Quasi-Relativism

*[NOTE: In order to correctly view characters in this post, your browser must be compatible with HTML4.0. One such browser is available free here.]

Modeling moral truth in S4 and S5

In chapter 6 of Simon Blackburn’s Spreading the Word [Simon Blackburn, Spreading the Word, Oxford UP: 1984.] he addresses various charges of `relativism’ and `subjectivism’ against his account. In doing so he introduces the attitudes associated with obligation and permission, and their relation to rival moral systems. He claims that if two moral systems, s and s’, cannot be improved, such systems are permissible for each other; call this Blackburn’s improvement criterion. So, let s be my moral system, s’ be another system, and A be some action. If we read □ and ◊ as unary operators corresponding to `it is obligatory that’ and `it is permissible that’, then we have the following:

1. □¬A, s [Premise]
2. ◊A, s’ [Premise]
3. ◊◊A, s [By 2, and the Improvement Criterion]
4. ◊◊A ⊃ ◊A [Reduction Law]
5. ◊A, s [3,4]
6. □¬A ≡ ¬◊A [By Definition]
7. ¬◊A, s [1,6]
8. ¬◊A ∧ ◊A, s [5, 7 Contradiction!]

Premise 4, the reduction law, will only be valid in modal logic S4 or stronger; the access-relation between systems requires transitivity for the reduction to hold. The conclusion requires the agent holding s to give up the commitment to □¬A. Would the reverse be true? Namely, must the person holding s’ give up ◊A?

1. □¬A, s [Premise]
2. ◊A, s’ [Premise]
3. ◊□¬A, s’ [By 1 and the Improvement Criterion]
4. ◊□¬A ⊃ □¬A [Reduction Law]
5. □¬A, s’ [3,4]
6. □¬A ≡ ¬◊A [By Definition]
7. ¬◊A, s’ [5,6]
8. ¬◊A ∧ ◊A, s’ [2, 7 Contradiction!]

The reduction law here is `◊□¬A ⊃ □¬A’. But this will only be valid in S5 where the access-relation between systems is both transitive and symmetric. But this reduction law when taken in its informal reading is much less intuitive. Being permitted to permit A seems to imply permitting A intuitively. However, it is much less plausible to think that being permitted to obligate A implies A is actually obligatory. And this fact might take one to regard S5 as not accurately modeling moral talk.

One reason for this is the nature of the access-relation. It makes sense to think of it as accessing different moral systems. In S5, s accesses s’ and vice versa; whereas in S4 s accesses s’, but s’ does not access s. Informally, my moral system, when it accesses another unimprovable system, must permit those things which it permits (thus transitivity). However, my moral system does not obligate what another’s obligates; another’s system does not access mine in return (and thus is not symmetric). This account is intuitively correct, I think, since my being aware of another’s system does not force me to adopt that system in total; it simply forces me to permit that system. But permitting it does not force me to obligate what it obligates.

The Rational Nazi and S4

We have seen that S5 is too strong. But is S4 too strong as well? Suppose there is some agent, the rational Nazi, who holds that people should be killed for various ethnic reasons. Moreover, let us stipulate that this view is a system, s*, that is supremely rational, i.e. one for which there can be no improvement. Clearly, in s*, `◊K’ is true, when K is read “killing for ethnic reasons”. But if this is so, then it follows that when I am aware of the rational Nazi, `◊◊K’ is true in s, and thus so is `◊K’. But surely, I want to reject that it is permissible to kill someone for purely ethnic reasons, even if s* is maximally consistent. But how can I avoid this conclusion? (Note: It is unclear to me whether Blackburn does want to reject this conclusion. Perhaps he will want to bite the bullet and maintain that there simply are no obligations, since any maximally coherent moral system immediately counts as a defeater to them. For me, this isn’t a desirable result.)

I think Blackburn has two options here. He can reject S4 as too strong, and thereby avoid the reduction law `◊◊A ⊃ ◊A’. This looks to be promising. After all, even if we are prepared to admit that it is permissible (by my lights) for the rational Nazi to permit killing for ethnic reasons (by his lights), we will want to block the inference to the permissibility of killing for ethnic reasons (by my lights). But Blackburn, I think, wouldn’t be attracted to this route. He seems to think the reduction law in S4 is intuitive. [See especially p.202]

Blackburn seems more inclined to reject the claim s* is maximally coherent, or unimprovable. Thus, on his view, the rational Nazi is impossible. He writes, “Fortunately, all this is ridiculously beside the point. [...] [O]ur natures and desires, needs and pleasures, constrain much of what we can admire and commend, tolerate and work for. There are not so many livable, unfragmented, developed, consistent, and coherent systems of attitude.”(p.197) The trouble here, of course, is that we often do seem to come up against those with differing but consistent moral views. Blackburn may wish to maintain that, although some systems are different, they never are so radically different as to permit such atrocities as genocide. For him, perhaps our natures and desires constrain moral systems to such an extent. This may be the case, but, as I far I know, he has not argued for this claim or defended it . The charge of relativism still stands until he does so.

~ by acotnoir on August 28, 2006.