Precis of Van Inwagen's "Modal Epistemology": Classic Papers in Modal Epistemology Series

In “Modal Epistemology” (Phil. Studies, 1998), Peter van Inwagen defends a mitigated form of modal skepticism. In particular, he argues that while we have knowledge of many modal claims that are close to the practical concerns of everyday life, science, and even (in some cases) philosophy (e.g., Gettier possibilities), we can’t have knowledge, or even reasonable belief, regarding possibility claims remote from ordinary experience (e.g., the possibility of disembodied souls, Anselmian beings, and instances of gratuitous evil). Van Inwagen further clarifies his version of mitigated modal skepticism by means of an analogy. Just as, in prescientific times, humans confidently formed false perceptual beliefs about distant objects (e.g., the distance of the celestial bodies from the Earth) on the basis of their reliability regarding perceptual judgements about nearby objects, so in the present day philosophers erroneously form confident beliefs about “distant” possibilities on the basis of their reliability regarding modal judgements about “nearby” possibilities. 

 

Van Inwagen gives a more precise characterization of his mitigated modal skepticism by listing and illustrating the categories of modal knowledge and reasonable belief he thinks we have: 

 

(i)    Basic or non-inferential knowledge of possibilities that are close to the practical concerns of daily life (e.g., my table could’ve been two feet to the left of where it is).

(ii)  Knowledge of simple Moorean necessities and impossibilities (e.g., there cannot be liquid wine bottles; it is necessary that there is a valley between any two mountains that touch at their bases).

(iii)  Knowledge of necessary truths and necessary falsehoods of mathematics and logic

(iv) Modal knowledge based on inferences from actuality to possibility (e.g., It’s possible that there are trees.).

(v)   Knowledge of modal propositions deduced from such knowledge plus facts about the way the world is put together (e.g., water is H20, and things have their composition of necessity; therefore, water is necessarily H20).

(vi)  Reasonable belief inferred from such knowledge (e.g., it is impossible for a moon to be made of cheese).

Since Van Inwagen thinks (i)-(vi) exhaust the categories of modal knowledge and reasonable belief, we can add a closure clause to the above:

 

(CC) There are no categories of modal knowledge or reasonable belief beyond (i)-(vi).

 

Van Inwagen offers two main arguments in support of his mitigated modal skepticism. The first is that arguments containing a modal premise beyond categories (i)-(vi) are subject to parity arguments (e.g., G.E. Moore Shifts) with equal epistemic force, thereby invariably leading to dialectical stalemates. Here’s an example:

 

DUALISM:

1. It’s possible that I exist and nothing material exists.

2. Whatever is immaterial is essentially immaterial.

3. Therefore, I am not a material thing.

 

MATERIALISM:

1. It’s possible that I exist and nothing immaterial exists.

2. Whatever is material is essentially material.

3. Therefore, I am a material thing.

 

The second argument is that anyone who accepts the most sophisticated account of modal epistemology at the time of  van Inwagen’s writing – viz., that explicated and defended in Stephen Yablo’s (1993) account – should be a modal skeptic. This is because Yablo’s account requires (roughly) that one imagine possible world W such that (i) W appears possible to one, and (ii) the de-modalized target claim appears to one as true in that world. But the problem is that one cannot imagine “distant” possibilities that meet these conditions – i.e., the “worlds” which one imagines fail to generate the appearance that they are genuinely metaphysically possible. This is because (a) Yablo’s account requires that one’s imaginings have enough detail that they rule out alternative possibilities that are compatible with the falsity of the relevant modal claim, and (b) imagining distant possibilities invariably fails to rule out such alternative possibilities. To use of of van Inwagen's examples, suppose you imagined a scientist accepting the Nobel Prize to a cheering audience, who then thanks all those who helped him in his long and difficult journey to create transparent iron, and then holds up something that looks like a sheet of glass.  That wouldn’t prima facie justify the claim that there is a possible world in which there is transparent iron.  For the thought experiment is equally supportive of other claims, such as that a group of jokester scientists have gotten together to enact a burlesque of a Nobel Prize ceremony, or that a group of scientists have conspired to fool the public that they’ve created transparent iron, etc.  

Beyond this, van Inwagen seems to be gesturing to a more general point, viz., that on any adequate account of conceivability as a guide to possibility, justified belief in distant possibilities will require a level of detail in their imagining that is sufficient to rule out relevant alternative possibilities to the truth of the modal claim at issue -- something that we are unable to do. 

Finally, Van Inwagen offers three explanations for why many philosophers feel justified in their acceptance and assertions of “distant” modal claims (despite lacking adequate grounds for doing so). The first is the simple yet common fallacious slide that’s often made in philosophy:  We see that logical inconsistency is a reliable guide to metaphysical impossibility, and then fallaciously infer that therefore failure to demonstrate logical inconsistency must likewise be a reliable guide to metaphysical possibility -- i.e., they fallaciously infer from this that all propositions that don't entail a contradiction in first-order logic are metaphysically possible.  But this slide is due to a failure to see that logical possibility isn’t a real kind of possibility, or one that that maps onto metaphysical possibilities 1 to 1.  Van Inwagen illustrates this as follows.  Suppose you had a map that listed a number of landmasses as uninhabitable, none has habitable, but made no claim to completeness regarding the labeling of uninhabitable landmasses.  Suppose further that someone were to say that an island is cartographically habitable just in case it isn’t marked ‘uninhabitable’ on the map.  In such a case, although cartographical uninhabitability is a reliable guide to “metaphysical uninhabitability”, one isn't thereby justified in taking cartographical habitability as a guide to “metaphysical habitability”. Similarly, from the fact that logical impossibility is a reliable guide to metaphysical impossibility, one shouldn’t thereby infer that logical possibility is a reliable guide to metaphysical possibility.  For there is good reason to think that the set of statements that can't be shown to entail a contradiction in first-order logic isn’t coextensive with the set of metaphysical possibilities. Although not mentioned in the paper, the following simple example from Plantinga is helpful here: The statement "There is a prime number that's also a prime minister" doesn't entail a contradiction in first-order logic, but for all that, it's metaphysically impossible for there to be a prime number that's also a prime minister. 

The second is the analogical argument mentioned earlier: Just as, in prescientific times, humans confidently made false perceptual judgments about distant objects (e.g., the distance of the celestial bodies from the Earth) on the basis of their reliability regarding perceptual judgements about nearby objects, so in the present day philosophers make confident yet dubious modal judgements about “distant” possibilities on the basis of their reliability regarding modal judgements about “nearby” possibilities. 

The third is connected to the second. Philosophers' erroneous belief that our ability to reason reliably about "nearby" possibilities extends to "remote" possibilities gives rise to a kind of  institutionalized pretense regarding our competence in possibilities the latter sort. We find our philosophical peers asserting claims about distant possibilities, and they let us get away with asserting similar claims without raising an eyebrow or putting up a fight. The result is a sociological environment that reinforces an unjustified confidence in the modal claims about "distant" possibilities one accepts and asserts.