Re: [lojban] Uses of Language

[pycyn@aol.com]

In a message dated 7/6/2001 8:51:42 PM Central Daylight Time,
jjllambias@hotmail.com writes:

I'm rather thinking of a semantic classification of the structures
(which syntactically are all alike) but independent or prior to the
use of those structures. Questions seem to be the easiest to identify
(and maybe should be subdivided into fill-in-the-blank questions and
true/false questions).

The program would be to classify sentences of the form <UI> <bridi>
in terms of what the bridi refers to depending on the UI. (Then we
can consider what happens with subclauses and so on, but we should
start with the simplest cases.) I suppose I'm restricting myself to
uses of bridi instead of considering all uses of language

Well, aside from the word "form" this looks like a useful thing to do and it
focuses on what has been the main problem in the current discussion (the
expressive - assertive distinction aside).  But, once you get the relevant
categories 9and tha comes out of the uses, I think), you can do it pretty
much a priori:
Does the speaker indicate that he believes the bridi true?
Yes: assertion
No: Is it because he cannot interpret the bridi coherently?
Yes: metalinguistic (maybe better as "para") response.
No: Is it because he believes it false?recommendation
Yes: contrary-to-fact moves: speculation, directive
No: Unclear about whether or exactly what has happened: question
      Has preference about what happens: hopes, wishes (NOT expectations,
despite the notes).
<I'm not sure whether in this scheme performatives should be in a
different class than assertions. In them the bridi refers to a
situation that comes to hold in the world as a result of or in
conjunction with the utterance, so in a sense they could be
considered to refer to a situation that holds in the world.>

I put them in as minimal directives: you know what you say is not true and
you act to change it and thus change status of others and perhaps self.