[YG Conlang Archives] > [ceqli group] > messages [Date Index] [Thread Index] >
Rex: > I'm mainly against them. There is > this notion in Loglan that I think is completely unreal, that > It is probably that he's asleep. > He's probably asleep. > are profoundly different, somehow. I don't agree. Now, there may be room > for some pinvor attitudinals of some kind, but certainly not with the ones > that have to do with conviction. I prefer the setup: > > X ke Y. > > Where X is a predicate and Y is the base sentence. Often the predicate will > just be an observative that can take an unspoken Go as a subject. Yes and no. Yes, what you say is correct, but X is not an ordinary predicate, because it doesn't contribute to the truth conditions of the sentence. Contrast 1a. Go away. 1b. I command you to go away. 2a. He's mad, of course. 2b. It is obvious that he's mad. There needs to be a way of distinguishing predicates that are part of the propositional content of the sentence, and predicates that express propositional attitudes. However, I don't think that issue has to be resolved at this stage. The distinction between 'function' and 'content' words or between 'cmavo' and 'brivla' is a relatively spurious one: in Lojban the difference is essentially phonological rather than grammatical, despite superficial appearances to the contrary. What I mean is that while it is necessary to distinguish between the two kinds of predicate, I don't think it makes much sense to try to define one lot as 'pivor' and the other as nonpivor, for the grammatical basis of the pivor class is built on quicksand. > Actually, I see no use for attitudinals with special forms at all. > Does anyone disagree? The Lojban 'attitudinals' represent a ragbag of different word classes all bundled together for purely syntactic reasons. So I see little use for recognizing them as a valid class. > Now, the Loglan position is that attitudinals do not change the truth value > of the sentence, which is absurd. To add 'maybe' to a sentence certainly > changes its truth value, not to mention an eo, which puts it in the > imperative mood, unless I'm missing something. > > To say 'probably' usually just modifies a whole sentence, or is a statement > about a whole sentence. > > To say 'please,' is to change the meaning of the sentence entirely. Speaking for Lojban, the idea that attitudinals don't affect truth values is an idealization which, because so much heterogeneous stuff got bunged into the class, has exceptions. That said, core attitudinals, including 'please', don't affect truth values. Yes, 'please' changes the meaning of a sentence, but to understand how it changes the meaning you need to realize that it doesn't affect truth conditions. --And.