3.1. The difference between commitments and intentions

While commitments and intentions are often closely related, they differ in important ways. A key difference is that commitments are made with consideration of what is genuinely worth pursuing. When we intend to go for a walk or pour another cup of coffee, we are not expressing deeply held values or judgments about the worth of such actions. However, commitments are closely linked to our sense of who we are and what we value. After all, we would not invest our valuable time or forgo other pleasurable pursuits unless we genuinely believed that our commitments align with what matters most to us. In the case of commitments, we not only simply use our capacity to decide, but we also use our capacity to assess reasons in considering whether to make a commitment. But we are also good at tricking ourselves into believing that we will genuinely put the effort required to keep our promises. This reveals a crucial point: the disparities between judgments and beliefs are also observable in the case of commitments because commitments are closely linked to value judgments. Our self-bias and optimism about our abilities create a gap between identifying a good reason for a commitment and following through with the necessary effort to honor our goals.Footnote ¹⁰ We make decisions based on what we consider valuable to achieve, but we cannot reliably predict if we will develop the habits needed to stay committed. While all commitments may involve decisions, not all decisions qualify as commitments. So, the breakdowns in the cases of judgments and beliefs are also common in the case of decisions and commitments. Just as judgment and belief can diverge, so too can our decisions and our ability to sustain our commitments because commitments (unlike ordinary intentions) are not immediately permeable by conscious decisions.

This brings me to the second related point. Unlike our ordinary intentions, which we might form out of compulsion or whim, commitments require us to resist psychological challenges and maintain consistent self-discipline. While the function of a decision and intention, as Paul proposes, is to activate standing dispositions, the function of a commitment is to resist them. Thus, while a decision and an intention to Φ might overlap, the decision to commit and the state of having a commitment cannot overlap, because the function of a commitment is to counteract our standing dispositions. We make a commitment when we know that the intended goal cannot be brought about without some diachronic self-discipline and regular patterns of action. We choose our commitments freely, but in doing so, we bind ourselves to a series of decisions and actions needed to sustain the commitment and achieve the desired outcome. Because we are weak-willed agents, making a commitment is a way of holding ourselves to a standard, disciplining ourselves, organizing our desires across time, and putting ourselves to the test – to see if we will show ourselves the resolution to continue, despite wavering motivations. When we commit to being in a relationship, we promise not to give up after a trivial disagreement; when we take responsibility for writing a book, we make the commitment not to give up on it after a minor setback, or whenever there are more interesting things to do. If someone claims to be committed to writing a book but keeps changing his mind when a more interesting topic arises, we might judge that he lacks real commitment and has mistaken desire for a sincere decision. Insisting he is committed in this situation would either show a misapplication of the concept or an inability to understand one’s commitment.

In contrast, ordinary intentions are compatible with being tempted by alternatives. Indeed, for example, if I intend to go to the shop to buy a skirt I saw online, but if I change my mind after seeing another one I like better, this does not negate the existence of my initial intention. However, unlike these (ordinary) intentions, knowing one’s commitments requires whether one typically has the determination to resist constant changes of mind.Footnote ¹¹ After all, commitments like writing a book or completing a PhD involve taking on challenging tasks that demand the development of self-discipline to resist self-sabotaging tendencies.

Third, commitments often require regular patterns of action to make progress or complete a task. While we can fulfill an ordinary intention with a single action, commitments involve taking on challenging tasks that demand ongoing effort through repeated actions to reach the goal. Thus, as I will further argue, knowing a commitment requires stronger evidence than knowing an ordinary intention.

A potential objection to this view is that failing to follow through on an action might be attributed to a mere lapse in willpower rather than an actual lack of commitment. I will now discuss why this is not necessarily the case and show how weakness of will can, in fact, surprisingly reveal a fundamental failure to uphold the commitment.

3.2. Weakness of will as a defeater of the existence of commitments

As many have pointed out in the case of intentions, when there is a breakdown between a decision or an intention about what one ought to do and one’s action, that breakdown does not impugn the existence of the former (Broome Reference Broome, Bermúdez and Millar2002; Raz Reference Raz1978; Shah Reference Shah2008).Footnote ¹² According to Shah, what we achieve through deliberation is forming an intention to act, while the ability to carry out that intention lies beyond the scope of deliberation itself. He rejects the idea that deliberation and intention are limited to matters within one’s “voluntary control” (Shah Reference Shah2008: 4). That is, we can intend what we ought to do, and whether we believe we will succeed in it is a separate matter.Footnote ¹³ Sometimes, the breakdown merely reflects fear, which might hinder the initiation of an action, but it does not mean that one misjudged or changed one’s mind about what one ought to do. This view may be accurate regarding our understanding of ordinary intentions, but it does not apply to commitments. When we fail to act on something within our control, it can cast doubt on whether we genuinely made that commitment in the first place. After all, the whole point of a commitment is often that we undertake concrete efforts to resist weakness of will.Footnote ¹⁴

Here, I want to flag two features of weakness of will that challenge the existence of commitments: first, foreseeable and preventable weaknesses, and second, regular patterns of failure to manage such weaknesses.Footnote ¹⁵ Specifically, foreseeable, preventable forms of weakness of will that are in tension with our commitments serve as crucial evidence against the commitment, whereas an accidental uncontrollable impulsion or overwhelming desire does not. For example, if I intend to give a talk but am unable to attend due to a debilitating panic attack, my deliberation and initial decision remain intact, despite my physical inability to follow through. In such cases, the issue is not a failure of agency but rather external factors such as compulsion, fear, or biological constraints.

However, it’d be misleading to treat weakness of will as too close to physical inability or biological factors that are completely external to our will. Not all kinds of weakness of will are beyond our agential control. To be sure, we can encounter an unpredictable physical or psychological impediment that we could not foresee. However, weakness of will cases are not those in which one is simply overwhelmed by emotion or impulse, in which there is no internal failure of agency. While some forms of compulsion and intense, sudden emotions that are wholly uncontrollable may in fact be caused by physical or biological factors outside of one’s will, we can also form commitments precisely to combat such weaknesses and to make our dispositions consistent with our commitments. As Odysseus shows us, it can be necessary to deprive oneself of the opportunity to choose wrongly such as by tying himself to his ship so that he cannot join the sirens. We tend to judge that one only made a shallow commitment if he allows his dispositions or other desires to easily get in the way of realizing his goals. Weakness of will cannot be explained by “unconscious” intentions or dispositions either, because when we experience it, we feel guilty for not exerting adequate self-control.

Even when weakness of will results from a physical or psychological issue, we can still consider whether we can manage these weaknesses to prevent them from hindering our ability to make progress on a task. If I am indeed committed to making public speeches, I could look into the solutions about what I could do to prevent my panic attacks, whether I could use therapy or more practice beforehand. We might have psychological or physical impediments that will foreseeably sabotage our goals, and we can reasonably take precautions to prevent them. This is not to say that we are always responsible for weakness of will, regardless of our efforts. To be sure, even with therapies and interventions, one might not fully overcome a panic attack. The key point is that to evaluate the degree of one’s commitment, we consider the efforts we put to resist or prevent foreseeable and preventable weaknesses of will, rather than merely considering whether one has made a conscious decision. Therefore, it is necessary to distinguish between wholly uncontrollable emotions and impulses, and foreseeable and preventable weakness of will, which is within one’s control and can be prevented by appropriate countermeasures. Only the latter impugns the existence of a commitment.

Considering a different example of weakness of will that is less physical in nature might be more helpful. Take, for instance, a writer who decides to finish her book but finds that spending excessive time in coffee shops impacts her writing productivity and recognizes that this habit interferes with her goal. If she believes that she is committed to working on her book project regularly but continues to spend too much time in coffee shops, knowing full well that this habit distracts from her writing time and focus, her consistent failure to alter her routine reveals that her commitment to the project is not particularly strong.

That said, we should concede that a single instance of accidental weakness of will may not be enough to undermine the commitment, as occasionally yielding to weakness of will might actually be beneficial for regaining energy and sustaining motivation.Footnote ¹⁶ However, as I pointed out in the previous section, unlike ordinary intentions, commitments involve undertaking challenging actions that require regular patterns of actions to make progress or complete a task. Therefore, the claim is that it is the regular patterns of foreseeable, preventable failure to resist one’s weakness of will that ultimately calls into question the existence of a genuine commitment. And this is supported by the widespread experience that when agents fail to follow through on significant commitments, they often find themselves questioning the existence of those commitments. When weakness of will becomes a recurring theme, it compels us to critically assess the sincerity of our commitments, urging us to confront whether our initial resolutions were genuinely made or merely aspirational.

Of course, one might do everything to control one’s temperament, take precautions, yet one still might fail. In such cases, it would be unfair to think that these entirely accidental weaknesses of will immediately defeats the existence of a commitment. However, even if we do not fully succeed in overcoming these weaknesses, when assessing our commitments, what matters is the effort put into resisting to our weaknesses or dispositions. The existence of a commitment would not, and should not, be judged solely on a mental state or a single moment of decision.

4.1. An externalist proposal

Given that regular patterns of weakness of will undermine the existence of commitments, I can now sketch why we should adopt an externalist view to understand knowledge of commitments, drawing a surprising contrast with the recent view of intentions introduced in (1) at the outset of the paper. My argument for this externalist view appeals to the following claim: if S systematically fails to Φ in relevant circumstances, then S should infer that S is not committed to Φ-ing in those circumstances. We may believe that we are committed to Φ in particular circumstances, but we cannot know this until we have sufficient evidence that we actually will reliably Φ.

Knowledge of commitment cannot depend only on one’s internal mental events, a traditional assumption that is associated with Descartes. I argue for an externalist view because a commitment to Φ has its own external correctness conditions, which are objective insofar as they cannot be altered purely through acts of will or thought. For instance, there are external conditions of satisfaction for whether a person is committed to writing a book, which cannot be altered through pure acts of will. While we may discover and explore new ways to write a book, the introduction of new correctness conditions is appropriate only if these adjustments enhance efficiency in making progress toward our goals. Self-knowledge of commitments requires a “reality check”: one must look to external evidence to determine whether one typically reason and acts accordingly.

Contrast this with our self-knowledge of ordinary intentions. Ordinary intentions are linked to short-term actions, and in many cases, my intention to Φ and my knowledge of that intention are inseparable; for example, my intention to stretch my limbs is typically intertwined with my knowledge of that intention. In contrast, commitments are not merely one-time actions; they involve a broader range of actions and attitudes over appreciable periods of time. And there are correct and incorrect, feasible and infeasible, ways to go about realizing one’s goals. They are the kinds of actions which, if one is disastrously wrong or ignorant of them, or fails to act in accordance with them, undermine one’s pretention of being truly committed to Φ. So, knowledge of commitments can be confirmed or undermined by agents’ regular patterns of action. Whether you truly are committed to the relationship depends not just on what you intend at the time of undertaking it: it depends also on how you act and react over time (in which case, there might not yet be a fact of the matter about whether you are committed or not).

One might object that even if S’s failure to reliably Φ implies that S is not committed to Φ-ing, it does not follow that, in order to know in advance that S is committed to Φ-ing, S must be certain that she will reliably Φ in relevant circumstances.Footnote ¹⁷ In other words, one might contend that I have demonstrated only that if an agent fails reliably to Φ, it undermines her knowledge of her commitment to Φ – not necessarily the commitment itself. If I believe I am committed to Φ-ing but then observe that I am not reliably doing so, I should abandon that belief, not the commitment.

My response to this objection is that commitments function precisely to make us think about how to overcome temptations, weakness of will, and other challenges that we are capable of preventing. So, if an agent does not start to resist her inclinations or trying to improve her relevant skills, then it is hard to say that she still has or did make a genuine commitment. That said, I acknowledge that consistently acting in ways that are incompatible with Φ-ing is a more conclusive defeater than merely failing to act in ways that would bring about Φ-ing. Acting in ways that directly contradict our main goal is a stronger indication of its defeat than simply not trying enough to achieve the goal.

Further, we might desire to have a commitment, and we often make irrational commitments, knowing full well that we might fail. In those cases where we do not try enough, the commitment remains merely aspirational rather than a genuine one.Footnote ¹⁸ To be sure, this does not imply that we cannot (or should not) make commitments if we are unsure of our ability to follow through. As Marušić (Reference Marušić2015) argues, we can resolve to do something even if the evidence from our own past makes it likely that we will not follow through. We can commit to actions on the basis of practical reasons, and despite the evidence, even when we are unsure of our chances of success. Marušić does not aim to explain how one can know that one has a commitment. His point is that one can resolve to do things, despite contrary evidence from one’s past conduct. This is compatible with my claim that knowing whether one is committed requires evidential support related to the actions necessary to achieve one’s goals. My claim is that independent of whether we are rational or irrational in making commitments, the epistemic story about how we discover them will be the same. We learn about the extent of our commitments by observing our regular patterns of action, how much we struggle with our weakness of will, how well we control our behavior, and what steps we take to maintain them. Whereas ordinary intentions are not defeated by a lack of follow-through, commitments are undermined or confirmed by what we do.

One might also object as follows: suppose someone is committed to helping the poor and starts an organization that sends college students to build houses in impoverished areas.Footnote ¹⁹ It may turn out that this effort further impoverishes the area by taking jobs away from local construction workers. Do we then want to say that this person was not truly committed to helping the poor?

In response, I identify two forms of extra-mental evidence about propositions of the form “S is committed to Φ.” The first facts are about what we intentionally do: that is, whether S acts in such a way as to intentionally Φ in a consistent fashion. The second facts are regarding which actions actually promote achieving Φ effectively. In the example above, our good doer is ignorant of the second set of facts. But the first body of facts strongly supports the view that they are committed: their attempts are serious and consistent with their stated aim, and with what they think is necessary or conducive to it. Yet, I hold that these discrepancies are less likely to occur in the case of commitments than in ordinary intentions directed toward single, one-off actions because commitments require learning from trial and error through regular patterns of action. That is, commitments demand staying attuned to reality and the effects of our actions in the world, as this is the only way to succeed or make progress toward our goals. Further, ideally, discrepancies between these two types of facts will be rare, as committed agents should be sensitive to external feedback about the efficacy of their efforts. If one truly cares about and commits to Φ, they will consider which actions are most conducive to achieving it. While we may sometimes fail to choose wisely, what matters is the body of evidence showing that one sincerely, consistently, and justifiably believes that actions x, y, and z will bring about Φ.

It is true that some features of our actions are strongly external. For example, whether you end up in Boston depends on whether the plane actually takes you there as promised. You might intend to go to Boston but end up in Moscow due to completely external, unforeseen circumstances. Other conditions are only weakly external. In such cases, what typically justifies one’s commitment is whether the agent made sufficiently reasonable, determined, and consistent efforts to accomplish her goal and whether any mistakes made about how to achieve it were non-culpable (and, if mistaken about how to get there, whether those mistaken beliefs were formed non-culpably).

My claim is that self-knowledge of commitments is external in a weak sense. The relevant external conditions are the foreseeable norms for fulfilling an action. For example, if we are committed to running a marathon, we need to follow a training regimen, maintain proper nutrition, and build endurance. There is shared, intersubjective knowledge about how to train for a marathon effectively. While we cannot control every external factor – such as weather or injuries – and might need to adjust or even abandon our intentions when those conditions are unmet, we are still bound by the norms for preparation if we intend to complete the race. Evidence about how well those norms or conditions are met justifies our beliefs about our commitment.

4.2. Denying first-personal privileged epistemic access

Now that an externalist proposal is on the table, I can elaborate how this view helps us evaluate and justify our commitments both to ourselves and others. If knowledge of commitments can be confirmed or undermined by agents’ regular patterns of action, then these actions are, in principle, observable by others. This is why, while we might have first-personal authoritative epistemic access to our ordinary intentions, we lack such access to our commitments.

This is evident in our tendency to challenge the commitments of others when they repeatedly fail to take adequate steps toward achieving their stated goals or meet the external conditions necessary for success – conditions that, in principle, can be observed by others. Consider the case of a CEO who publicly claims to prioritize inclusivity but consistently neglects critical issues like pay gaps between men and women, reports of discrimination, and promotes the same homogeneous leadership team. While their public statements may suggest a commitment to inclusivity, the absence of concrete actions raises serious doubts about the existence of that commitment.

To be clear, the focus of this paper is on intrapersonal commitments rather than interpersonal ones, as the latter is not a form of intention.Footnote ²⁰ For example, if A tells B, “I promise to attend your talk,” A has made a promise, regardless of whether A shows up or consistently breaks promises. A’s patterns of behavior, past or future, are simply irrelevant to whether or not he has promised B that he will come to her talk. In such cases, the key fact is that the promise-maker took responsibility to keep his promise, not for his actions. Interpersonal commitments are performative statements that might have no bearing on achievement, achieving any kind of accomplishment, or intending to achieve any such success. My analysis focuses only on intrapersonal commitments, which require a different analysis. However, it is still worth noting that A might still be criticized, not just for breaking the promise, but also for failing to make a genuine effort to keep it. This distinction between lacking commitment and breaking a promise highlights the relevance of intrapersonal commitment, even in cases of interpersonal unfulfilled promises.

In some cases, an agent might insist that he is committed but consistently takes the wrong means to achieve his goal. This situation is less likely to occur with commitments than with ordinary intentions because commitments involve regular patterns of action that can be evaluated and adjusted over time; they thus allow more room for a reality check on whether we are on the right track. In fact, we would and should also assess whether an agent is committed, considering to what extent they are sensitive to reality checks and the relevant outside feedback in making progress toward their goals. If we care about our commitments, we engage in ongoing evaluation which makes it harder to overlook the constraints that affect the effectiveness of our actions. Commitments require us to be in touch with reality and assess the effects of our actions in the world to make progress toward our goals. Just as we risk being irrational if we maintain a belief despite contrary evidence, we risk being irrational if we insist that we are committed while consistently using ineffective or contradictory means to our ends. Authoritative first-personal privileged epistemic access to our momentary decisions does not eliminate this risk.

I noted at the outset that my view is neither behaviorist nor dispositionalist. Dispositionalists hold that we know ourselves through observation, just as we observe and infer what others believe and intend. As Schwitzgebel insightfully argues, believing primarily involves having stereotypical dispositions to reason and behave in ways that we would typically associate with having that attitude (Schwitzgebel Reference Schwitzgebel2002: 250–53). And, indeed, others might be better judges of what we believe. Although I propose an externalist view according to which knowing one’s own commitments is much like knowing those of others, I do not endorse a dispositionalist account of commitments. On an extreme version of a dispositionalist analysis, any time I act against an expectation that I will do A, it is because I have a disposition to do B instead. Yet, this would be problematic for a theory of commitment. It would fail to explain why we regret those deviations or pride in our successes, or “reactive attitudes”Footnote ²¹ toward others’ failures to uphold commitments, when their commitments matter to us.

By contrast, commitments are quite different from other goal-directed actions to which we are not committed. So, the evidence for them is not purely observational. As a type of intention, our commitments are not simply facts we uncover by observing ourselves; they are actions we choose to perform. While such observational data can undermine a claim of being committed, our reactive attitudes, such as regret when we fail and pride when we succeed, indicate that we take responsibility for our commitments, recognizing them as more than just dispositions or habits.

Denying the privileged epistemic authority does not require choosing between a first-personal and a dispositionalist approach. An extreme dispositionalist view would imply that any failure to act in accordance with our commitments implies we never truly had the relevant dispositions. However, our actions might vindicate the presence of a genuine commitment as we can work on overcoming the dispositions that hinder our goals. By adjusting our dispositions or acquiring new ones, we prove our commitment. While I agree with the transparency view that commitments are first-personal because it is up to the agent what to commit to and initiate the action, I reject the idea that the first person is in an epistemically privileged or authoritative position to know her commitments. Therefore, we can still preserve some of the virtues of the transparency view without losing the epistemic merits of an externalist, evidentialist account.