The Paradox of Supererogation: Can we choose to be kind?

Supererogatory acts are those acts which go ‘above and beyond’ the call of duty: freely chosen acts which are morally better than our obligations. The existence of supererogatory acts forms one point of an inconsistent triad, which – along with moral continence and the ‘very weak classical view’ (VWCV) – is known as the paradox of supererogation. Donald Davidson (2001) provides an influential solution to this paradox, by rejecting supererogation. However, due to the intuitive attraction of supererogation, I provide a more compelling, novel solution by denying the existence of positive moral obligations, thus only requiring a modification to moral continence.

Note that the VWCV is not discussed in detail, both for a concise analysis, and because the questions of how the paradox should be resolved can be answered by solving the tension between supererogation and moral continence alone. As such, this judgement depends both on internal consistency, and its convincingness as an account of our moral intuition.

As mentioned, the paradox of supererogation comprises three terms: supererogation, the VWCV, and moral continence. Before we can solve the paradox, we must understand the relationship between these three terms.

First, supererogation holds that there are acts which are morally better than our obligated acts. James Dreier (2009) gives the example of riding a bus, when a lady’s hat blows out the window; in getting off, running after the hat, and returning it to the woman, you are going above what is required of you (2009:147). This is an incredibly intuitive phenomenon; examples of people’s unexpected kindness are easily recognisable – Daniel Muñoz (2020) presents doing favours, or even rescuing drowning children (2020:699). Here, despite the requirements for ethical action, there is clearly a range of supererogatory acts available.

Second, the VWCV holds that moral goodness and moral reasons move in tandem, such that if X is morally better than Y, then there are more moral reasons to do X than Y. Third, moral continence holds that we are morally obligated to do that which we have the most moral reason to do – combined with the VWCW, we are obligated to always do the morally best act (2009:148). Yet, it cannot be both true that we are obligated to do the best act, and that there are non-obligated acts which are morally better.

Davidson argues that the paradox should be resolved by rejecting supererogation, as moral continence remains the stronger principle. To prove this, he outlines moral incontinence – where an agent intentionally acts, despite knowing “that another course of action is better” and has more reasons (2001:21). For example, Davidson presents a man settling into bed, realising that he has forgotten to brush his teeth. Whilst judging that the cost of not brushing once is offset by the benefit of being well-rested, he thinks: “my feeling that I ought to brush my teeth is too strong for me” (2001:30). Despite having more reasons to stay in bed, the man intentionally gets up. As such, Davidson argues that – though perhaps for different reasons – any agent who is incontinent is irrational, and “cannot understand himself” (2001:42). From this, Davidson concludes that, if we are to be rational, moral agents, we ought to act by the ‘principle of continence’: to “perform the action judged best on the basis of all available relevant [moral] reasons” (2001:41).

Davidson’s argument seems initially compelling – choosing a sub-optimal action is intuitively irrational, leaving us with unsatisfying answers about an agent’s reasoning. Yet it is unclear why this rationality must be morally obligated, especially since it leads us to overly-demanding requirements for ethical action: to always do the most moral act, with no choice from a selection.

Anticipating this objection, Davidson takes a case of having the most reasons to do X yet being obligated to do Y. Here, if you are not obligated to do that which you have the most reason, then any subsequent explanation of why you have an obligation to do Y simply provides further reasons for Y over X (2001:42). As such, we are led back into a conclusion that the obligated act and the act with the most compelling reasons are one and the same. This defence indeed strengthens Davidson’s argument – under this understanding, the link between our moral reasons and obligations is well-explained. However, it remains unconvincing that that any act with fewer than the most moral reasons (and thus, less than highest moral goodness) is just as morally unacceptable as the act with the least. That is, the implication that not retrieving the lady’s hat and murdering her remain equally unacceptable for Davidson, gives us strong reason to question his understanding of obligation.

Alternatively, I argue that the paradox should be resolved by modifying moral continence and loosening the demands of obligation. Specifically, there are no positive moral obligations, only negative obligations; that is, there are no demands from morality to do certain acts, but instead, to merely abstain from them. This is intuitive: well-recognised moral obligations regarding murder, theft, rape, etc., all remain prohibitions from causing unjust harm to others – yet obligations which demand certain acts are far less common (and are dealt with shortly).

However, it may be fairly levelled against this argument that – since positive obligations only relate to morally good acts – it worryingly creates limits responsibility towards moral goodness. Yet, I argue that this is simply a more accurate depiction of how we value goodness. In Dreier’s example, would we describe the man as doing ‘simply what was expected of him’, or praise him because he went above and beyond his obligation? I argue the latter; of specific acts which are obligated, such as a parent feeding their child, it seems strange to praise the goodness of the act – since this is simply what is expected. Likewise, of acts which are morally good, I argue that our praise necessarily excludes obligation. John Taurek (1977) convincingly makes a similar argument, in which an agent is not obligated to sacrifice their life to save five others; it would certainly be ‘nice’ to be altruistic, but this description is only possible since the agent has the free choice to save himself (1977:299). As such, whilst non-obligated good acts exist, since no obligated acts are described as morally praiseworthy – and no genuinely praiseworthy acts are also obligated – more fundamentally, all morally good acts are supererogatory. Indeed, if this was not sufficiently convincing, the very fact that so many people do not meet Davidson’s minimum requirements for action – yet maintain their status as moral people – gives us another strong reason to adopt my account of obligations instead.

On the other hand, upon hearing this counter, Davidson may still reject the theory, on account of its worryingly limited responsibility to action in general. For instance, he may argue that, clearly, a parent has a positive moral obligation to protect and care for their child, or an employer to actively ensure a fair and safe working environment to their employees. This is initially very compelling – there are countless instances of there being special duties to act in specific ways, which are enforced by positive moral obligations.

Whilst these cases indeed threaten my argument, Taurek’s arguments can be used to internalise these exceptions. Taurek explains that, whilst altruism is not ordinarily required, there are indeed cases of “a special obligation…arising out of a contract”, such as when “a doctor has a special obligation to his…patients” (1977:296). Under this interpretation, these obligations are ‘secondary’ positive obligations – only created when parties freely enter into a contract together. These are secondary since obligation is created from the ‘primary’ negative obligation to ‘not break your promises’. This is highly convincing; just as a parent enters into an implicit contract to care for their child when freely choosing to conceive – and an employer with their employees when freely choosing to hire – so it is the case that any proceeding obligations only have moral force because their violation would violate the obligation to not break one’s contracts.

Given that primary obligations are only negative, we can solve the paradox of supererogation in a more compelling manner. Specifically, we can formulate a ‘negative principle of continence’ – to abstain from the acts which you have more moral reasons to not do. This does not violate the VWCV, since moral goodness and moral reasons moving in tandem also implies that the morally worst acts have the fewest moral reasons to do them, and thus the most moral reasons not to do them. Unlike Davidson, however, we are not comparing within a selection of actions, but instead, within a single action. For instance, in considering whether I will help a cat out of a tree, I consider each of my possible acts – shooting the cat, ignoring it, calling the fire service, climbing up the tree myself, etc. In judging whether each has more reasons to not act than to act, I am left with a revised selection of acts. Since the revised selection is composed of acts not prohibited by negative obligation, obligation is no further applied – thus, if I freely choose to save the cat, this act is supererogatory. This formulation is more compelling, since it solves the paradox of supererogation – by holding supererogation, the VWCV, and the negative principle of continence as all true – whilst avoiding the overly-demanding and unrealistic description of obligation found in Davidson’s solution.

In conclusion, whilst Davidson’s solution to the paradox of supererogation comes close to providing a satisfying account, it leaves too many serious concerns unresolved. Specifically, its overly-demanding requirements, difficulty differentiating permissibility observed in real life, and rejection of the existence of supererogatory acts render it unconvincing. In contrast, by providing a more compelling account – albeit admittedly radical – understanding of obligations, my theory of negative moral continence avoids the aforementioned issues, whilst still solving the paradox of supererogation.

Bibliography:

• Davidson, D. (2001), “How is Weakness of the Will Possible?” in “Essays on Actions and Events”. Oxford University Press, pp. 21-42.
• Dreier, J. (2009), “Why Ethical Satisficing Makes Sense and Rational Satisficing Doesn’t” in “Satisficing and Maximizing”, edited by Michael Byron. Cambridge University Press, pp. 131-154.
• Muñoz, D. (2020), “Three paradoxes of supererogation”. Noûs, Vol. 55, No. 3, pp. 699-716.
• Taurek, J. (1977), “Should the Numbers Count?”. Philosophy and Public Affairs, Vol. 6, No. 4, pp. 293-316.