Last week I was at the Arizona Ontology Conference, held at the White Stallion Ranch outside Tucson. This was a memorable occasion, with some philosophy papers interspersed among the horseback riding, hiking, and other cowboy activities. (As a bonus, there was a freak snowstorm on the Sunday evening, delaying my flight back by a day.) Joe Salerno has put some great photos online, and I've put some online as well, while Berit Brogaard has a report on some of the philosophy. Thanks to Laurie Paul for putting such an excellent event together. (Update: still more photos from Brian Fiala, Andy Egan, and Benj Hellie and Jessica Wilson.)
I gave my paper on "Ontological Anti-Realism" at the conference, with excellent comments by Jonathan Schaffer (who has given permission to put them online). One of Jonathan's points is that I "half succeed" because my view yields "half-realism": realism about the fundamental but not the nonfundamental. I'm happy enough with the half-realism, as this is more or less the line I take in the paper. But it's worth noting that the realism about the fundamental need only be a metaphysical realism, where reality objectively and determinately has a certain fundamental nature. This needn't be an ontological realism, where this fundamental nature involves a domain of fundamental objects having fundamental properties (it might be, on some versions of the view, but the framework doesn't require it).
Jonathan also gives three interesting arguments against the distinction between ordinary and ontological existence assertions of a sentence, at least construed as involving a difference in truth-conditions (as opposed to pragmatic correctness conditions). I think these arguments can be answered, but in any case it turns out that Jonathan and I were interpreting the claim that there is such a distinction in different ways: on my reading it's compatible with the claim that these assertions have different underlying logical form (e.g. involving covert variables), while on Jonathan's reading it is not. So where Jonathan proposes what he takes to be an alternative to the distinction, involving covert variables for furnishing functions, I'd happily endorse this proposal as an implementation of the distinction as I construe it. In any case the exchange and the discussion at the conference were very useful, and I hope to have a revised version of the paper online before long.