Wright's and Wald's comments on Dixon's "The rise and fall of languages"

[Robert R. Ratcliffe]

----------------------------Original message----------------------------
re: Roger Wright's question:
 
 
>Am I really the only person on the List to find
>actual language *splits* (as opposed to obvious and normal dialect
 
> >diversification) inherently unlikely, and thus in need of some kind
> of
> >non-linguistic explanation?
 
No, I would agree that clear splits are unlikely, at least in the case
where language spread doesn't involve loss of contact.  But I don't
think it is right to see the development of national orthographies as a
CAUSE for the diversification of languages.  I think the more important
point may the development of new patterns of political division and
concentration, of which the development of national orthographies is
merely one linguistic symptom.
 
The linguistic situation in the Arab world today is quite comparable
with the Romance situation in the 12th century. To begin with, you have
a huge dialect continuum stretching from Mauretania to Iraq, with more
or less mutual intelligibility between any two neighboring communities,
and unintelligbility between more widely separated groups. The only
official written Standard throughout the region is Classical Arabic,
different from all of the dialects, and comparable with Latin.  At the
same time in areas which have a stable, historically-established
political identity, like Egypt or Morocco, you do have a de facto
national spoken standard, based on the language of the capital cities,
which is understood and recognized throughout the country in spite of
regional dialect variation, and which could (if the political will were
present) be developed as national (written as well as spoken)
languages.  Levelling of regional dialects toward the standard of the
capital is already apparent in these countries.  In countries whose
political history is newer, Algeria, Jordan, Saudia Arabia, e.g., there
really isn't  a clearly defined national dialect, though if the
political boundaries remain as they are perhaps one will eventually
crystalize.
 
Another point about Arabic dialects ties in (I think) with Benji Wald's
point:
 
> Meanwhile, historical linguists sub-classify reconstructed languages
> without this social implication, although there is clearly some kind
> of
> social implication for any form of diversification
 
It is strange to me that people who work on sub-classification of Arabic
dialects use entirely different methods and assumptions from people who
work on sub-classification of prehistoric Semitic.  In the first case,
no one ever tries to use a tree diagram, because everyone realizes it
wouldn't have much explanatory value. The more isoglosses you use the
more confused and contradictory the situation becomes.  In addition to
isoglosses defining areas, and the usual sort of Rhenish fan effect, you
have isoglosses which crosscut areas and correlate with social factors--
such as communal membership (Muslim, Christian, Jewish) or socioeconomic
stratum (Bedouin vs. Rural vs. Urban).  But Arabic dialectologists are
not disturbed about this complex linguistic situation because we
actually know from historical and other non-linguistic sources that the
spread and diffusion of Arabic was a complex historical process, and it
is quite easy to match the linguistic situation with the actual
historical events which brought it about. (For example the Hasaniyya
dialect of Mauretania has a pattern of verb conjugation which is typical
of the dialects of North Africa, but in some of its phonology and
lexicon it is quite different from these dialects and close to the
Bedouin dialects of the Arabian peninsula.  This correlates with the
fact, known from history, that this dialect was brought into the area by
a bedouin group from Arabia some six hundred years after the initial
Arab settlement of North Africa.)
 
In Semitics, on the other hand, many people seem terribly disturbed
about the fact that some languages (notably Classical Arabic) can't be
fit neatly into a tree diagram because they share some isoglosses with
languages on one branch of the tree, and another set of isoglosses with
languages on another distinct branch.  When confronted with such a
situation what we should be doing is asking "what kind of
socio-historical situation (what patterns of migration or contact) does
this contradictory linguistic situation imply?"  Instead many Semitists
have focussed on trying to theorize away the contradictory isoglosses in
order to preserve a sharply delineated tree model.  This procedure is
tantamount to asserting that the normal or expected process of language
spread in pre-history involves a sharp split in a speech community
followed by little or no contact, whereas in historical times this
pattern of langauge spread seems very rare indeed.
 
 
 
 
 
+++++++++++++++++++++++++++++++++++++++++++++++++
Robert R. Ratcliffe
Senior Lecturer, Arabic and Linguistics,
Dept. of Linguistics and Information Science
Tokyo University of Foreign Studies
Nishigahara 4-51-21, Kita-ku
Tokyo 114 Japan