Two-level Deontology?

A neglected region of logical space

It’s standard for consequentialists to distinguish “criteria of rightness” (or what considerations, in theory, have fundamental normative force) from “decision procedures” (or what real-world practical norms are worth endorsing and inculcating). I’d be interested to learn of non-consequentialist discussions of this distinction; my sense is that it doesn’t tend to be so much on their radar. But I think it’s worth considering.

After all, deontological views still owe us some answer to the telic question of what (act-inclusive) outcomes we should hope to see realized. And there’s no empirical guarantee that naive instrumentalist pursuit of their (even distinctively deontological) moral ends will be the most reliable means of securing them. For example, suppose that (i) we have duties of respect not to exploit beings above a certain level of cognitive development (only), and yet (ii) anthropocentric and self-interested biases will predictably lead people to underestimate which non-human animals thus qualify. One might then support veganism as a practical norm to more reliably secure the relevant duties, even while believing that there are actually some animals which it would (strictly speaking) be objectively permissible to “exploit” (or treat in a way that would constitute exploitation were they a relevantly different sort of being).

If you don’t like that example, we can imagine more stark cases involving weirdly anti-reliable agents. If people routinely achieved the very opposite of what they were superficially “aiming” at, then deontologists should presumably want these people to “aim” at violating their most important duties. (If the clumsy agents did this with an eye to ultimately securing the relevant duties, it seems they could even qualify as virtuous and praise-worthy while doing so.)

So, for the same basic reasons that consequentialists adopt two-level views, it seems that deontologists should do likewise. Their two levels may more commonly coincide in real life (not a coincidence, since I think commonsense deontology just is the theoretical reification of plausibly good practical norms). But the possible divergence in hypothetical cases suffices to establish the conceptual distinction between good practical norms and objective normative reasons, even for deontologists.

Once you grant the distinction, that naturally opens up my master argument for consequentialism: considered as rival normative explanations of our shared practical norms (supporting rights, opposing pushing people in front of trolleys), the consequentialist account has significant advantages, and no such obvious disadvantages. That’s why I think it ends up being the most intuitive view, in reflective equilibrium. (Part IV of Beyond Right and Wrong will develop this argument at greater length.)