tag:blogger.com,1999:blog-5275657281509261156.post7365041888421419816..comments2024-03-28T04:04:55.806-07:00Comments on Faculty of Language: The 4th Hilbert Question: Is There Repair by EllipsisNorberthttp://www.blogger.com/profile/15701059232144474269noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-5275657281509261156.post-48216586055765550002015-01-08T11:33:21.464-08:002015-01-08T11:33:21.464-08:00I imagine Craig and I will want to engage the subs...I imagine Craig and I will want to engage the substantive discussion after he gets back from LSA, but in the meantime let me address some apparent presuppositions behind Dennis's comments.<br /><br />1) While Haider & van Riemsdijk did invoke Hilbert in their invitation that led to our squiblet, they wanted to select 100 questions from a larger pool of submissions, while Hilbert apparently only managed 23 questions in total, so they/we weren't necessarily aiming for Hilbert(i)an heights to begin with. In abandoning their book project, they said most submissions failed to even pose questions in the way they had in mind, but we believe we did, which is why we submitted the squiblet to Norbert. You may think his standards for posting these are too low, but I think his reply speaks for itself.<br /><br />2) They imposed a 1400 word limit. Norbert imposed a 4-page limit. We barely/sort of stayed within those limits, so we really *didn't* have space to say more.<br /><br />3) We were using "islands" the way most people now use the term--as a way to refer to constraints on extraction that have been rigorously defined and established empirically, e.g. in experimental work by Jon Sprouse and colleagues. In fact, Jon and Norbert edited a volume about islands recently, so we can defer to it/them for justification of the notion. If no one has a good *explanation* of islands at the moment, that shouldn't preclude talking about them, otherwise we couldn't talk about much of anything.<br /><br />4) Someone who is unwilling to accept that "ill-formed" and "acceptable" can be part of a "substantial" discussion of syntax must believe that such discussion is impossible, as far as we can tell. Some readers may be aware that a great deal of ink has been spilt on this topic since the mid-90s, but this isn't the place to delve into that. Anonymoushttps://www.blogger.com/profile/01589309048777608111noreply@blogger.comtag:blogger.com,1999:blog-5275657281509261156.post-82299476171790650722015-01-07T11:55:06.379-08:002015-01-07T11:55:06.379-08:00Not sure I agree: that's a very high bar for a...Not sure I agree: that's a very high bar for a Hilbert Question, and I am told that even Hilbert's questions didn't always meet such a bar.<br /><br />In fact, it seems that Bruening did not find the question too abstruse to provide evidence against it. <br /><br />As for the rest, it sounds to me like the question raised a lot of good issues: for example is deletion generally of constituents. You seem to be suggesting that it is not. I am less sure than you are about this. It certainly seems like a good 0th assumption given the role that constituents play in other parts of the grammar. You also question the link between 'ill-fomred' and 'acceptable.' As a methodological observation this seems reasonable. But as a matter of practice, do you really want to claim that there is NO relation between unacceptability and ungrammaticality? Is the former not a prima facie indicator of the latter in the kinds of cases linguists look at? If not, then more than the ellipsis stuff is at risk, no? And as for islands, yes they are theoretically opaque given current assumptions. But doesn't how else to investigate better theories than to consider cases where apparent generalizations are violated. This has been the Merchant-Lasnik strategy and I for one find it to be a perfectly coherent one. In fact, I cannot think of another that won't benefit from these sorts of inquires.<br /><br />Btw, what are Bruening's counter arguments. These would be worth knowing., for if B is right, then at the very least the problem has to be reframed, if not discarded. Why not post a few of them and let's see what Sailor and Schutze have to say. Norberthttps://www.blogger.com/profile/15701059232144474269noreply@blogger.comtag:blogger.com,1999:blog-5275657281509261156.post-24056333136006254762015-01-07T11:40:41.423-08:002015-01-07T11:40:41.423-08:00In my view, there are way too many dubious assumpt...In my view, there are way too many dubious assumptions in here for this to be a proper "Hilbert question." It is simply stated, for instance, that while "[s]pace restrictions preclude exemplifying each phenomenon, [...] they can all be shown to involve the derivation in (1)" -- but this is not clear at all. (See, e.g., Bruening's recent refutation of the Sailor & Thoms analysis.) The claim that ellipsis must in all cases be implemented as constituent deletion has never been properly justified, to my knowledge. (Merchant's main motivation was the E-feature, but that is an ad hoc, descriptive device.) And the question if ellipsis can "repair" islands can only be asked once we have an idea what islands are -- and we currently don't. The central question, "To what extent, if any, can ellipsis make an ill-formed structure acceptable?" is substantial only to the extent that the notions "ill-formed" and "acceptable" are given some meaning; as it stands, those are just informal notions.Dennis O.https://www.blogger.com/profile/07200488340449742505noreply@blogger.com