Saturday, March 11, 2006

how could possible possibly be ambiguous?

reposted from coments here.

The Baudrillard passage at the centre of the rumpus:
"Before the event it is too early for the possible. After the event it is too late for the possible. It is too late also for representation, and nothing will really be able to account for it. September 11th, for example, is there first—only then do its possibility and its causes catch up with it, through all the discourses that will attempt to explain it."

BigSecretDog's conclusion:[2]What he seems to be telling us is that prior to an event there are no possibilities or causes, that the possibilities and causes of an event come to be only through the discourse that places them prior to that event, and thus that discourse is prior to the possibilities and causes of the event that it describes or explains.

My response: I’ll argue that there's a reading of the passage in question that vindicates Baudrillard. The maxim of charity is (roughly) that one ought to attribute to one's opponent the most reasonable construction of his views. There's a construction of B's passage available that keeps it consistent. So, we ought to adopt that view. That, however, requires an appeal to epistemic possibility. You don't like that. So I'm going to try and defend the notion.

Your crucial inference is from :
[A]'Before the event it is too early for the possible'
[B]'What he seems to be telling us is that prior to an event there are no possibilities or causes.'

You take B in the metaphysical sense i.e. there are really no causes… etc. That reading of B is justified if A is a metaphysical proposition, otherwise not. A is a metaphysical proposition if the notion picked out by the word ‘possible’ is a metaphysical one. You clearly think so. So, your case relies at a crucial point on the thought that the modal adjective 'possible' univocally refers to metaphysical possibility.

Let a statement S be epistemically possible if, relative to some knower’s knowledge at some time, S could be true. Let metaphysical possibility be absolute possibility, i.e. only what is possible to be (in the 'to exist' sense). Suppose a statement S describes a state of affairs SA. SA is metaphysically possible if it is the case that SA could come to be. Epistemic possibility and metaphysical possibility do not coincide - something can be metaphysically possible without being epistemically possible, and something can be epistemically possible without being metaphysically possible. Note that there's no decisive reason to think that we have exhaustive knowledge of what is absolutely possible.

Now, your claim about the meaning of the word 'possible' is false, as a matter of easily-ascertainable fact. Modal verbs (I'm thinking about your use of the word 'can') and adjectives are often ambiguous between their epistemic and primary (e.g. normative) uses, as you'll discover if you google. Thus, when one says, 'X is possible', there usually remains the question – what sort of possibility have you in mind?

A vivid case: [G] It is possible that God can damn an innocent man.
If 'possible' in G is read as metaphysical possibility then the sentence is false. If 'possible' means consistent with God's omnipotence, and supposing that to mean that God can bring about any consistently describable situation, then it is true.

Another example: Suppose I solve a complicated sum, and someone asks me whether it is possible that I'm wrong. If I'm wrong, then I'm necessarily wrong; if right, I'm necessarily right. Either way, it's a matter of necessity. So what sort of possibility is in question here? epistemic possibility.

Yet another example (the modal fallacy): If I know p, it is not possible that p is not the case. But if it is not possible that p is not the case, then p is necessarily the case. The sort of possibility that is required to defuse the difficuly is epistemic, not metaphysical possibility

Example: It is possible that I don’t exist. If I say this, I’m not asserting of myself that I don’t exist (or else it is necessarily false). Rather, it means that for all I know, it could have been the case that I did not exist.

Yet another example: Someone comes up to you at a ball and says: ‘That masked man is your father’. You say: ‘That’s not possible!’ In fact, the masked man is your father. And given that necessarily your father is identical to himself, and that necessarily your father is your father, it looks like you had something like epistemic possibility in mind.

Examples could be multiplied.

The point is that the word 'possible' often expresses different modalities. One can't deduce from the fact that a sentence has the form 'X is possible' that it makes a statement about metaphysical possibility. [You might (!) like to see this note on modal confusions]

Now, suppose that there's a fact, F, such that at a given time, for some person, or group of people, F is epistemically inaccessible. There is no reason to think that that F is not also metaphysically possible. Further, actuality implies possibility. If F should now become the case, then it was always metaphysically possible that F, and it is now (let us suppose) known and hence possible to know that F is the case. But then, one can truly say that there was a time in the past when it was epistemically impossible that F. A fortiori, B, exploiting the ambiguity of 'possible', can say that "impossible things can . . . happen", and all the rest of it.

So, we've established that 'x is possible' doesn't always mean 'x is metaphysically possible'. And we know that there's a perfectly intelligible notion of epistemic possibility. B's extract, read in that light, is not only coherent, it is rather banal. I suggest we attribute banality rather than vice.

No comments: