In a message dated 9/24/2001 11:10:18 AM Central Daylight Time, arosta@uclan.ac.uk writes:
Much though I rue it, I remain much in the dark about the formals of lambda, Well, much as I hate to put Logic up against Reason (I have less worry about yours or xorxes' reason), it does seem to work out diffferently, and the things that you have said actually seem to support that view (not unusual: Quine once wrote a paper that clearly recommended a certain logical device but which he claimed ever after refuted that use definitively). If we shift back to lambda (and I admit I haven't messed with this for thirty odd years), what you want in the first case is ^x^y Lxm<y>, where both lambdas are on the sentence level. I would read your sentence as ^xLx^ym<y>, where one lambda is on the term level, creating the name of a function just as the first does of a property. Clearly, we need a way of saying ^xf<x> in Lojban and we need an explanation for {le broda be ce'u} in Lojban. The analogy with the abstractors other than {ka} suggests that the two problems have the same solution, i.e., that the _expression_ is the way to express the function. Xorxes, I suppose, wants the function to have an initial flag, like {ka}, though of a rather different nature. You want {ce'u} to be transitive over some contexts, though not over others (else the extension-claims explanation of indirect questions will get into trouble -- the set of answers one as well, of course). Maybe the only context involved is bare LE, without NU (but I wonder if the analogies won't render actual cases of {nu} suspect to you too). Or perhaps your point is that {ce'u} is only sensible in the scope of an abstraction -- an maybe only certain abstractions at that. How do you feel about {la djoun mamta ce'u}, for example, or {nu da mamta ce'u}? They are grammatical and have natural clear interpretations: do you reject them or give them different interpreations or what? |