate ends stand in a subordinate relation to ultimate ends, pre-existing ultimate ends do not straightforwardly veto the adoption of a new ultimate end that conflicts with them.

I have characterized an ultimate end as a final, hard-to-dispose-of end, and as having greater normative significance than non-ultimate ends. I want to claim that conceiving of hard choices as among ultimate ends grounds the impermissibility of choosing arbitrarily by plumping but not by picking. Indeed, it is the role of ultimate ends in our practical identity that explains why many are uneasy with the idea of arbitrarily choosing between them. However, I will suggest, the unease is fitting when it concerns choosing arbitrarily by plumping but not by picking.

Since ultimate ends together constitute an individual’s practical identity, their pursuit enables one to express one’s normative standpoint—one’s deeply held values. As such, if all our ultimate ends turned out to be the result of an arbitrary choice made outside the scope of rationality, their normative outlook would be bleak. It would imply that "the trajectory of [our] life … is itself profoundly arbitrary … But this makes a travesty of the human condition" (Chang 2021, p. 105).⁶ My illustration of Sophie’s choice above shows that not all forms of arbitrary choice display a lack of care towards the choice itself or are outside the scope of rationality. One may greatly care about the outcome of one’s choice and yet think that, even if one does not have an all-things-considered reason in favour of one option over another, one has sufficient reason to choose either. In these cases, an individual may permissibly pick without thereby plumping. That is true of other hard choices too. If a choice between jobs is hard, it is permissible for one to choose arbitrarily by picking a job, recognizing that one has sufficient (but not conclusive) reason to choose either. One could do so by tossing a coin, to either display a refusal to choose one job over another or to signal that there is sufficient reason to choose either. And yet, I submit, it would be impermissible to plump. Imagine that one chose to be a doctor rather than an engineer after walking past a stand of ‘doctors without borders’ and forming a fleeting desire to join them because of the colours of their logo. Or think of one who, burdened by the choice, drowned his sorrows and, in a daze, chose to be a doctor. Choosing arbitrarily in these ways would amount to choosing without responding to reasons, but only out of a passing desire or an impulse. It would display a lack of care towards the outcome of the choice.⁷ If my analysis is correct, the well-grounded aversion towards plumping for an ultimate end does not justify the aversion towards picking, which is, instead, a legitimate choice process, given that it displays sufficient consideration towards the process and the outcome of the choice.

Why, then, is it permissible to plump between dinner plans but not between paths in one’s life? What to eat for dinner does not, in most cases, characterize an individual’s practical identity. A path in one’s life, just like one’s job or ideology, does.

IV

Agential and Moral Constraints on Hard Choices. Construing hard choices as choices in which it is impermissible to plump, but not to pick, because they are between ultimate ends sheds light on the constraints on an individual’s practical rationality when faced with hard choices.

One agential consideration constrains the adoption and disposal of ultimate ends. Since ultimate ends jointly constitute an individual’s practical identity, disposing of them impacts an individual’s agential unity by fracturing one’s self-perception. It is commonly held by many that to be an agent one needs to have some minimal degree of agential unity (Korsgaard 1989; Millgram 1997; Nguyen 2020, p. 4). That is, one needs to see oneself as motivated by ends that are not incompatible with each other, minimally stable over time, and as not prey to merely passing desires, but rather as the legislator of whether those desires should count as reasons (Korsgaard 1989, pp. 109–12; Millgram 1997, pp. 51–3; Garnett 2011). This consideration does not imply that having some ultimate ends that are incompatible or sometimes acting on passing desires is necessarily objectionable. It may be prudentially good to pursue new ultimate ends that disrupt one’s agential unity when doing so adds meaning to an individual’s life by removing her from the boredom that a ‘fully unified agent’ (that is, a person who has stable unchanging ultimate ends) may experience (see Millgram 2004). Still, from the perspective of agency, it seems impermissible to adopt a new conflicting ultimate end unless it will bring about desirable agential disunity. I do not, and neither does Millgram (2004), spell out what counts as ‘desirable disunity’. While it is beyond the scope of this article to do so, it is important for theories of practical rationality to address this concern.

Besides this agential consideration, there are two moral considerations that constrain the adoption or disposal of ultimate ends.

The first regards the legitimate grounds to choose an ultimate end. Recall that an ultimate end is pursued for its own sake. To be able to pursue an end for its own sake, an individual, or so I will argue, must choose it based on an appreciation of its intrinsic value. This does not rule out that the appreciation of the intrinsic value can be attenuated or that ultimate ends may be chosen for additional considerations. Ultimate ends, I argued above, partly structure what is normative for a person, in so far as they are reason-giving, endowing a person with pro tanto reasons to promote them. To exert this normative role legitimately, they must be chosen adequately, on the basis of an appreciation of their intrinsic value. It is thus morally required to choose an ultimate end based on at least an attenuated appreciation of its intrinsic value. Incidentally, note that this constraint implies that it is prohibited to choose an ultimate end only because it would create pro tanto reasons to pursue an otherwise morally impermissible end that one has. For instance, pursuing a career as a policeman is impermissible if it is chosen only because it allows one to use violence.

There is a second moral consideration, which constrains the disposal, rather than the adoption, of ultimate ends. Since ultimate ends jointly constitute an individual’s practical identity, they have more normative significance than non-ultimate ends. I suggested that ultimate ends may veto the adoption of non-ultimate ends that conflict with their promotion. Specifically, ultimate ends are more demanding than non-ultimate ends—they are ‘robustly demanding’ (see Pettit 2015). That is, under a wide range of circumstances, the weight of the pro tanto reasons generated by an ultimate end is greater than each individual reason generated by non-ultimate ends. The conditions under which an individual can dispose of her ultimate ends are more stringent than those for disposing of non-ultimate ends. For instance, one should—or at least could—abandon one’s end of winning at a leisure activity if one’s opponents are crying, but it would be unfitting to abandon one’s ultimate end of winning a Grand Slam tournament as a professional tennis player because one’s opponent is in tears.

Where does this discussion leave us? My novel characterization of hard choices as comprising ‘classic’ hard choices, transformative experiences, and aspirational choices makes explicit their shared features. These shared features, in turn, help explain why it is impermissible to plump and yet permissible to pick between them. They also delineate moral and agential constraints on the permissible decision-making procedures faced with these choices. These considerations offer a new lens to appraise the merits of theories of practical rationality specifically developed to address hard choices. In future work, these considerations can hopefully be used to constructively challenge prominent theories of practical rationality, such as those put forth by Ruth Chang, Laurie Paul, and Agnes Callard.⁸

References

Footnotes