By Kelly Trogdon
This is my first post here – thanks Gualtiero for inviting me.
As I read Chalmers’ “Does Conceivability Entail Possibility?” (2002), he claims that a state of positive conceivability is positive in virtue of involving a “positive intuition” of a situation, and to have such an intuition is just to represent a situation, judging that it verifies a proposition. The claim is that you positively conceive P when you represent a particular situation, judging that it verifies P, while to negatively conceive P is just to judge that ~P is not a priori. (Here I’m glossing over various distinctions that come up in the paper that aren’t relevant to what I’m going to say.)
Let P be the proposition expressed by “Situation S is F”. Suppose that you negatively con
ceive P (i.e. you judge that ~P is not a priori) and then judge that S, qua situation with F, verifies some proposition P*. By Chalmers’ lights, your state of conceivability here is positive. I don’t think, however, that we should be comfortable with this result. It’s true that in this case you judge that S and P* stand in a certain epistemic relation, but it’s unclear why this alone should engender or require anything like a “positive intuition” of S, provided that I correctly understand what we’re trying to capture here.
Let’s say that you perceptually imagine P just in case you form a mental image of a situation S (a representation of S relevantly similar to how you would represent S in perception) and judge that S verifies P. I think I understand how mental images in this sense are positive, independently of the role they play in verification judgments. So I likewise think I understand why it is that perceptual imagination should count as a form of positive conceivability.
What I struggle with, however, is non-imagistic positive representation in this context. Assuming that such representation is absent in cases of negative conceivability, our example above shows that to form a non-imagistic positive conception of a situation is not merely to form a non-imagistic representation of that situation that you take to verify a proposition.
Everything seems right here, except (i) in your definitions of positive conceivability, “represent” should be replaced by “modally imagine”, and (ii) for just this reason, the sentence starting “By Chalmers’ lights” is false. Here maybe you’re representing S in some sense, but you’re not imagining it. Modal imagination is supposed to be broader than perceptual imagination but narrower than representation in general. Of course there is plenty of work to be done to clarify the notion (I do a little in the paper, as does Yablo in his paper), but in any case that’s the general structure.
Thanks for the quick response! As I was interpreting your view, the idea is that what makes a representation a case of modal imagination is (at least in part) that its content is a situation that you judge to verify a certain proposition. I thought that in your view it’s the “objectual” nature characteristic of positive conceivability – the fact that here you judge that certain situations verify certain propositions – that sets it apart from negative conceivability. The worry I had for this idea – that objectuality is the differentia between positive and negative conceivability – is that it’s unclear that such judgments are that in virtue of which states of positive conceivability are positive, for we can make such judgments about situations that figure in the contents of states of negative conceivability.
“Let’s say that you perceptually imagine P just in case you form a mental image of a situation S..”
I wonder if there is a multiplication of a single idea or entity here. P is the marks through which S is conveyed. As such, there is no relationship between P and S. P is either S or else is merely the marks on paper associated with S.
Let me put it another way. P is not a hybrid of marks and the meaning that is delivered by those marks, it’s either one or the other. If it is the latter then P is S. If it is the former then P (marks on paper) has no truth value associated with it.
I know that point might seem unpopular; the notion that a statement is a viable independent truth-tracking entity is popular even among logicians, yet its logical status is itself never examined.