Dixon's "The rise and fall of languages"

[bwald]

----------------------------Original message----------------------------
Roger Wright writes:
 
>- Manaster Ramer discussion seeems not to refer to literate communities,
>so it seems sensible to point out that if we find out that relevant
>groups are still physically together, over whatever length of time, we
>would not expect a split at all.
 
That's wrong.  Dialect diversification does not depend on loss of physical
or communicative contact.  Since such diversification is true of all
languages known, it is reasonable to assume that it has other causes, some
of which are fairly well understood now, e.g., social differentiation,
local identification, etc.  Also, if what Roger says were true, then the
historical notion that the *area* of greatest diversification is the most
likely origin of a particular language would be completely illogical.
 
Scott Delancey wrote:
 
>I (along with a long list of others) would argue that there is
in principle no way to understand a lot of facts about synchronic
structure except in terms of grammaticalization.  The fact, for
example, that syntactic categories are *normally* not airtight
and completely discrete.  How can there be any useful account of the
English "quasi-modals"--gonna, oughta, usta, etc.--that doesn't have
>a clear diachronic dimension?
 
The remarks I made about the relation between synchrony and diachrony to
which Scott addressed this were not meant to be all-inclusive, but a
minimum for which I felt there was a general consensus.  Meanwhile, Scott's
suggestion is not totally clear to me.  I don't understand what he means by
'*useful* accounts', unless it tautologically means 'useful' to him as a
historical linguist.  I don't really think that.  I think he has adopted a
point of view of (synchronic) grammar as diachronic pragmatics, putting
priority on pragmatics as an explanatory principle underlying "grammar",
where, I suppose, "grammar" is a conventionalised set of pragmatic
strategies  (Givon proposes something like that).
 
The issue was raised by Janda about whether synchronic applies to an
individual speaker in space and time, so that such a person is not privy to
historical deductions.  To me the interesting issue is that in many cases,
built into the speaker's unconscious "knowledge" of his/her language is a
great deal of the history and even imminent, as well as possible, future
directions of change.  This was advanced, for example, for the variable
constraints on copula deletion in Black English, synchronic reflexes of a
non-English type copula system, and a number of other cases, to which I
could add from my own research on linguistic residues in speech production,
e.g., some tense-markers I have studied in Mombasa Swahili; speakers use
them in speech in ways they are not aware of, but that reflect
*DE-grammaticalisation* of a distinction which can be deduced to have once
been obligatory in the earlier history of East Bantu.  For me, the boundary
between synchrony and diachrony is empirically problematic, and does not
reside in the difference between an individual speaker and the language
spoken by a larger community of speakers over more than an individual's
lifetime.
 
An example of the problem of synchrony vs. diachrony that I have been
recently studying has to do with NV compounds in English.  Until recently,
the received wisdom was that they arise by backformation from NN (e.g.,
NV-er/ing) and NA (e,g, NV-en/ing) compounds.  That is, as epiphenomena of
NN or NA compounds.  But more recent proposals are that NV has the same
status as NN and NA in synchronic analysis.  Neither position is obvious,
and, indeed, I think the different proposals reflect both a change in point
of view (from diachronic to synchronic) and an actual change in English
"productivity" of NV, to which these points of view are reacting.  There's
more to it, but you get the gist.
 
Next, Miguel quotes Dixon's:
        The most important task in linguistics today  -- indeed, the
        only really important task -- is to get out in the field and
        describe languages, while this can still be done.  [Other
        things] can wait; that will always be possible.  Linguistic
        description must be undertaken now.
 
One cannot argue with this to the extent that ANY description is better
than none, no matter how limited, cf. Etruscan and many others.  In fact,
at the present time I think there is division of labour such that some
linguists are better at describing languages than others, and that some of
the others are better at raising interesting issues about language and
languages which the language describers can use to make better and fuller
descriptions.  Obviously, no language has been or ever will be fully
described, nor will it ever be obvious what issues will emerge from
linguistic phenomena that we do not currently pay attention to but which
will turn out to be important.  I would even say that Dixon is good at both
description and theory, and has not by any means forsaken his theoretical
excursi in order to concentrate solely on preservation of endangered
languages, even at the risk of garbling information about languages he is
not familiar with.  By all means, go out and preserve undescribed
languages, but note the problems that arise, and whether "theory" can be
dispensed with.