consequently.org/writing/invention

Forgive what may be an impertinent comment, Greg, as I’m yet to read your paper. But a single word in the first sentence of your abstract stuck in my craw: “essential”. Do you mean to claim that one cannot understand modal logics and modal concepts by means other than a possible worlds semantics? This strikes me as (a) a very strong claim, and one which would take significant efforts to demonstrate; (b) very implausible, given that one could (for example) potentially gain an understanding of these logics and concepts through (say) a category-theoretic semantics. Perhaps a different semantics would give a different type of understanding, something we see often in mathematics; moreover, there may be semantic models for modal logics which are not yet known and which could provide even deeper and more profound insights than any we have today.

Posted by: Peter McBurney at October 13, 2005 10:27 PM

I certainly don’t mean to think that you can’t understand modal concepts without possible worlds semantics. I was thinking of “our understanding of the logical features” very broadly in the sense of “everything we know about normal modal logics” — we wouldn’t gave got that far without possible worlds semantics. I think that I’d be happy to claim that any successor “semantics” for modal logic will have to incoroporate the insights of possible world semantics in some way or other. How does that sound?

Posted by: Greg Restall at October 18, 2005 01:13 AM

I’ve not read very far into this paper yet, but I like the claims you are making. I’ve a list of corrections for the first few pages: would you rather get them in dribs and drabs or altogether when I’ve finished the paper?

One main criticism I have so far: you don’t distinguish between Kripke semantics and possible world semantics. There is a looseness in talk of the two kinds of semantics that doesn’t distinguish between them, but whilst this looseness is covenient, it also confuses, and there were places where I found myself suspecting your text of being a bit confused about this. Thomas Forster wrote a nice paper, The Modal Aether [ps] which is about this, amongst other things. The general point I’d make, is that while possible world semantics has been an inspiration for Kripke semantics, and still provides a useful intuition, there must be more to Kripke semantics than the possible worlds intuition can give you, since it is applicable to non-alethic modalities.

More to follow, not necessarily all that soon…

Posted by: Charles Stewart at November 22, 2005 06:09 AM

Link to Thomas’ paper: http://www.dpmms.cam.ac.uk/~tf/modalrealism.ps

Posted by: Charles Stewart at November 22, 2005 06:09 AM

Thanks, Charles. They’re useful comments. I agree that I’ve slid between possible worlds semantics and Kripke semantics in some of the discussion (let alone between explication, explanation and other issues when it comes to the relationship between what we might call modal ontology and modal ideology).

I’ve given a talk based on this material a few times (Nottingham, Bristol, St. Andrews) and I’m doing a few more (Oxford, Paris) and then I’ll get back to the written version on the basis of the comments I get.

So, you can hold off on little fixes until you get to the end. Big mistakes, of course, I’ll deal with anytime.

Posted by: Greg Restall at November 22, 2005 09:31 PM

I listened with great interest to your presentation of this paper in Paris. I have much sympathy for what you label Option 2 : let’s explain modal ontology by modal ideology, by the way we know and use concepts such as necessity and possibility. It is an inherent feature of Option 1 (the option that accepts the Kripkean semantic framework with no qualms on the way we come to possess modal concepts) that possibility and necessity behave in what I would call a symmetrical way - there’s no question of conceptual priority between the two notions. However questions about the conceptual asymmetry of modal concepts may arise if you opt for Option II. There’s no in-built motive, in modal epistemology, that one concept should not be more primitive than the other. I think that I remember that Hale’s paper on “Knowledge of Possibility and of Necessity” tends towards such an idea in favor on the primitiveness of necessity. I think, on different lines, that the original motivation you give for Option 2 in quoting Prior indicates some kind of asymetry in modal ideology in favor of necessity too : ” we understand “truth in states of affairs” because we understand ‘necessarily’”. So given those conceptual bases - which I suspect are corroborated by empirical data about ordinary modal cognition - there’s admissible leeway in the way you introduce modal operators in your framework. You may decide - in order to reflect basic ideological asymmetries or at least their conceptual possibility - to make an operator essentially more primitive than the other. One broad advantage, I think, in choosing - as you do - a proof theoretical account of semantical features of modal operators is that you are not constrained as a truth-theoretical account is compelled to - to inherit this symmetrical account from Kripkean semantics - and you are freer to make your logical systems more expressive of possible features of natural modal cognition.

Posted by: Sacha Bourgeois-Gironde at December 7, 2005 04:52 AM

I think I agree — while, as I mentioned at the talk, I actually have fairly strongly symmetricallist inclinations, I can see that there’s more room to wiggle things around given the inference-first account of things. Where can I go to get a good picture of the empirical data?

Posted by: Greg Restall at December 14, 2005 11:50 PM