In a message dated 9/29/2001 9:33:33 PM Central Daylight Time, a.rosta@dtn.ntl.com writes:
You Ah, the old words are the best ones -- though I do like {ce'u}. Indagations was just what I was about, despite some other helpful suggestions of what I might have been about. And, in the process, I notices a nice pattern emerging, which I then took on to another case as a test, where it worked perfectly. I suppose that that case could be called an addition, since it had not been used before, but it was part of a pattern and used an already existintg -- though unexplained -- construction. So far nothing has been demolished, so I feel safe to proceed with any interference from &&X. What is clearly known about {ce'u}? The Refgram gives an obviously confused report about {ka}, making it a property with nothing for it to be a property of, and then introduces {ce'u} -- first one, then several per {ka} -- to make properties of things (it still is not clear what a property that is not a property of anything is -- a proposition is the best guess and does fit the examples fairly well -- and also the logic). Elsewhere the Refgram says that {ce'u} is used with {ka} and perhaps other abstractors, but -- unlike the neighboring case of {ke'a} -- does not say that {ce'u} is used *only* with abstractors, though that may be implicit (always a bad plan, we have found). The other "certainty" about {ce'u} is that it is "a lambda variable." This is not in the Refgram but comes, apparently from Cowan's presentations which led to introducing {ce'u} in the first place and his later comments at various times (but I can't trace a decisive case -- Elephant?). The one limited and slightly inaccurate discussion of the lambda calculus in Refgram makes no mention of {ce'u}, nor of something like its presumed role in Lojban, but rather talks about representing the calculus in MEX, a totally different task. Of course, in the process it gives a uniform way of representing functions (and properties and relations and the rest of the abstractions in Lojban without the messy problems of NU -- but with the far more messy problems of MEX). It is clear that the idea of a lambda variable is behind {ce'u} but it cannot be claimed that {ce'u} is one to the extent of using such a notion to the full. At least not as part of the accepted wisdom. Since the question of {le mamta be ce'u} has noot been considered before now, there does not seem to be an accepted wisdom about it. The claim that {ce'u} is restricted to abstraction phrases is based on a parallel with {ke'a}, but {ce'u} differs from {ke'a} precisely in lacking a restriction to one kind of phrase. The two factors can be weighed either way, so probably come to a draw. The notion that {ce'u} is a lambda variable is held pretty widely as a claim, but without any significant understanding of what it means. It is also not in the Refgram, so cannot claim that kind of authority. It would seem then that any attempt to introduce a new explanation of a form on that basis is probably not going to be successful. So, I guess that, if I am going to argue for {le mamta be ce'u}, I have to tie it in with other structures and shoow how well it fits. Unfortunately, the other structures with which it fits are indirect questions, for the most part, an thye are still a matter of controversy, with one fairly precise view about them and one that wanders around a bit and gives apparently counterintuitive results, but is vigorously defended by And. So, I think for the nonce, I'll make it my task to nail down questions. Then, when they are part of the accepted wisdom (extension claims or whatever having received a well-deserved burial -- like that ever happens here) I can come back to make the case about {le mamta be ce'u}. |