Re: Unambiguous languages (was: EU allumettes)

Ray:
> On Monday, May 3, 2004, at 04:06 PM, And Rosta wrote:
> > Ray:
> >> On Sunday, May 2, 2004, at 02:22 PM, And Rosta wrote:
> >> [snip]
> >>> I was thinking rather that since this hypothetical EU language would
> >>> be for official & legalistic use, it might as well confer the
> >>> additional boon of being unambiguous, in which case some kind of
> >>> amalgam of several dozen European languages would be quite the
> >>> wrong way to go about it.
>
> I agree the amalgam of several dozen European languages is not going to
> produce an unambiguous language; indeed, there already exist more than
> enough such amalgams.
>
> >> But isn't this precisely what 18th century conlangers like Dalgarno &
> >> Wilkins thought & aimed for.
> >
> > No AFAIK. I may be wrong, but I thought their main desire was for
> > a common scholarly language that, perhaps as a beneficial byproduct,
> > systematized enlightenment knowledge.
>
> Let Dalgarno speak for himself:
> "About twenty years ago I published....a Synopsis of a Philosophical
> Grammar and Lexicon, thereby showing a way to remedy the difficulties and
> absurdities which all languages are clogged with ever since the Confusion,
>   or rather the Fall, by cutting off all anomaly, taking away all
ambiguity
> and equivocation, contracting the primitives (primary words) to a few
> number, and even those not to be of a mere arbitrary, but a rational
> institution, enlarging the bounds of derivation and composition, for the
> cause both of copia and emphasis. In a word, designing not only to remedie
> the confusion of language, by giving a much more easie medium of
> communication than any yet known, but also to cure even Philosophy itself
> of the disease of Sophisms and Logomachies."
>
> >> Fortunately, neither of their conlangs, nor any other similar 'logical,
> >> philosophic' conlangs of the time caught on. Now 3 centuries later, we
> >> see that their philosophy & logic was a wee bit mistaken.
> >
> > To me they're no better or worse than your average IAL
>
> They were engelangs - rather different from the usual auxlangs encountered
> on that other list.
Yes. Perhaps on reflection I would say that they are better than the
average IAL, at least by virtue of their more estimable intent.

> > (all of which
> > are worse than your average natlang). If, though, you are right that
> > they aimed for unambiguity, then I'd count that as a big plus in
> > their favour.
>
> They surely aimed for unambiguity.
OK, but see below.

> >> But if unambiguity is the aim, there's always Classical Yiklamu   ;)
>
> I thought an aim of CY was to maximize unambiguity (as far as it's humanly
> possible). Am i mistaken?
Semantic 'structures' have a paradigmatic dimension and a syntagmatic
dimension. Avoidance of ambiguity in the paradigmatic dimension would
be avoidance of polysemy. This is what CY does (IIRC) and what I
think Dalgarno was probably thinking of primarily. The syntagmatic
dimension would have to do with such things as logical scope and
with matching arguments to predicates. I tend to be more preoccupied
with the latter (and to simply take the former for granted), but of
course a fully unambiguous language would have to pay due heed to both
dimensions.

> >> [snip]
> >>> to an equal degree. The one wholly official language would merit
> >>> its status by virtue of its superior qualities of unambiguity
> >>> (superior for its legalistic purposes),
> >>
> >> I fear this is like looking for a chimaera.
> >
> > Politically, of course. But not linguistically. I think it is
> > instructive to realize that a language that has the expressive
> > capabilities of a natlang but that is unambiguous is
> > linguistically achievable.
>
> Doesn't lojban attempt to achieve this?
It's probably fair to say it does attempt to achieve this, but
without success.

> Isn't Livagian meant to be an attempt to achieve this or have I
> misunderstood its aims?
No, indeed Livagian is such an attempt. But it is incomplete and
unpublished, a mere promissory note.

> Hasn't this been attempted many times over the past three hundred
> years or so?
I don't know. I wouldn't be surprised if many conlangs have claimed
to be, or to attempt to be, unambiguous. I doubt that many, if any,
have really tried. And I doubt that any have succeeded. Success
requires not only considerable skills as a language designer, but
also particular knowledge of linguistic semantics, especially its
logical aspects.

--And.