WikiDiscuss

---

> On 5/17/05, John E Clifford
> <clifford-j@sbcglobal.net> wrote:
> > But the definitions given preempt their
> future
> > use, where they might have served for some
> real
> > need and now are stuck with a formulaic but
> > useless sense. (People manage to ride over
> this
> > kind of problem all the time, but it would be
> a
> > courtesy not to stick them with it in the
> first
> > place.)
>
> I wish you'd come up with at least one
> suggestive
> example of how this might happen. All this
> theoretical
> speculation of what might be the case if we
> don't do
> what we are doing is not very convincing.

"Give me an example of a non-existent object or I
won't be convinced" is not a very interesting
comment. I have cited cases from the past where
the rules being in force would have prevented a
useful usage; I extrapolate from that.

> > > None of the rules are exceptionless, but
> they
> > > are quite extensive.
> >
> > Yes, things tend to turn out in a certain
> way,
> > just as lujvo tend to fall into a small
> number
> > (say two) of patterns. But that hardly is a
> > basis for claiming that future items will
> always
> > fall into the predominant pattern, as the
> > definitions given here do.
>
> For example, which definition do you find
> overly restrictive?
I would say any one that says the meaning of
this form has to be such and such, based on some
rule rather than some usage. Examples seem to be
too numerous to list: all cases with made up
examples would do in principle.

> > Let actual usage
> > decide; the remarks should give the sort of
> hints
> > that might lead one to find a word that would
> > work for a notion that one has in mind but
> ought
> > not preclude the use of a rare — even so-far
> > unexemplified — pattern.
>
> For example?

Well, back to the already cited past cases. I
have said all that I need about the demand for a
nonexistent object.

> > But so, of course, is the original listed
> form
> > {ri'a nai} (indeed, it is better yet) even
> though
> > its justification is not just in terms of
> > patterns of connections between a BAI and
> some
> > brivla, but rather in terms of general
> linguistic
> > principles applied to such a connection (but
> note
> > that even the connection is not actually
> > necessary for finding uses for these forms,
> > merely — as CLL says somewhere --
> suggestive, a
> > mnemonic, not a definition).
>
> I prefer compositionality to suggestive
> mnemonics
> whenever possible, and in this case it is
> possible, so
> my preference goes to {to'e ri'a nai} despite
> the two
> additional syllables.

Well, we have different tastes at this point; I
prefer the possibility of creativity to the
strict requirement that all be done by rules. It
does not seem to me appropriate for the grammar
to take sides on this, hence my suggestion that
the unused forms be declared undetermine but als
show which way the predictions of their use lie.