* Saturday, 2014-10-11 at 12:42 +0100 - And Rosta <and.rosta@gmail.com>:
> Martin Bays, On 11/10/2014 03:12:
> > it looks like {tu'a} is in LAhE only syntactically, not
> > semantically, and must be handled separately.
> >
> > (So then tu'a needing opacity is no longer an argument that the rest of
> > LAhE should get it...)
>
> Given that syntax is logical form -- rather than combinatorics of
> morphophonological forms, which is pseudosyntax -- your framing of the
> issue should not be accepted. (I realize I've expressed that in an ex
> cathedra way, but I'm happy to argue the point if it is contested.)
I'm not familiar with the syntax vs pseudosyntax distinction. Probably
there are multiple competing definitions involved, but at least one
meaning of "syntax" has the question of what strings are accepted by
a formal grammar to be a matter of syntax. That's what I meant.
> Me I would advocate throwing away the pseudosyntax as the unlinguistic
> junk it is, but anybody set on keeping it as the basis for actual
> syntax couldn't get away with this wishful distinction between
> 'syntax' and 'semantics' where 'semantics' is used to mean 'structure
> of logical form'.
Of course logical forms are also syntactic, but the (probably not fully
realisable) aim here is to translate lojban to a logical formalism whose
semantics is standard or relatively straightforward to define, so the
logical forms are at least a useful proxy for the actual semantics. Of
course there's still a distinction; equivalent non-equal formulae exist.
Attachment:
signature.asc
Description: Digital signature