From: Michael L. Siemon (firstname.lastname@example.org)
Date: Wed Oct 25 1995 - 00:11:03 EDT
>I'm kind'a hoping we can end this thread soon, but since I've gotten
>flooded with questions re "future tense" I think it's only fair to
>respond a little.
Well, the thread has been fascinating and extremely suggestive to me,
and your contributions not the least -- however, I think here that your
attempt to squeeze a lot into a "little" response exposes your smooth
sailing to an iceberg or two ...
>first, I'm not sure what's wrong with either formalism or theoretical
Nothing, per se; but there is a problem with over-systematization of
language, which you illustrate below:
>the other problem is consistency of form to meaning. and I'm sorry,
>but you're not going to get this except maybe in Esperanto.
Exactly -- no REAL language HAS such a consistency.
>if I say
>that the English plural is the stem plus /-s/, some young'un will
>point out that women obtains and not womans.
The problem is more interesting than this quick dismissal -- the
very young WILL pick up your elegant "stem plus /-s/", and then
after much correction by older speakers of English, will tend to
err in the opposite direction, by making irregular plurals or
irregular pasts or whatever in cases where the typical speaker
*does* use the rule. :-)
In fact, there *is* no consistent synchronic system of any real
human language -- just a near approximation to one such system
or another, varying over time. In your quest for "compositional"
semantics derived out of consistent pieces, you are implicitly
assuming a uniformity of language without the "fault lines" along
which language change happens diachronically. With no assured
"tracking" of syntactic and semantic elements in correlation.
As in Mari's (entirely laudatory) goal of teaching language compre-
hension with a minimum of "exceptions" to the "rules", I am quite
in favor of the systematizing (and the goals of simple elegance)
you urge so passionately. But you seem to be being carried away
in your passion, if I may say so :-)
>well, that 's quaint.
>should we give up the analysis of the regular plural on the basis of
>less productive morphology?
No -- of course not, as long as we recognize that there are in fact
*several* rather incompatible things going on at once! It is (I think,
though I will accept correction readily on this) a fact of life that
verbs in Greek *are* systematically something of a mess (e.g., in
comparison to Latin.) However much tidying up of this mess you do,
you are still left with a whole raft of things like irregular English
plurals and highly sporadic ablaut. And those things ARE part of the
language, however inconvenient for synchronic theorizers.
There is a quotation running around the net, apocryphally ascribed
to Einstein, that "things should be made as simple as possible, but
no simpler." I believe that rule should be inscribed on the lintels of
all schools of linguistics :-)
But again, I have to concede that my gut entirely approves of the
style of thing you are doing, or that Mari is doing. *Given* that our
synchronic systems are *not* going to be an exact match to the
language as she is spoke, you have all the motivation in the world
to construct a model that conveys the greatest possible sense of
the actual usage with the least need for ad hoc patches, and I at
least rather like some of the stuff you guys are saying about TMA
(at least as regards Greek; my Semitic is nearly non-existent, so
I have no opinion at all on that.)
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:31 EDT