synchronic and diachronic "explanation".

[R. Rankin]

> At the risk of making myself unpopular on the list,
I'd like to stick up
> for the idea that synchrony is NOT always just frozen
diachrony.  People do
> certainly memorize fossilized sequences, and many
things can certainly be
> explained that way, but speakers do also sometimes
reanalyze, reorder,
> create novel combinations...

Don't worry, you won't be unpopular; in fact I was
hoping/expecting you might weigh in.  Let's be clear
though on the fact that NO linguist claims that
"synchrony is ... *always* just frozen diachrony."  I
think the reverse is more often true, especially among
(the increasing number of) linguists who are totally
unfamiliar with historical/comparative work.

I guess I do have one or two more observations that
bear on our discussion though.

>  Anyhow -- while it may well be true that affix
sequences in Siouan are
> memorized chunks, historical fossils that children
learn by rote,
> ridiculing the idea of even looking for possible
synchronic analyses
> strikes me as much too harsh.

Ridicule might be a little strong, BUT, it seems to me
that there is no reason to look for synchronic
explanation of word or morpheme order that has not
actually changed.  And to know that, the linguist (not
the speaker/learner) has to know what has changed and
what hasn't.  The ordering of Dakotan /wicha-/, /uNk-/,
/ma- & ni-/, /wa- & ya-/ does not require synchronic
"explanation" because we know it represents the order
of addition to the system.  If that order were to
change, and, say, /uNk-/ were to suddenly join the same
position class as /wa- & ya-/, THEN an explanation is
called for.

I think theoreticians miss a really good bet by not
studying the morphosyntactic histories of their target
languages, because it is precisely where the word or
morpheme order has changed that they will find their
BEST EVIDENCE for UG ordering principles.  Whether the
learner knows language history is irrelevant to the
argument that the linguist and ESPECIALLY the
theoretician needs to know it.  If the theoretician
doesn't know the history, how can s/he know what
orderings represent changes and, thus, are due to
internalized rules/principles/ constraints, and what
orderings are due to accidents of history (and, thus do
not require explanation in terms of rules/constraints)?

The REordering of postposed pronominals in, say,
Mongolian, obviously requires "explanation", but the
(historically conservative) ordering of the Siouan
pronominals (above) does not.  It's a waste of time to
search for "principles" where the learner just took the
path of least resistence and learned the order mama's
way.  But ya gotta know which is which!

> As Shannon says, speakers don't know the history (and
no, I don't consider
> this a canard) --

I guess, from my perspective, I still do, for reasons
given previously and above.  I find it a poor excuse
for ignoring much of linguistic scholarship (and an
excuse that Slavicists and Romanists like ourselves
were never allowed to get away with!).  I have to admit
though, that historical linguists have been piss-poor
ambassadors for their subject.  Historical and
comparative linguistics has often been taught from
materials and by instructors that ignored all the
synchronic and theoretical consequences of language
history.  I've surveyed the historical chapters of a
number of popular intro textbooks, and I must say I've
been very dismayed by what I've found.  Many -- no,
most -- such chapters are just a string of cute
anecdotes about vocabulary, a boring presentation of
the Pater Noster in Old and Middle English, a handful
of Indo-European cognates and a couple of trivial
examples of "drive:drove::dive:dove" analogy -- plus a
few paragraphs of (forgive me)Labovian drivel.  We
should make sure that these chapters are written by
real practicioners of the discipline, and not just
assigned as an additional task to the poor fellow who
penned the phonology chapter.

But I'm preaching again . . . .   :-)

Bob