Thomas Hurka and Esther Shubert (henceforth, “the authors”) offer an interesting account of permissions. The authors’ starting point is that it can be a “primitive truth that you are permitted [to pursue your own happiness].” (6) They then construct a system of ratios for a “band of permissions” to favor one’s own happiness (and freely sacrifice one’s own happiness). This paper argues that the authors’ model has an issue with self-respect and enforceable duties. By straying away from the “reason-based” accounts, the band of permissions account has tensions between moral agents who act permissibly; in other words, it is untenable within a moral community. This issue points to an underlying problem with the authors’ starting point of permissions rather than duties.

The authors begin by outlining the deficiencies in the “reason-based” accounts of agent-favoring permissions. The authors say that we generally have permissions to care more about ourselves, and the “reason-based” account says that these permissions come from a clash between “moral reasons to promote the good of all people impartially” (2) and “prudential reasons to care disproportionately about your own [good].” (2) Derek Parfit’s nuanced “reason-based” account suggests it is sometimes obvious which choice to take when impartial and partial concerns clash – for instance, “if you can give either one unit of happiness to yourself or a thousand units to another, your impersonal reason outweighs your personal one and you should prefer the thousand.” (3) However, it is not clear when the clash is smaller, like “one [unit of happiness] for yourself and six, four, or three for another.” (4) These uncertain areas, according to Parfit, are the basis for agent-favoring permissions.

The authors push back on this sort of “reason-based” account. They suggest that this approach is “misguided” (5) because the permission is derived from “more basic normative factors that aren’t themselves permissions but count positively in favour of an act.” (5) The “reason-based” accounts begin with duties, what we “ought” to do (ceteris paribus), and try to outline what we are permitted (or “may”) do.  The authors set their primitive truths as permissions (ceteris paribus) rather than constraints or duties. Parfit’s “reason-based” account starts with clashing duties and has areas of permissions where the clashing duties cannot be outweighed one way or the other. The authors find this “recherché” view counterintuitive.

Let us take a closer look at this critique. It seems right to say that Parfit’s account has the wrong rationale. It gives a sort of explanatory account of permissions which derives from some “underivative truths” (6) about duties; the authors, however, give a normative explanation which gives a justification of permissions from the outset. The aim is to justify permissions without appealing to duties (or uncertainty of weighing some duties with another), although these permissions can be “weighed against the impartial duty and in some cases will lose to it.” (5) This approach appears to capture our common sense morality and gives us a working “band of permissions.”

The idea of a “band of permissions” is supposed to demarcate just how much is permitted for “agent-favoring choices” and “agent-sacrificing choices.” (13) Reduced to units of happiness, the authors propose a ratio scheme to track agent-favoring and agent-sacrificing choices. For example, an agent-favoring ratio would be choosing one unit of happiness for yourself instead of choosing five units of happiness for another person. An interesting feature of the authors’ view is that it has room for agent-sacrificing permissions. The importance of this feature is evident when considering choices involving intimates – for instance, a husband might want to choose one unit of happiness for his wife instead of five units of happiness for himself. In essence, there is a ratio scale and a “band of permissions” designating which range of ratios is permitted. There are other various interesting aspects of this “band” theory, but only a grasp of the general idea is needed for the rest of this paper.

Let us return to the original motivation for this band theory, namely as a better alternative to “reason-based” accounts. There seems to be an issue that arises with the band theory when it is practiced within a community of other moral agents. Specifically, tensions arise when moral agents have differing bands of permission. Here I explore two instances of such tensions: areas of self-respect and areas of enforcement.

Self-respect is closely tied with agent-sacrificing permissions. In brief, self-respect is thought of as a duty to the self that encompasses things like integrity and dignity. To illustrate, a husband who constantly aims at pleasing his wife over any of his own aims to the point of undermining his own projects, needs, and general happiness can be said to be violating self-respect – for instance, (say) when he is not busy earning money to shower his wife with gifts, he spends the rest of his time washing his wife’s feet with his tears. Where does this fit on the “band” theory? It would seem that a case of self-disrespect is beyond the band of permission for agent-sacrificing. But suppose the husband believes that his band of agent-sacrificing permission extends much longer and that his actions are within the band of permissions; then, it appears his self-disrespect is merely a case of relatively extreme self-sacrifice. Luckily, the authors provide stipulations for self-respecting agent-sacrificing. They say that one must “have the following beliefs: that you’re only permitted to prefer another’s lesser happiness but have no duty to do so; that you’re also permitted to prefer your own lesser happiness, and even to do so to a greater extend; and that your mix of permissions and duties is exactly the same as everyone else’s.” (19) The final clause is of particular importance, but we will return to it after flushing out a similar tension with the authors’ model.

Now looking at the enforcement of duties, it seems that one might sometimes have a duty to enforce the duties of other moral agents. For example, you might have a duty to enforce the duty of not killing innocent people, say, if you had sniper scoped on me and I was on a murderous rampage at the local daycare. On the “band” picture, some have suggested that there are certain thresholds for enforceable duties – for instance, if I chose to favor one unit of happiness instead of a million units for somebody else.[1] With this in mind, we ask the following question: is the threshold identical for everybody? If it is not, we run into some issues. What would be over the threshold for enforcement for one person could plausibly be under the threshold for enforcement for another. It is unclear who is wrong in this picture; there is some disharmony with taking such a view. There is moreover a less ambitious point to be made. Even without talk of thresholds, what is one person’s duty can be considered a permission on another person’s account. There appears to be no unity within this “band” picture.

Both of these areas of tension, self-respect and enforced duties, point at the issue with having band of permission that is subjectively derived. There is no objectively correct band of permission to compare and judge moral choices with. Let us now return to the authors’ suggestion that the “your mix of permissions and duties is exactly the same as everyone else’s.” (19) Perhaps this means that human psychology is such that we have a univocal moral intuition so all of our bands and ratios are identical. This would solve tension; however, this seems rather ad hoc and unfeasible. It would be fair to say that there is some general form that the band of permissions takes, but tensions would arise unless they were completely identical – really, the devil is in the details. Two self-proclaimed morally ideal agents would likely question at least some of each other’s choices. By and large, permissions seem to be fairly controversial, or at least not univocal.

Perhaps the authors want to suggest that there is some band of permission that is objectively correct. This approach however seems to be a case of having your cake and eating it too. The authors jettisoned the “reason-based” account which may have this feature of objectivity. When duties are primitive, they can be appealed to as an objective criterion to analyze permissions. Take Parfit’s view for example: moral agents would not have different permissions because they have the same criterion (i.e. duties that cannot be weighed). The authors rightly point out that even Parfit’s view is not completely satisfying with respect to resolving conflicting views on permissions. That is, perhaps Parfit’s criterion for permission is equally vague since epistemic reasons are not as strong as “there being, metaphysically, completely determinate truths…” (4) The authors’ picture does not include a robust band of permission to model after or use to adjudicate what is and what is not within the band of permission.

This paper cannot offer any substantive revisions to the authors’ model, but there might be some grounds for questioning the authors’ shift of starting with permissions as a primitive truth. One peculiar aspect is how the authors’ define the nature of permission. The authors suggest that “the concepts of ought and permissions are interdefinable.” (6) It seems right to say that permissions and duties (or “oughts”) should make reference to each other if they are to be coherently defined, but the authors’ conception of permission seems too ambitious. The authors say that the strength of a permission is “its tendency to determine a normative outcome,” (7) but this makes the domain of permissions too big. Such normative outcomes can involve morally trivial cases, such as choosing the color of one’s shoes. It seems morally trivial cases do not enter at all into the calculus of the band of permissions – they are irrelevant and do not require further analysis in this manner. If duties and permissions are “interdefinable,” then permissions should be limited to morally significant choices.

As such, a slightly more refined view of permissions undercuts some of the motivation for thinking permissions should be taken as a primitive truth. On the authors’ view, a “reason-based” account seems to ground some cases, like the choice of shoe color, in a convoluted calculus of duties and derived permissions. Yet, understanding that these morally trivial cases do not enter the picture, the “reason-based” accounts seem at least more elegant than what the authors’ see it to be. Nevertheless, this new outlook carries the assumption that there is a different class of morally trivial choices which do not enter the discussion of “ought and permissions,” but this seems relatively uncontentious.

The “band” theory seems to have some issues that the “reason-based” accounts do not. The main issue that is that it creates tensions between moral agents who act permissibly, and such tensions include self-respect and enforced duties. The fundamental issue is that jettisoning the “reason-based” account also took away any objective criterion to compare bands of permission. This might be a reason to reconsider taking permissions as a foundational primitive truth.

Source

Hurka, Thomas and Schubert, Esther. ‘Permissions to Do Less Than the Best: A Moving Band’, Oxford Studies in Normative Ethics 2 (2012): 1-27.

---

[1] This is not too convincing to me. Would we have a duty to make sure that all the billionaires in the world donate to charity?