Hope as a Source of Grit

1. Hopeful Perseverance

Psychologists have argued that success at long-term, difficult endeavours (such as getting back in shape, mastering a musical instrument or getting a PhD) crucially turns on the capacity to persevere, called “grit”. This capacity is supposed to explain why, holding circumstances like talent and social context fixed, some press on and succeed in the face of adversity while others give up, despite still seeing their initial commitment as valuable (see Duckworth, Peterson, Matthews, & Kelly 2007; Duckworth & Quinn 2009). Some have emphasized grit’s dependence on the notions of willpower and epistemic resilience: psychologists have found a high correlation between the capacity for perseverance and the self-control facet of conscientiousness (see Meriac, Slifka, & LaBat 2015), while philosophers have argued that grit involves an epistemic disposition to require especially strong evidence to revise one’s initial positive assessment of one’s chances of success (Morton & Paul 2019).¹ By contrast, both popular discussions of grit and assessments of its value in educational contexts portray hopefulness as constituting a distinctive source or dimension of the capacity for perseverance. For instance, in a book aimed at popular audiences, Angela Duckworth (who proposed the grit construct and developed the “Grit Scale”) writes the following (2016: 169):²

What is hope? One kind of hope is the expectation that tomorrow will be better than today. It’s the kind of hope that has us yearning for sunnier weather, or a smoother path ahead. It comes without the burden of responsibility. The onus is on the universe to make things better. Grit depends on a different kind of hope. It rests on the expectation that our own efforts can improve our future. I have a feeling tomorrow will be better is different from I resolve to make tomorrow better. The hope that gritty people have has nothing to do with luck and everything to do with getting up again.

Such emphasis on a form of “perseverance-conducive” hope echoes recent philosophical discussions. Cheshire Calhoun, for instance, has recently highlighted the importance of a distinctive kind of hope for temporally-extended agency, which she calls “substantial practical hope”, and which she defines as a “hope for the realization of one’s ends through one’s own efforts” (Calhoun 2018: 69). In a similar vein, Adrienne Martin (2020) has contrasted “personal hope”, defined as a hope to bring about a desired outcome through one’s agency, with “impersonal” or “idle” hopes for outcomes lying outside the scope of one’s influence. Both authors also oppose substantial practical or personal hope to “mundane” or “prosaic” hope, such as the hope for good weather, or the hope that one’s neighbor has a good day.³

Substantial practical hope, like all hope, appears to involve the belief that an outcome is possible and a desire for that outcome.⁴ But if it is to constitute a unique source of grit—as suggested by some depictions of that capacity, it must go beyond a belief-desire pair. For as we will see, it seems possible for two agents engaged in a difficult project, one having grit and the other lacking it, to desire success equally strongly, to both see its occurrence as unlikely, to exhibit equal willpower and epistemic resilience, while still differing in their motivation to bring success about.⁵ In such circumstances, which form of hopefulness do gritty agents exhibit, such that it can provide them with a unique motivation to persevere, irreducible to that of strong desire? And if a form of hopefulness can indeed count as an independent source of grit, how should we understand its contribution to gritty agents’ practical rationality?

My aim is to answer these questions. I want to explain why the kind of practical hopefulness highlighted by Calhoun and others counts as a standalone, philosophically interesting dimension of grit. In particular, I will argue that when facing decisions that can impact whether the outcome they desire materializes, gritty agents who have practical hope go beyond desiring success and seeing it as possible: they also exhibit what I call a “risk-inclined attentional pattern”. Gritty agents’ tendency to focus their attention on their desired outcome’s goodness as opposed to its odds of materializing leads them to adopt a specific weighting of two competing concerns we have over action: they prioritize the concern with choosing a course of action which might go very well over that of choosing a course of action which is very likely to go fairly well. By depicting gritty agents’ form of hopefulness as involving risk-inclination, we can explain their disposition to persevere when their non-gritty counterparts would give up, as well as the practical rationality of such perseverance. The result is a more inclusive picture of the sources of grit.

Sections 2 and 3 turn to existing views of substantial practical hope, to conclude that they either cannot account for gritty agents’ sustained motivation, or for their practical rationality. Sections 4 and 5 then introduce the notion of a risk-inclined attentional pattern as a dimension of gritty agents’ practical hope. Finally, Section 6 explains how practical hope can rationalize perseverance, in particular on a theory of practical rationality more permissive than standard expected utility—namely, Lara Buchak’s (2013) risk-weighted expected utility theory. Section 6 also draws lessons from the study of grit for decision theory more generally.

2. Grit-Conducive Fantasies

To motivate the intuition that the capacity for perseverance can be rooted in a form of hopefulness, consider the following two cases:

Aspiring Novelists

Gabriel and Jules have both completed their MA in Creative Writing three years ago. For the past two years, they each have invested themselves fully in the project of writing their first novels. They have eschewed all work commitments and have managed to spend time on their book manuscripts almost every day. They have submitted their manuscripts to the very same six publishers: all rejections. After the third rejection, Jules was still committed to the project of writing his first novel. But he has now decided to give up and try to find other work instead. Gabriel, on the other hand, keeps sending modified versions of his manuscript to publishers. He hasn’t lost hope to become a novelist, and it’s this hope that keeps him going.

Missing Child⁶

Emily and Tony Hughes are on holiday in a rural French town when their son Oliver disappears. Eight years later, he still has not been found. In the meantime, Emily has started a new life with one of the detectives who initially worked on the case. She has stopped looking for their son. Tony, on the other hand, is still continuing the search. After having seen a recent photograph of a boy wearing a headscarf identical to the one Oliver was wearing on the day he disappeared (and made for him with a unique insignia), he contacts the retired detective who originally led the search. The case gets reopened. Tony becomes actively engaged in the investigation, moved by the hope to find his son alive, while Emily refuses to participate.

In Aspiring Novelists and Missing Child, Gabriel’s and Tony’s form of hopefulness seems to provide them with an especially resilient motivation to pursue the outcomes they desire. By contrast, Jules and Emily do not display Gabriel’s and Tony’s resolve, despite sharing the latter’s strong desires for the relevant outcomes, as well as their verdicts concerning those outcomes’ low probability of occurring. Furthermore, Gabriel’s and Tony’s form of hopefulness appears to contribute to their practical rationality: whereas giving up seems like the rational option for Jules and Emily, persevering seems best for Gabriel and Tony. A satisfying account of hopefulness as a distinctive source of grit should make sense of these intuitive verdicts concerning hope’s motivational power and rationalizing influence.

Interestingly, cases structurally similar to the ones just presented figure prominently in the recent literature on substantial hope. Adrienne Martin (2014: 14–16), in particular, introduces the case of Alan and Bess, two cancer patients who have enrolled in the early-phase trial of an experimental drug. Both patients strongly desire to be cured and know that being cured is extremely unlikely. But whereas Bess’s hope that “she will be the 1 percent” is “what keeps her going” (Martin 2014: 15), as she plans for and projects herself into a future in which she lives, Alan instead dedicates greater energy preparing for his death. Martin takes such a difference in motivation to underlie a difference in motivationally potent, substantial hope. Her account of substantial hope as partly constituted by a “licensing stance” aims to capture what distinguishes hopeful agents such as Bess. On Martin’s view, those who have substantial hope endorse their desire for an uncertain outcome and see it as giving them sufficient reasons to engage in various hopeful activities. These include planning around the hoped-for outcome’s realization, experiencing positive feelings of anticipation and, importantly for our purposes, “fantasizing” about the hoped-for outcome’s materialization. In fact, Martin takes hope’s “typical expression” in fantasizing to explain why hope is often seen as having a unique motivational influence (2011; 2014: 95–103). Fantasizing is supposed to explain why hopeful agents are motivated to make decisions that their non-hopeful counterparts would not make. It is also supposed to explain why these decisions can count as practically rational.

In contrast with Martin (2014), my interest lies in the form of hopefulness displayed by gritty agents. That form of hopefulness seems quite different from other instances of substantial hope for outcomes lying outside the scope of one’s agency (such as Bess’s hope that her treatment will work). Substantial practical or “personal” hope, unlike “idle” or “impersonal” substantial hope (Martin 2020) for outcomes lying outside the scope of one’s influence, is directed at the realization of a desired outcome through one’s own efforts. As such, it seems unique in its capacity to motivate actions directed at that outcome’s realization.⁷ At the same time, despite differences in the actions they motivate, both substantial practical and idle hope are endowed with a motivational power going beyond that of desire (as the aforementioned cases illustrate). We can thus begin our inquiry into hope as a dimension of grit by turning to Martin’s (2011; 2014) account of the mechanism underlying substantial hope’s motivational influence. As her account was introduced prior to distinction between idle and practical hope, it might plausibly apply to both hope variants.

Martin argues that hope-generated fantasies tell stories in which our desires are satisfied. In doing so, they present us with previously unrecognized, valuable aspects of our ends, thus reinforcing our motivation to pursue them. Martin uses the case of a young writer who is trying to get pregnant and who is hopeful that she will succeed (2011: 163; 2014: 88–89). This writer has various fantasies about motherhood, one of which pictures her in the front porch, working on her first novel, while her baby naps in a swing. As the writer pictures this scene, it occurs to her that becoming a mother could affect her writing in interesting ways, by giving her the experience of a new range of emotions. In turn, this possibility appears to her as a new reason to become a mother, thereby strengthening her commitment to this project. Martin insists that the sense in which fantasies present one with reasons is distinct from that in which desires do. Since fantasies result from the “free play of the imagination”, they propose a world which is to some degree distant from the actual world. It is then up to the fantasizer to determine whether that worldly representation is accurate, and whether she should take up the reasons presented in fantasizing as the basis for pursuing certain ends (2011: 163–64; 2014: 99–100).

Applied to Aspiring Novelists, Martin’s account would yield the following result: if Gabriel remains motivated to pursue his project when Jules does not, it must be because he has discovered an additional reason to value becoming a novelist through having fantasized about it. Likewise, if Tony preserves the motivation to find Oliver even when Emily has already given up, it must because he sees finding his son as more valuable than Emily does, as the result of having fantasized about finding him. Martin’s proposal doesn’t allow non-hopeful, non-gritty agents to differ from their hopeful, gritty counterparts in terms of motivation to persevere while entertaining just as strong a desire in success. But such a possibility is salient in our cases: after all, as parents, both Tony and Emily might very strongly and equally desire to find their son alive, identify with their desire, and yet differ in their motivation to try to bring about that outcome. In fact, we can easily imagine Tony’s and Emily’s respective desires to find their son to already be so strong that no activity of fantasizing could possibly strengthen them. The notion of fantasizing thus seems insufficient to account for our cases’ apparent motivational asymmetries.

Martin’s proposal also identifies substantial hope’s contribution to practical rationality with that of desire: her explanation of why Gabriel can be rational in persevering while Jules can be rational in giving up would be that the former’s desire is stronger than the latter’s as the result of fantasizing. On a desire-belief model of rational action, as well as on an orthodox decision theory model, one single restricted set of actions counts as practically rational, given one’s beliefs and desires.⁸ Gabriel performs the action that counts as rational for him, given his beliefs and desires, while Jules performs the rational action for him, considering his different beliefs and desires. No room is left for the possibility that somehow owing to a difference in substantial hopefulness, both agents’ diverging courses of action count as rational for them, even assuming they share all relevant beliefs and desires.

Finally, Martin’s view depicts substantial hope’s sustaining force as contingent—a feature which seems at odds with hope’s role in grit cases. Martin acknowledges that fantasizing sometimes reinforces one’s desires, and sometimes does the opposite: it can “cut both ways” and propose a world containing undesirable features instead, thus undermining one’s motivation to pursue a previously desired end (2011: 164; 2014: 101–3). Such a contingent impact on motivation seems in tension with the possibility that hopefulness constitutes an important source of the capacity for perseverance. For we want an account of gritty agents’ mental states and dispositions which explains why exercises of grit are reliably linked with a motivation to keep pushing and follow through: grit, after all, is defined as the “capacity for perseverance”.⁹ Agents possessing grit and exercising it won’t always end up persevering in their projects: relying on one’s capacity for perseverance (and in particular on hopefulness as a source of that capacity) is compatible with giving up, as we shall see. But continued effort in trying to bring about a desired outcome should nonetheless be seen as the “default” posture of agents possessing and exercising grit. We need an account of their state of mind which makes sense of that default, as opposed to a view portraying their choice to persevere as contingent upon the deliverances of an imaginative activity.

3. A Phenomenological Idea of the Future Containing Success

To uncover the state of mind of gritty agents like Gabriel and Tony, we should turn to a different view of substantial practical hopefulness. Calhoun (who has introduced the notion of practical hope [2018: Chapter 4]) highlights its contribution to the management of our bounded rationality. She argues that practical hope, as a “hope to realize one’s ends through one’s own efforts” (2018: 69), facilitates sticking with commitments that we see as reasonably formed. It achieves this by replacing one’s “phenomenological idea of the future containing failure”—namely, one’s broadly unreflective sense of what the future is going to be like—with a different phenomenological idea, containing success. On this picture, practical hope is a state of mind needed to manage to do what one already ought to. It does not provide one with additional reasons to stick to one’s project, extraneous to those related to success’s value and likelihood already considered in initial deliberation. Hope instead counterbalances some detrimental psychological tendencies that can potentially sap our motivation to persevere and perform the actions that we in fact have most reason to do. Practical hope is needed because our motivation, as real human agents, is not solely determined by cognitive states apt to enter into practical deliberation, such as beliefs and desires. According to Calhoun, our motivation also depends on our “sense of the future”—contrasted with “beliefs or a conceptualization of the future” (2018: 71). This sense of the future has both a content, “much of which is unreflective” (2018: 74), and a qualitative character. It “operates as a background to propositional states” (2018: 73).

Calhoun’s view of practical hope as partly constituted by a phenomenological idea of the future containing success is supposed to explain why hopeful agents remain motivated in the pursuit of difficult goals. Calhoun argues that low odds of success can be demotivating, mainly because they encourage us to see our current activity as wasted effort whose costs will not be redeemed (2018: 85). Hopeful agents, however, do not let low odds pull themselves toward the idea of a failed future. Rather, their phenomenological idea of the future containing success enables them to act on the reasons they considered when they first decided to commit to their difficult endeavours.

In insisting on practical hopefulness’s role in structuring our background assumptions about how the future will unfold, Calhoun attempts to single out a unique motivational role for hope for success at projects we care about. But her notion of a phenomenological idea of the future could be more explicit, to make good on the claim that substantial practical hope has a distinctive motivational influence. For some manifestations supposedly characterizing one’s phenomenological idea of a future containing success can also be seen as manifestations of a desire to succeed—for instance, the disposition to “previsage a particular future in our imagination” (2018: 72).¹⁰ Another issue is that one’s phenomenological idea is supposed to have a particular content (about how the future will unfold), without however counting as a belief or a conceptualization of the future (2018: 71), since it is supposed to act as a “background to propositional states” (2018: 73). We might wonder whether this is psychologically plausible. For instance, could, on Calhoun’s view, two agents possess the same background understanding of the future while differing in their disposition to bring this understanding to the “foreground”? And in general, what does it mean to say that a background understanding and an explicit view of the future both share a determinate content?

A more important problem for our purposes with Calhoun’s view is that it does not portray practical hope as making any difference to the outcome of practical deliberation. Practical hope only comes in after one has already settled for oneself the “question of whether the pursuit is worth it” (2018: 85). Its impact occurs at the “phenomenological level”, that of one’s (often unconsciously held) background assumptions about how the future will unfold. In fact, Calhoun goes as far as suggesting that if we endow hope with a unique role within practical deliberation itself, we end up with the implausible result that hope pushes us into epistemic and practical irrationality: “The only option [if one holds that hope figures into deliberation] appears to be the unacceptable one that hope involves a distortion of the probability assessment one would have had, had one not been so hopeful” (2018: 83).

Such a statement downplays the possibility that substantial practical hope could both reinforce our pre-existing desires (owing to fantasizing in the ways Martin describes), while also “seconding” pre-existing commitment in Calhoun’s sense. Gritty agents who have substantial practical hope could tend to persevere both owing to their especially strong desires and to the kind of “acedia-preventing” phenomenological idea of the future Calhoun describes. But even a complex account along those lines would fail to address an important question raised by our original cases: why can persevering count as practically rational for a hopeful, gritty agent such as Gabriel while practically irrational for a non-hopeful, non-gritty counterpart (such as Jules) sharing all relevant beliefs and desires? In the next sections, I will argue that to answer this question and account for gritty agents’ unfading motivation and practical rationality all at once, we should appeal to the notion of a risk-inclined attentional pattern. By uncovering gritty agents’ disposition to attend to their desired outcome’s goodness as opposed to their low odds of success, we can capture both the motivating and rationalizing roles of substantial practical hope as an independent source of grit.

4. Practical Hopefulness and the Aims Over Action

I propose that gritty agents rely on a particular kind of substantial hope, centrally involving a belief that success at their difficult projects is possible, a desire for success, and a risk-inclined, attention-fueled way of balancing two competing aims we have over action. These are the aims of choosing a course of action which might go very well and that of choosing a course of action which is very likely to go fairly well.¹¹ When facing the decision of whether to persevere, gritty agents prioritize the former aim, as the result of focusing their attention on their desired outcome’s goodness (as opposed to its odds of occurring). I argue that gritty agents’ particular weighting of those two competing practical aims plays a role distinct from that of beliefs and desires in determining how they act. In particular, gritty agents’ risk-inclined attentional pattern explains their tendency to persevere when their non-gritty counterparts would give up. It also explains why such perseverance can count as practically rational, thereby accounting for substantial practical hope’s motivating and rationalizing roles.

One’s way of balancing the competing practical aims in a given case can be seen as a practical analogue to what some studying belief have called one’s “evidential threshold”. Epistemic agents have often been portrayed as balancing competing epistemic aims, namely those of believing truth and avoiding error.¹² The idea that practical agents also balance competing aims hasn’t loomed as large, but it can be made more precise by a parallel with the notion of an “epistemic weighting” or evidential threshold. It is often said that an agent who, in his belief-formation, puts more weight or emphasis on the aim of avoiding error than on that of believing truth will end up requiring comparatively more evidence before settling belief—a belief-formation practice captured by the notion of a “high evidential threshold”. By contrast, one who puts greater emphasis on the aim of believing as many truths as possible is often thought to require comparatively less evidence before settling belief, and thus viewed as having adopted “low evidential thresholds” for his various beliefs.¹³ An epistemic agent’s evidential thresholds are thought to guide her belief-formation without directly figuring as normative considerations or evidence justifying belief. For instance, on Jennifer Morton’s and Sarah Paul’s (2019) view of grit as centrally involving epistemic resilience, gritty agents form their beliefs about their chances of success on the basis of the evidence, but possess a disposition to adopt an especially high evidential threshold to revise their initial positive assessment of their chances of success, and conclude that they will fail at their projects. The present proposal points to an analogue weighting with respect to one’s competing practical aims of choosing a course of action which might go very well and of choosing a course of action which is very likely to go fairly well. On the present view, an agent’s way of balancing these competing practical aims guides her decision-making and practical deliberation without figuring as one more desire for particular outcomes or as a belief about which outcomes to pursue. One’s way of balancing the competing practical aims instead determines how one structures the attainment of the various goals one wants to achieve—either through “venturesomeness” and risk-taking, or through “prudence” and risk-avoidance (see Buchak 2013: 55–56). I propose that “venturesomeness” with respect to a particular goal, combined with a desire for success and the belief that success is possible, amounts to a substantial practical hope for the realization of that goal. Such a form of hope can count as an independent dimension of grit, complementing epistemic resilience.

To illustrate the role of one’s weighting of the aims over action in motivating and rationalizing perseverance, we can go back to Gabriel and Jules, who each have to decide between persevering in the project of becoming a novelist, or giving up and try to find other work instead. Gabriel and Jules each know that if they persevere, they have a small chance of enjoying a great outcome (actually becoming novelists), but an important chance of a rather bad outcome (in particular, ending up poor and without work). They also know that if they give up now, there is an important chance that they will end up with a job not as great as that of being a successful novelist, but that they will still be better off than if they continue trying only to have failed in the end. I hold that having access to all of these facts does not suffice to settle Gabriel’s decision, nor Jules’s. In order to arrive at a decision about which path to pursue, each agent also has to factor in their own weighting of the competing concerns with respect to action. In particular, if Jules prioritizes ensuring that his course of action almost certainly gets him a fairly good outcome, he is going to pass up some opportunities—such as the opportunity to persevere—involving the possibility of an even better outcome, considering that those opportunities also involve the possibility of rather bad outcomes. Likewise, due to his prioritization of the aim of choosing a course of action which has some probability of yielding him the best possible outcome, Gabriel is also going to pass up opportunities. In particular, he is going to pass up the opportunity to give up: even though giving up would guarantee him a fairly good outcome, it does not involve the possibility of getting the best possible outcome, which is that of actually becoming a successful novelist.

It seems that both Gabriel’s and Jules’s way of balancing the concerns over action is rationally permitted, and that this explains why persevering is practically rational for Gabriel, whereas giving up is practically rational for Jules. I come back to this point in the last section. For now, we can note the parallel with the epistemological view that there are various rationally permissible ways for agents to balance the competing aims of believing truth and avoiding error. Just like epistemic agents can plausibly be rational in holding different beliefs on the same body of evidence (as long as they do so through having used different rationally permitted, truth-conducive evidential thresholds),¹⁴ practical agents can also plausibly be practically rational in performing different actions despite sharing all relevant beliefs and desires, as long as they have used different rationally permitted weightings of the competing aims over action. I propose that agents who seem to have access to a source of grittiness going beyond willpower and epistemic resilience do not necessarily differ from their non-gritty counterparts in their beliefs, nor in the intensity of their desires. Instead, gritty agents often stand out due to their specific weighting of the competing practical aims. It is this weighting which, combined with their desire for an uncertain but possible outcome, makes them count as entertaining a form of substantial hope in their future success.

5. Attention and Risk-Inclination

We saw that substantial practical hope, as a hope to bring about a desired outcome through exercises of one’s agency, must go beyond a belief-desire pair to constitute a distinctive source of motivation and practical rationality. My proposal is that an emphasis on the aim of choosing a course of action which might go very well makes up substantial practical hope’s “third dimension”, beyond belief and desire. I now want to argue that practical hope’s weighting of the aims over action is attention-driven. This conception of grit-conducive, practical hope has its roots in classifications of hope (practical and non-practical alike) as an emotion (see in particular Ben-Ze’ev 2001: 475; Damasio 2003: 44; Helm 2009: 252; Tappolet 2016: 25–26), as well as in the view that emotions are constituted by patterns of salience and concern (see de Sousa 1987; Ben-Ze’ev 2001; Evans 2001; Elgin 2008). Katie Stockdale (2020) notes that most philosophers working on hope have so far set aside the question of whether it is an emotion, despite its important similarities with emotional states. Emotion theorists, by contrast, have included hope in their classifications, thereby emphasizing its intentional and evaluative character, its distinctive phenomenology and its motivational power.¹⁵ In fact, it appears that by treating substantial hope as an emotion, we might see why it can count as a unique source of motivation, going beyond its desire condition. For on the influential view of the emotions just mentioned, their special motivational force lies in their capacity to direct and focus attention. Fear, on such a view, motivates a flight response due to its constitutive tendency to direct attention to the feared prospect’s badness and to escape strategies. Likewise, guilt motivates behavior in making the agent attend both to the badness of what he did and to ways of making reparations (Brady 2013: 20–23; 2014: 62). In general, on the view that emotions are constituted by patterns of salience and concern, their role is to “fill gaps left by (mere wanting plus) ‘pure reason’ in the determination of action and belief” (de Sousa 1987: 195). Emotions circumscribe our practical and cognitive options through their constitutive influence on both perceptual and intellectual attention. Their unique motivational force thus derives from their capacity to alert us to objects and strategies of importance.¹⁶

Calhoun entertains the possibility that uncovering substantial practical hope’s motivational power requires viewing it as an emotion constituted by a specific attentional pattern. As she writes (2018: 79–80):

After all, emotions and emotional attitudes are, in part, distinctive patterns of salience. Danger is salient in fear, offenses and wrongs in resentment. Perhaps hope does double duty, making psychologically salient the desirability of the end and at the same time suppressing attention to what we actually believe about the odds. Hope would then consist in a belief that a future outcome is possible, a preference for that outcome, and a disposition not to think about the low probability of the outcome but instead to keep one’s eyes on the prize.

Calhoun, however, is quick to dismiss the possibility that substantial practical hope is constituted by what she calls an “Eyes-on-the-Prize” attentional disposition (2018: 80). She equates a disposition to focus on the desired outcome’s goodness and away from its actual odds of materializing with one to “ignore discouraging news”. She argues that such a disposition is bound to interfere with rational reconsideration, by keeping out of view evidence relevant to abandoning one’s pursuit.

Calhoun is right that an “Eyes-on-the-Prize” attentional disposition can lead one to miss out on evidence relevant to reconsideration. However, there is an important difference between paying less attention to one’s actual odds of success than one would have had, had one not been so hopeful, and being broadly insensitive to incoming evidence bearing on whether one will succeed if one continues trying. Focusing one’s attention on a desired outcome’s goodness and away from its probability seems compatible with retaining sufficient sensitivity to the fact that we are pursuing our goal under low odds, to reconsider if need be, and make contingency plans. Of course, such a stance can be somewhat difficult to maintain and cannot successfully be adopted by all gritty agents. But this might just be why, as Martin (2020) points out, practical hope is sometimes seen as a virtue—one which requires striking the right balance between sustained motivation to bring about a desire outcome and acknowledgment that success also depends on contingencies falling into place. Again, a parallel with epistemic resilience as a dimension of grit seems useful: one’s disposition to require especially strong evidence for the belief that one will fail can “backfire”, leaving one unable to revise one’s confidence in one’s future success when one should. But epistemic resilience can—and often does—also fall within the bounds of epistemic propriety, especially when displayed by epistemically virtuous gritty agents.¹⁷

On my view, gritty agents like Gabriel and Tony are in a mental state which does what Calhoun (2018: 80) calls “double duty”: it increases the salience of their goal’s desirability while decreasing the salience of success’s probability.¹⁸ I take such an attentional pattern to structure gritty agents’ stance with respect to the twin practical aims: through devoting considerable attention to their desired outcome’s goodness and not as much to its low probability, gritty agents end up giving greater weight, in their decision-making, to the aim of choosing a course of action which might go very well. On one way of explaining the relation between hopeful, gritty agents’ attentional pattern and their practical weighting, that connection is causal: gritty agents’ attentional dispositions cause them to adopt their particular weighting of the aims over action. Alternatively, exemplifying an “Eyes-on-the-Prize” attentional pattern might also simply constitute what is it for a gritty agent to prioritize, in a given case, the practical aim of choosing a course of action which might go very well. According to this latter suggestion, the relation between gritty agents’ attentional pattern and their particular way of balancing the twin practical aims would instead be constitutive. I propose to remain agnostic on which of these alternatives holds.¹⁹ What matters for our purposes is instead the plausibility of there being a connection between gritty agents’ attentional dispositions and their weighting of the aims over action—be that connection causal or constitutive.

All substantial hope—practical and non-practical alike—might plausibly involve the kind of attentional pattern discussed by Calhoun.²⁰ However, as we saw, practical hope is a specific kind of substantial hope: it is a hope to bring about a desired outcome through exercises of one’s agency or a “hope for the success of our own activities” (Calhoun 2018: 3)—the kind of “active”, agential hope displayed by gritty agents. As such, it seems unique in its power to impact one’s way of balancing the competing practical aims. Non-practical substantial hopes for the occurrence of some desirable thing, such as Bess’s hope that her treatment will work, can enable us to live well and cope with difficult situations. But since their object is not the realization, via our agency, of an outcome that seems to lie more squarely within our influence (such as becoming a successful novelist, or finding one’s son), idle substantial hopes seem unable to impact our weighting of the twin practical aims.²¹

On my view, we can criticize gritty agents for being mistaken about whether the outcome they desire is possible and can indeed be realized via their agency, as well as for wrongly evaluating its probability of occurring. My view also leaves room for criticisms directed at gritty agents’ desire for success: when success is not as valuable as it appears (especially considering one’s other projects and goals), a strong desire representing it as being “good” can be deemed “unfitting” (see Howard 2018). Do these norms exhaust the rational standards governing substantial practical hope? In particular, if a gritty agent’s practical hope centrally involves an accurate probability assessment and a fitting desire, can it nonetheless, on my view, rightfully be deemed irrational? We might think that there is something objectionable in a gritty agent’s activity of pinning a very strong hope on a highly unlikely outcome, even when that hope is evidentially supported and grounded in a fitting desire.

My view accounts for such an intuition: evidentially-supported practical hopes directed at truly valuable prospects can nonetheless fail as hopes, due to the specific, attention-fueled weighting of the practical aims they involve. Especially when one’s initial odds of success are very low, placing heavy emphasis on the aim of choosing a course of action which might go very well will often be “unreasonable”, in the sense of constituting an obstacle to the realization of other desires directed at more probable state of affairs. To put the point differently: someone who places great emphasis on what happens in the best-case scenario will sometimes forgo many good things, as the result of failing to appreciate the importance of what happens in non-best-case scenarios. (Consider, for instance, a version of Gabriel’s case in which his continuing to send modified versions of his manuscript prevents him from working toward other, more probable things that he also values, such as achieving a work-life balance.) I hold that when an agent’s risk-inclined weighting of the practical aims undermines the realization of his overall set of projects and goals, that weighting can be deemed “unreasonable”. I thus side with Martin (2014: 50–51) in thinking that “prudential norms” or “norms of reasonableness” constraint substantial hopefulness’s “third dimension”.²² As is the case with one’s “evidential policies”, however, the “permissible band” of what counts as a reasonable way of balancing the twin practical aims seems to be quite large, such that many weightings will count as acceptable—a point to which I turn in the next section.

6. Practical Hope’s Rationalizing Influence

The present view traces gritty agents’ unfailing motivation to persevere back to their attention-fueled, risk-inclined weighting of the concerns over action. It thereby makes perseverance in difficult projects not simply be a matter of epistemic resilience and willpower. I will now argue that the notion of a risk-inclined attentional pattern can also clarify the contribution of grit’s hopefulness dimension to practical rationality. In particular, it can allow us to explain why perseverance can both be practically rational for a gritty agent and practically irrational for a non-gritty counterpart sharing all relevant beliefs and desires. The key to understanding how practical hopefulness can rationalize perseverance is to turn to a theory of rational decision-making more permissive than standard expected utility (henceforth, EU), namely risk-weighted expected utility theory (henceforth, REU), as defended by Lara Buchak (2013). We will see that homing in on the phenomenon of hopeful perseverance in difficult projects can pave the way for the best possible formulation of REU.

Risk-weighted expected utility theory holds that rational agents’ answer to the question of what to do is determined by their utilities, their credences and their risk-attitudes. One’s risk-attitudes express one’s weighting of the concerns over action (Buchak 2013: 53–56) and are different in kind from beliefs and desires about how much risk one should tolerate. REU is partly motivated by cases where expected utility seems too Draconian—in particular, cases where EU predicts that agents should be indifferent between gambles, but where those who prefer one gamble over the other do not intuitively seem practically irrational.²³ REU is also motivated by the related thought that practically rational agents care about “global properties of gambles” (centered around the mean, spread out, etc.), as opposed to always preferring gambles with the highest expected utility.

According to REU, risk-attitudes determine how agents weigh the utilities of the various outcomes associated with a gamble—such as the gamble of persevering in one’s project of becoming a novelist. First, a rational agent can aggregate utility values by weighting utilities by their probabilities. Such an agent would, for instance, sum up the utility of each of the possible consequences of persevering, individually weighted by their respective probability of occurring. If, for instance, persevering in the novelists’ case has three possible consequences (let’s say, ending up very poor and without work (C), or ending up poor and without work (B), or successfully becoming a novelist (A)), and each of these consequences has a respective probability of ½, ¼ and ¼, then the expected utility of that gamble would be the following:

This is equivalent to:

Here, the value of a gamble is represented as its worst possible value (ending up very poor and without work), plus the interval difference (in terms of utility) between the worst value and the second worst value, weighted by the probability of getting at least the second worst value, plus the interval difference between the second worst and the third worst value, weighted by the probability of getting at least the third worst value; and so on.

But REU holds that there are other rationally permissible ways for agents to aggregate utility values, besides the one just presented, in order to determine the overall value of a gamble. In fact, it holds that the weight that each outcome’s value gets in an agent’s evaluation of a gamble as a whole is “up to the agent”, and determined by her risk-attitudes $\text{(}r\text{)}$. This means that the REU of the perseverance gamble is the following:

An agent can set her risk-attitudes $\text{(}r\text{)}$ such that $r\text{(}p\text{)}=p$ (where $p$ is the probability of a given outcome). Such a person would aggregate utility values in line with EU.²⁴ But according to REU, one can also permissibly set $r$ such that $r\text{(}p\text{)}=p^2$. For that person, the REU of the perseverance gamble would be the following:

For someone with that risk-function, benefits that are obtained in a smaller proportion of states get proportionally less and less weight. This means that worse outcomes will count proportionally more, and better outcomes will count proportionally less in her determination of the overall value of a gamble. Buchak (2013: 64) deems such an agent, who sets $r$’s value such that $r\text{(}p\text{)}$ is smaller than $p$ for all $p$, “risk-avoidant”. By contrast, an agent who sets $r$’s value such that $r\text{(}p\text{)}$ is equal or greater than $p$ is “risk-inclined”. As benefits obtain in a smaller and smaller portion of states, they get proportionally more and more weight in her evaluation of gambles. Worse outcomes will count proportionally less, and better outcomes will count proportionally more.

The present model of rational decision-making explains why Gabriel and Jules can both be practically rational in persevering and giving up (respectively): Gabriel is rational in persevering because he is risk-inclined, whereas Jules is rational in giving up because he is either “risk-neutral” or risk-avoidant. We also have an explanation of why Gabriel’s mental state as a gritty agent makes a difference as to what counts as a rational action for him. Through directing his attention toward the hoped-for outcome’s goodness and away from its low odds of materializing, what Gabriel is actually doing, from the point of view of REU, is to give a great weight to a benefit that is obtained in a small proportion of the states associated with the decision to persevere. Because of his hopeful attentional pattern, Gabriel can have the acceptable risk-function of someone who is risk-inclined.

I say “can”, since it is possible that Gabriel’s preferences cannot be represented by a risk-function deemed acceptable by REU. An agent who follows maximax, for instance, can plausibly be seen as prioritizing the concern with choosing a course of action that might go very well as he conducts his practical deliberation. But maximax (the rule which says to always prefer the gamble with the highest maximum) counts as a limiting case of the $r$-function (Buchak 2013: 68–69). The reason is that for the agent following maximax (who can be represented in REU as having a discontinuous $r$-function),²⁵ the value of a gamble is that of its best possible outcome. But this rule has the unhappy consequence that such an agent is indifferent between all gambles, since every act has at least some chance of leading to a really great outcome. Such observations have led Buchak to discard maximax as an acceptable $r$-function, and assume that the $r$-function is continuous. (This fits well with our previous observations that some weightings of the aims over action should be deemed practically pernicious and unreasonable, and that the warrant for them is therefore partly “strategic”.)

In her initial defense of REU, Buchak (2013) assumes that one’s risk-function (as set by one’s risk-attitudes) remains constant across domains of decision. On this view, if an agent prioritizes the aim of choosing a course of action which might go very well (thereby setting $r$’s values such that $r\text{(}p\text{)}$ is equal or greater than $p\text{)}$, her risk-inclined risk-function would, for instance, carry over to inform both decisions in the health domain (such as the decision of whether to buy health insurance), and decisions related to one’s career (such as the decision of whether to persevere in the project of becoming a novelist). On this way of developing REU, either one is risk-inclined across the board, or one is simply not risk-inclined.

The view that risk-attitudes are domain-invariant does not sit well with the present account of grit’s hopefulness dimension. For we seem to be more disposed to have practical hope in certain domains than others, and thus seem to have different weightings of the aims over action across domains. For instance, when it comes to making decisions about whether to keep competing in a recreational sport (in the face of setbacks, injuries, etc.), one might prioritize the aim of choosing a course of action which is certain (or at least very likely) to go fairly well, whereas when it comes to making career choices, one might prioritize the aim of choosing a course of action which might turn out very well. It appears that one does not possess the same attentional dispositions in both of those domains: for instance, one might have a tendency to dwell on the low odds when considering the possibility of recovering from an injury, while being free of such a tendency when thinking about one’s chances of succeeding at one’s chosen career. Such differences in grit-related attentional dispositions indicate that one’s risk-attitudes and risk-function vary across domains of decision.

This implies that REU’s assumption concerning the invariability of one’s risk-function should be relaxed if REU is to explain practical hope’s contribution to gritty agent’s practical rationality. Buchak’s initial reservations about making REU more permissive and allow risk-attitudes to vary across decision domains come from her commitment to the “Ordering Axiom” underlying REU. This axiom states that all outcomes are comparable. Buchak worries that if risk-attitudes are allowed to differ across domains, we will be drawn to thinking that we cannot compare outcomes that fall within different domains (“otherwise, it would seem odd to have different preference norms in different domains” [Buchak 2013: 80]), and thereby abandon the Ordering Axiom. But as Buchak herself points out when discussing that axiom (2013: 94), it is not even clear that all goods are comparable²⁶ and thus, that the axiom in fact holds. Incomparability has not been discussed much in decision theory, some assuming that it rationally licenses the agent in making an arbitrary selection.²⁷ But Buchak insists that choices made in incomparability cases are not simply arbitrary—something decision-theory should account for (2013: 94). She has even argued elsewhere (Buchak & Murray 2019: 6) that one’s risk-attitudes are part of the explanation of why decisions made in cases of incomparability are not arbitrary selections. Her view is that risk-attitudes resolve the underdetermination left at the level of utility and credence, by setting the agent’s answer to the question “How to aggregate?”. Buchak thus appears to grant the existence of incomparability cases, while even suggesting that risk-attitudes play a special role in rationalizing action in such cases. This indicates that proponents of REU need to reconsider the allegedly “global” character of risk functions, or at least provide better normative reasons in favor of the unique, global $r$. As Brad Armendt (2014) notes, as REU currently stands, none of its other axioms besides the Ordering Axiom seem in tension with permitting different risk sensitivity values in different regions of preference.²⁸

I have argued that when construed as involving a risk-inclined attentional pattern, grit’s hopefulness dimension can plausibly rationalize perseverance on a standard of practical rationality more permissive than EU. However, I should emphasize that practical hope, as a facet of grit, does not always suffice to make perseverance come out as the practically rational option. In particular, when one’s initial probability of success at a difficult endeavour is particularly low, giving up will often be better than persevering, even for a gritty agent who has practical hope. For rational action does not only depend on one’s risk-attitudes, but also on one’s utilities and on one’s credences: if an action can bring about a highly valued outcome, but its probability of actually yielding it is very low, then it often won’t be the rational option, even for a risk-inclined agent. Gritty agents’ practical hope is therefore not a panacea, with the power to rationalize perseverance in even the direst of situations.

Practical hope nonetheless appears to constitute an important source of the capacity for perseverance—one that exerts its motivational and rationalizing influence independently of willpower and epistemic resilience. Gritty agents can rely on their capacity to resist the “sirens of temptation” to continue taking steps toward their goals despite appealing distractions. They can also rely on their epistemic disposition to require especially strong evidence to revise their initial confidence in their future success. But since difficult endeavours are “difficult” precisely because they involve especially low odds of success and compelling evidence that one might fail, epistemic resilience and willpower will often have to be supplemented by practical hopefulness. After all, those who have epistemic resilience won’t always manage, in the face of compelling evidence indicating failure, to count as epistemically rational in maintaining the level of confidence required for continued effort. I have argued that when confidence that one will succeed if one keeps trying seems misplaced, but gritty agents still press on, their capacity to persevere cannot be primarily grounded in an epistemic disposition. One’s capacity for perseverance is also often rooted a form of hopefulness in one’s future success. Grit is thus multi-faceted: when willpower and epistemic resilience run out, practical hopefulness takes center stage.