Is it reasonable to regret things one did?

By Ben Long

In a majority of cases where regret informs our decision-making processes, it can be said to be reasonable. I contend that the reasonability of regret primarily stems from its role in allowing us to actualise as moral agents, through the functions of moral ownership, communication, and authorship. Views that deem regret unreasonable fail to account for these normative features of regret by focusing on a restrictive hedonic view.

To evaluate the reasonability of regret, we must first outline what regret is. I will rely on Bernard Williams’ notion of ‘agent-regret’, a negative emotion directed at an agent’s previous actions. Here regret relies on a counterfactual situation of how one could have acted and ‘the focus of the regret is on that possibility, the thought being formed in part by first-personal conceptions of how one might have acted otherwise’ (1981, p27). A key feature of agent regret is that it extends to the unintended consequences of one’s actions, such as the case in which a safety-conscious lorry driver runs over a young girl who runs out into the road (Williams, 1981, p28).

With the idea of agent regret in mind, let us now turn to a prominent case put forward by Bittner (1992) as to why regret may be thought of as unreasonable. According to Bittner, experiencing regret simply adds additional suffering into the world on account of the painful reflection on a previous wrongdoing by an agent. In this sense, regret makes one experience additional pain which does not (by itself) serve to rectify previous wrongdoing, giving us a prima facie reason to avoid it. Put crudely, the idea reflects the old adage ‘don’t cry over spilt milk’. Bittner also claims that feelings of regret could impede rational decision-making by clouding our judgement (p267), however, such claims are of an empirical nature and are presented without empirical backing, placing them outside the scope of this discussion.

Bittner’s argument unduly restricts itself to the hedonic aspects of regret whilst ignoring its normative functions. Bittner purposefully excludes instances of agent regret that pertain to unintended consequences (such as the lorry driver case) from his argument. This reduces regret to ‘a painful feeling about something we did which we think was bad’ (1992, p262), where the notion of having done something pertains only to intentional actions. However, as previously illustrated, we can feel regret towards outcomes that we did not intend. These consequences can also carry normative weight (in the case of running over a young girl) that are still causally brought about by the agent.

Intuitively, it would seem that the refusal of the lorry driver to regret the consequences of his actions would be inauthentic. Due to the moral significance of such an event, merely acknowledging one’s causal role in bringing it about (perhaps by saying ‘It was me who ran her over, and I acknowledge the tragedy of it.’) appears to miss an important aspect of the situation. Price (2020) highlights that such intuitions can be supported by the fact that regret can motivate the agent to remedy the situation by giving compensation or by consoling those affected. Even if nothing can presently be done to remedy the situation, Price argues regret keeps the value of its object in the mind of the agent, aiding future remedial action. 

Failure to regret such instances appears to externalise the costs (both moral and potentially material) of the consequences of our actions to the ‘insurance fund’ (Williams, 1981, p29) whilst trying to maintain our character as moral agents whose actions have normative weight. I shall call this function of regret (that assimilates even unintended consequences into our picture of moral agency) ownership, as it enables us to ‘own’ the normative consequences of our actions as moral agents instead of jettisoning them.

Bittner criticises this conception of ownership by claiming that there is no moral obligation to retain regretted outcomes as part of your identity as a moral agent (1992, p270). This comes from the premise that the value of ownership stems from recognising the fact that we did indeed cause a certain event. It is by being actively aware of the true reality of our actions that we attach value to ownership. However, as the reality of one’s moral identity is up to the agent, no truth can be reached by choosing to take ownership of an event via regret as opposed to jettisoning it by not regretting it. The fact that the lorry driver is able to choose between two options, A (regret) and B (not regretting), does not (by itself) make the choice of A in line with our true identity as moral agents, as we are free to choose either A or B. Therefore, Bittner’s objection holds that ownershiphas no value in shedding light onto the actuality of our status as moral agents- making regret (viewed as pain and ownership) unreasonable.

I contend that this issue of why it is morally costly to forgo ownership of certain regrettable actions and events is answerable only when we examine another core function of regret that ties it to moral agency, its communicative function. The interaction between these two functions will allow us to overcome Bittner’s objection.

Regret enables communication on both an interpersonal and first personal level. When one experiences regret, we can say that the emotion is directed at an undesirable outcome of one’s actions and aims at invoking a certain response in both the regretter and relevant third parties. Matheson (2017) uses the example of Alan knocking over an acquaintance’s (Bill’s) vase with an involuntary arm spasm to illustrate the point. When Alan experiences regret, he feels a pain directed at the knocking over of the vase which, when implemented in his speech or body language, promotes an uptake of empathy or forgiveness in Bill. This is not to say that communication is impossible without the pains of regret, but rather that it helps to make our communications with others appear more genuine and persuasive.

Regret also allows us to communicate with ourselves, due to the ‘epistemic gap’ we have regarding our own minds (Matheson, 2017, p676). When Alan knocks over the vase and feels regret towards the destruction of Bill’s possessions this may come as a surprise to him, as he was neither particularly close to Bill nor did he intentionally destroy the vase. In such a case, the feeling of regret could communicate that Alan cares more about Bill than he previously thought. Perhaps Alan viewed himself as a heartless misanthrope, however, this feeling of regret could communicate that he does in fact care about others. In such cases, we can say that regret allows one’s perceived self to better approximate one’s actual self. This self-knowledge appears to be prima facie valuable (Schroeder, 2019) whilst also allowing us to make decisions that are more in line with our actual preferences and desires (Matheson, 2017, p677). This is due to the fact that if one has a fuller grasp on their actual desires, day-to-day decisions can be better made that fulfil said desires.

Now that I have outlined both the ownership and communication functions of regret, we can now answer Bittner’s objection to why retaining regrettable outcomes as part of our moral agency is valuable. The first-person communication of regret is inherently self-regarding, meaning that it is directed at oneself and conveys self-knowledge. I contend that the nature of the self in question is one’s identity as a moral agent, a person concerned with the goodness and badness of outcomes that they cause. Therefore, for the communicative function of regret to work, one must first recognise the object of regret (such as the knocking over of the vase) as their own. Failure to do so renders the knowledge uniquely conveyed by regret irrelevant either to their future decision-making or in helping them discover their actual selves as it is not evaluated as pertaining to them.

Therefore, it is not the fact that ownership conveys the truth about our status as moral agents in general, but rather that it enables us to discover what type of moral agent we are by enabling first personal communication. 

The communicative function of regret also enables us to critique Bittner’s hedonic view of regret. According to Bittner, the pain of regret is a wholly negative experience which is best to be avoided. This condemnation of pain can equally extend itself to all negative emotions, with grief acting as a prime example. Just like in instances of regret, Bittner argues that ‘there is nothing recommendable or praiseworthy in grieving’ (1992, p273) and that it too is best avoided. However, one can find value in negative emotions if we consider that we attach value not only to our hedonic preferences for pleasure and pain but also to a set of second-order preferences about what type of person we are.

I take it as uncontroversial that most people wish to regard themselves as caring and compassionate for those they are close to. It can be said that grieving the loss of someone is a sign not only that they were close to you, but also that you are compassionate towards them as you are willing to undergo present pain in their memory. Even if this connection between compassion and grief is not built on strict logical implication, it is (at least in western cultures) culturally expected that one grieves the death of a loved one. Therefore, when someone feels the pain of grief it communicates to them that they are acting in accordance with the type of person they want to be, a compassionate person. 

Similarly, in the vase case not only does regret communicate that Alan cares more about Bill than he once thought, but it also communicates that Alan’s actual self is in line with a valued conception of the type of person he wants to be. Alan has a perceived notion of a caring person as one who would regret damaging someone’s vase (albeit accidentally), and through actually experiencing regret Alan embodies this notion, bringing him satisfaction. Negative emotions, including grief and regret, can now not only be seen as valuable but also instrumentally reasonable due to this communicative role.

The final role of regret stems both from its ownership and communication functions in the form of authorship.Schroeder (2019) argues that in addition to caring about the expected utilities of our choices in our day-to-day lives (or what we ought to do), we also care about how such events unfold in our lives to form a narrative. A key part of this narrative is that we are authors of it, with the events that make up our life being under our control. Of particular relevance to regret are major life decisions such as choosing a career, moving to a new country, or deciding to marry. These decisions allow us to ‘become authors of our own lives’ (Schroeder, 2019, p148) through their expected and seemingly persistent significance as to the type of person we are. These two aspects of a narrative conception of one’s life draw upon both ownership (as we claim our lives as our own) and first personal communication, as major life decisions allow us to become the persons we wish to be.

Consider the case where you are choosing whether to move to Greece to start a new life[1]. You have received a job offer that pays more than your current job and you know the locals well thanks to your many trips to the country in the past. Despite these positives, when you deliberate over whether to move or not you begin to think about what would happen if things didn’t turn out as planned, you might feel culturally isolated and find it difficult to see your family regularly. You then decide to stay in the UK as you feel that you would regret moving if things turned out badly, whereas if the situation in the UK worsens you could always consider moving again.

In the above example, you are anticipating what Schroeder calls narrative regret (2019, p151), where one feels as if they are ‘alienated from their own story about their own life’. If things don’t go as well as expected in Greece, you would be more unsatisfied with the narrative of your life than if you stayed in the UK. It is through narrative regret that you take up the position as an author of your life, exercising your capacities as a moral agent. 

Despite Bittner’s claim that regret is unnecessary in such cases as one could rationally deliberate what one ought to do by considering the costs and benefits, narrative regret is still of use as it does not enter the deliberative process as simply another ‘negative’. Rather, narrative regret acts to clarify the decisions one makes by framing the decision in terms of an overarching narrative as opposed to an isolated choice. Hence regret is not concerned with what one ought to do per se, but rather with which narrative you would like to integrate yourself with (Schroeder, 2019). Schroeder fits this aspect of regret into its phenomenology, as it feels as if it clarifies one’s choice once anticipated, rather than it being accounted for when you first consider whether or not to move to Greece.

I have argued that regret is reasonable when the functions of ownership, communication, and authorship are invoked, however, this does not mean that every instance of regret is reasonable. The more apparent cases of unreasonable regret occur when the object of regret is a minor non-ethical choice in which we have transparent first personal knowledge. A clear example of such a choice is picking between two desserts (a pie and some sorbet) which satisfy your preferences in qualitatively different but comparably satisfactory ways (Chang, 2017).

In such a case I shall assume that the decision is minor and will have no anticipated or actual meaningful effect on your life’s narrative, excluding the authorship function. Such a case is also of little to no moral significance as it is entirely self-regarding and bears no weight on the well-being of others, making ownershipof the decision less important. We can further assume that you have eaten plenty of pies and servings of sorbet before, so you have transparent and complete information as to how they satisfy your preferences for desserts, and what those preferences are- this excludes the role of communication. In such a case, regretting the choice of a particular dessert does appear to be unreasonable as none of the positive roles of regret are invoked.

More interestingly, we could say that one acts unreasonably when one regrets pathologically in a way that is disproportionate to the object of regret. Just as one would be unreasonable for continuously weeping over catching a friend’s jacket when closing the door, one regrets unreasonably once the functions of regret have been carried out. Ownership and authorship appear to occur as soon as regret is felt (or anticipated) by the agent, as one affirms that the thing that they are regretting is part of their identity as a moral agent in their life’s narrative. First-person communication also seems to come into force once regret is felt as the fact that one feels that way indicates an underlying concern for the object of regret (such as Alans’s concern for Bill and his vase). It is communication with others that appears to be the limiting factor, for once it is made clear to others how we feel, then continuing to regret appears to be futile and self-serving (Matheson, 2017).

To conclude, regret can be said to be reasonable in cases where we take the object of regret on board as a moral agent (ownership), where regret enables both interpersonal or first personal communication, or when helping us to decide which narrative we would like our lives to form (authorship). This analysis of regret enables us to critique Bittner’s view as restrictive by focusing only on the hedonic aspects of regret as a factor that makes us decide what we ought to do. My account also helps to explain why Bittner’s argument leads to unintuitive conclusions about negative emotions in general such as grief. Finally, this analysis does not render all instances of regret as reasonable, such as in the intuitive cases of regretting minor non-ethical decisions with transparent knowledge, or in cases of pathological regret.

Bittner, R. (1992) Is it reasonable to regret things one did? The Journal of Philosophy, 89 (5): 262-273.

Chang, R. (2017) Hard choices. Journal of the American Philosophical Association, 3 (1): 1-21.

Matheson, B. (2017) More Than A Feeling: The Communicative Function of Regret. International Journal of Philosophical Studies, 25 (5): 664-681.

Price, C. (2020) The Many Flavours of Regret. The Monist, 103 (2): 147-162.

Schroeder, M. (2019) Why You’ll Regret Not Reading This Paper. Royal Institute of Philosophy Supplement, 85 135-156.

Williams, B. (1981) Moral Luck: Philosophical Papers 1973–1980. Cambridge: Cambridge University Press.


[1] This example is based on those presented by Schroeder (2019)

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s