Objective and subjective modalities

Remember our list of kinds of modalities from the previous post:

• Logico-conceptual (Bachelor cannot be married)
• Epistemic (she might be at home)
• Metaphysical (water molecules are of the H2O kind)
• Natural (heavy objects must fall)
• Deontic (she must work tonight)
• Practical (we could use wood to build this plane)

The two first modalities above, epistemic and logico-conceptual, are often thought to be mind-dependent or subjective, while the two next ones, metaphysical and natural, are mind-independent or objective (as for the two last ones, let us put them aside for now). Objective modalities are often used in order to analyse explanations or causation. These are the ones typically involved in counterfactual talk (if you had got up earlier, you woudn't have missed your train), although sometimes conceptual necessities are thus expressed as well (if he were a bachelor, he wouldn't be married).

The distinction between objective and subjective modalities could be understood as the idea that the source of necessity is either “in the mind”, i.e. it is representational (our concepts, what we know, our evidence), or it is external, "in the world" (the laws of nature or essences). But is it really the case?

Take epistemic possibility. If we associate knowledge with truth, then whether something is epistemically possible or not depends on the external world, at least indirectly. Evidence is not clearly mind-dependent. In any case, it is not a priori. And the same would go for conceptual truth if one were to reject the analytic/synthetic distinction: the source of conceptual necessity would not lie entirely within us, but somehow depend on the fabric of the world (some concepts are more apt than others). However, there is still an important distinction with the two last kinds of modals: in the two first cases, the necessity is somehow relative to us: something can be epistemically possible for us at some time but become impossible at a later time or for someone else, and perhaps also conceivable by us, given our conceptual schemes, but not for our ancestors (note how in the epistemic case, progress restrains possibilities, while in the conceptual one, it opens new possibilities! I will come back to this aspect in the next article).

I believe that what is actually going on is that the target of necessity (what is deemed necessary or not) under these two modalities is representational rather than worldly, but not the source of necessity. What is epistemically or conceptually or logically possible (compatible with our reasoning rules, concepts and evidence) is a belief, a thought, or maybe more accurately the content of a belief or thought, but not what the belief is about (not the thing in the world). We often say that a state of affairs is conceivable or that for all we know, it might obtain. There is a slight difference between a state of affairs and the content of a belief (typically, belief contents are sensitive to modes of presentation in a way that states of affairs are not: “Clark Kent flies” and “Superman flies” correspond to the same state of affairs, but one can believe one without the other). Addressing all this would lead us too far, but that does not contradict the idea that with subjective modalities, the target of necessity is our attitude towards states of affairs rather than the states of affairs themselves.

So, the idea that it is thoughts that are constrained by our conceptual schemes, reasoning rules or by our evidence, and not things out there in the world, seems to make good sense. And the idea that in the case of metaphysical and natural modalities, it is worldly entities that are constrained, not thoughts, also makes sense, and all this does not depend at all on the nature of the source of these constraints (whether it is mind dependent or not), only the target. Now of course, the fact that natural constraints affect the behaviour of external objects also has an impact on how we decide to represent these objects, so it affects our representations, but the route is via epistemic modality: it is because we want our representations to be accurate that our representations (of worldly constraints) are affected. This complication should not lure us.

A criteria sometimes mentioned in order to distinguish mind-dependent and mind-independent modalities is that modals of the former kind should have their truth-value or truth-conditions preserved by substitution of co-referring terms. As a matter of logical or conceptual possibility, Clark Kent is identical to himself, but he need not be identical to Superman simply in virtue of logic nor in virtue of our concepts. Or, as a matter of epistemic necessity, Superman must be able to fly (I saw him), but not so for Clark Kent. But if we replace the conceptual necessity with metaphysical necessity, and the epistemic one with a natural one, then the substitution works. If Superman is able to fly by natural necessity (because of his physical constitution), then so is Clark Kent, since they are the same person. Is this criteria compatible with what we have just said, regarding the fact that the difference lies in the target of necessity? I think it is. The substitution concerns what is constrained by necessity, the target that is, not the source, and if the target does not change by substituting co-referential terms, as is the case for worldly targets, then it must be constrained in the same way, so modals that concern worldly targets have their truth-value preserved by substitution of co-referring terms, but if the target does change with substitution, as is the case with representational targets, then the constraints can be different in each case. And again, this does not depend on whether the source of the constraint is mind-dependent or not.

The upshot is that the division between mind-dependent and mind-independent modalities has to do with the target of the modality, and not with its source.