Mark Sacks at the Transcendental Arguments Workshop

Mark gave a paper entitled 'Phenomeno-logical proof'. In it he argued for a specific understanding of what makes transcendental arguments transcendental - there being descriptions in argument form of what Mark calls presupposition relations between situated thoughts. The idea is that appeal to situatedness explains how transcendental proofs make a synthetic or ampliative step. If such arguments are construed as concerned with meanings, propositional contents or conceptual analysis, then it is hard to see how we get more than we started with. And anyway, the inference from 'there is a tree in the garden' to 'there is a tree in front of me' is not analytic, for such inferences fail in cases where the perceiver is remotely located, or where the statements describe, for example, trees and gardens on different possible worlds. But, once we situate the thinker so that the thought 'there is a tree in the garden' is demonstratively directed at its objects, the inference is made good and can be sanctioned a priori (no specific empirical information over and above the presupposition relations between the situated thoughts themselves is required).

The general idea then, is that what we are doing when we engage in transcendental proof is thinking ourselves into the relevant situated thought. We then see that certain presupposition relations necessarily hold between such thoughts. These necessary relations are what transcendental proof describes - a synthetic a priori body of truths about what might be termed the structural features of situated thinking. The description of situatedness at work is thin, but this is to be expected, as in order to make necessity claims, not too much must be built into the various cognitive scenarios we investigate transcendentally. Moreover, the necessary truths we uncover in this manner only hold from within the relevant situated viewpoint - they are not 'world-directed' truths, but truths about how experience must be from such a viewpoint. In other words, they are necessary truths about possible experience.

One objection was that how do we know that our inability to, say, not ascribe two contradictory predicates to an object at the same time was not being driven by general metaphysical principles about possession of incompatible properties. Mark's reply was that we are asking how incompatible predicates get applied in experience, and this requires recourse to the situatedness of the thinker, and thus to necessary presupposition relations between a pair of thoughts like 'this is a cube that is red all over' and 'this is the same cube that is green all over', and the thought that 'this same cube has persisted through time'.

Another objection inquired after what made the putative necessary truths yielded by situated thoughts necessary - i.e. are we not mistaking the limits of our imagination for some sort of metaphysical limit? Appealing to the modesty of situated thoughts - there not being world-directed - will not help here, for they still make necessity claims about what any situated thinker of the relevant type must experience the world to be like. Mark's reply was that there was no way to guarantee necessity here. Kant's proofs have stood the test of time, and so that is a marker of their success. I was unclear whether Mark was suggesting that transcendental proofs operate with a deflationary epistemically constrained understanding of modal truth, or whether they purport to uncover full-blooded metaphysical necessities (albeit about subjective experience). In other words, is the suggestion that there is no way to definitively verify that such truths are necessary, but that there is also no reason to think transcendetnal proof an unreliable guide to uncovering necessary truths, or, is it that continual convergence over time in phenomeno-logical proof is what being a necessary truth consists in around these parts?