Thursday, December 01, 2011

Scalar Consequentialism and Constructed Permissibility

I take Consequentialism to suggest a fundamentally scalar picture. The most fundamental assessment of actions simply ranks them on a scale of better to worse, indicating our having more or less reason to perform them. That's what centrally matters.

But we may also be interested in other moral questions, such as whether we would be blameworthy for performing some act. Consequentialists traditionally haven't been much interested in questions of blameworthiness (as distinct from, say, whether it would promote utility to express blame in some circumstances), but I think there are real normative questions here, besides those that consequentialism addresses. For example, there are rational norms governing emotions and reactive attitudes, which we may reasonably theorize about. So we may ask whether certain negative emotional responses towards others are warranted, in light of their actions. This is to ask whether they are blame-worthy.

I think the best theory of blameworthiness is some kind of quality of will account, according to which people are praiseworthy or blameworthy, in performing some action φ, to the extent that their φ-ing manifests a good or bad quality of will (respectively). Here "bad" is to be understood as insufficiently good -- so acting in a way that isn't positively malicious, but demonstrates a lack of adequate concern for others, still qualifies as "blameworthy" on this account.

This distinction between adequate and inadequate concern introduces a binary element into our moral philosophy. [This comes with distinctive problems -- what determines exactly where the boundary is drawn? -- but I won't get into that here.] This can then be used to construct a derivative notion of 'rightness' or permissibility that could be of practical interest. For example...

(1) Impermissibility as hypothetical blameworthiness: There seems to be an intimate connection between wrong action and blameworthiness. But they're clearly not identical: Sometimes ignorance might excuse acting wrongly, and conversely, objectively harmless actions (e.g. voodoo) might be blameworthy if performed with ill intent. What's key to these cases of divergence is a mismatch between how things really are, and how the agent takes them to be. Quality of will (and hence blameworthiness) concerns the latter, whereas permissibility seems to be a more objective mode of assessment. But perhaps we can bridge the gap by defining 'impermissibility' in terms of acts that couldn't be blamelessly performed by competent agents who (momentarily) know all the relevant facts.

The rough idea (abstracting from the distorting effects of ignorance) is that an action is permissible if it is among the options compatible with exemplifying an "adequate" level of concern. (There are potential issues with the conditional fallacy here -- can it be avoided in this case by the proviso that the hypothetical agent is cognitively idealized only for the moment of decision?)

(2) We might also appeal to this notion of an "adequate" level of concern in order to determine a principled "effort ceiling" for Effort-Based Satisficing Consequentialism. Given a prior account of emotional norms, specifying what counts as an "adequate" level of concern (to avoid blameworthiness) in any given situation, we can then specify that the effort ceiling X is the amount of effort that an adequately concerned moral agent would be willing to expend (if necessary) in that situation. Thus explicated, EBSC amounts to the view that we're obligated to achieve the best results we can without being required to put in more effort than an adequately concerned moral agent would be willing to. And that sounds vaguely plausible, given the above noted connection between impermissibility and blameworthiness (inadequate concern).

Remaining Questions:

* Are (1) and (2) competing options? I'm hoping that (2) is what you naturally get when you supplement the general account in (1) with consequentialist norms, such that within the options compatible with one's limited degree of moral motivation, rationality requires that you choose the one with the best outcome. But it's a little slippery, so I'd welcome others' thoughts on this connection.

* I've suggested that even scalar consequentialists might be led to construct a derivative notion of permissibility, by way of an independent interest in (avoiding) blameworthiness. Do you think that this adequately captures the ordinary conception of permissibility and why it matters? (Or do you think, say, that permissibility is bedrock, and not to be analyzed in other terms? Perhaps its significance is supposed to be directly communicated through moral phenomenology? If so then I don't think I get it.)

* I'm drawing on non-consequentialist norms for emotions and reactive attitudes in order to construct this derivative notion of permissibility. Is this a problem? I'm inclined to think not, since I think that consequentialism only applies to actions (and intimately connected mental items like intentions and preferences). But others might think there's something illicit about mixing consequentialist and non-consequentialist norms in this way. If so, I'd be curious to hear the objection spelled out...

* Is it plausible that some non-optimal level of good will is "adequate" to avoid blameworthiness? How might the details of this go -- have theorists of blame said much about this? (If only perfect moral motivation is acceptable, then my proposed version of EBSC will collapse into maximizing consequentialism. Whereas if there are no objective norms governing the reactive attitudes, then it would seem we can't go beyond simple scalar consequentialism.)

Any other thoughts?


  1. “ I'm drawing on non-consequentialist norms for emotions and reactive attitudes in order to construct this derivative notion of permissibility. Is this a problem?”

    One possible worry here might be that having a non consequentialist account of the conditions where reactive attitudes are appropriate would create an implausible disconnect between the appropriateness conditions for instances of blame that are attitudes and the appropriateness conditions for instances of blame that are actions.

    It seems like in cases where resentment is appropriate, certain actions (deciding to think negative thoughts about the person, making faces at the person, etc.) are prima facie permissible. Furthermore, you might think that the features of the situation that make the blame-actions prima facie permissible are the same features of the situation that that make the resentment appropriate. For example, the fact that Bob’s act showed ill will makes resenting him appropriate and also gives you reason to perform actions that blame him.

    I do think you could account for this on a consequentialist picture if you had the right stuff in your list of goods (e.g. its intrinsically good to react negatively to those who act badly). But I’m not sure if you could maintain this connection between blame attitudes and blame actions on views where the only good to be promoted is pleasure or wellbeing.

  2. Thanks Philip, that seems like a good worry.

    Perhaps what it brings out is that consequentialists should be generally more suspicious of such "linking principles" that propose the prima facie permissibility of expressing warranted attitudes. Take belief, for example. I take it that what determines whether a belief is rationally warranted is its epistemic quality. But the epistemic quality of a belief can easily come apart from whether it's felicific to assert.

    In either case, consequentialists might suggest that we might reasonably default to the assumption that expressing warranted attitudes is prima facie felicific. Asserting warranted beliefs can help spread true beliefs, which are often useful (certain obvious exceptions aside). General practices of blaming the blameworthy may be expected to provide a useful disincentive against bad behaviour, etc. (Again, with certain exceptions.)

    So the consequentialist may be able to offer a prima facie case for these prima facie principles. But if it turned out that this consequentialist justification for them failed, then I'd be inclined to reject the linking principles on just that basis. (The proposed links seem less clearly true to me than the claims that (i) consequentialist norms are appropriate for actions, and (ii) non-consequentialist norms are appropriate for various other attitudes.)


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.