Age of various language families

[vovin at hawaii.edu]

----------------------------Original message----------------------------
> Sasha Vovin brought up some very good points about the non-
correlation
> between primary branches and language family age. I'd just like to
> pointout that although Robert Blust is an outstanding scholar and
> an expert in
> Austronesian linguistics, his account of the branching of
> Austronesian is
> not the only one out there. The late Stanley Starosta, another true
> expertin the field, held that Austronesian had just two primary
> branches, and
> that all the Formosan languages split of from one or another of
these.

Blaine, the idea that all Formosan languages are within the same
subgroup did not originate with Stan -- as far as I know it goes back
to at least Dahl's work. To refresh your fuzzy memory (:-), I believe
that you refer to Starosta 1995 -- "A grammatical subgrouping of
Formosan languages" in Symposium series of the institute of history and
philology, Academia Sinica #3. But with all my love and respect to Stan
-- I know that you know that this is the case -- his Formosan taxonomy
that is based on lexicase grammar is off the wall, to say the least.
And this is exactly what Blust debunks in his 1997 article (in the
papers for the 8th International Conference on Austronesian Languages,
Taipei December 28-30, 1997) under the subtitle "syntactic non-
evidence". Please read this most important paper. Certainly, Bob Blust
classification is not the only one, but I bet it is the one which is
the most accepted today (with minor variations). As in any field there
are crazy classifications of AN branches -- e.g., the one by Isidore
Dyen -- you are not going to invoke it, right? As you well know, the
only basis for subgrouping is the exclusively shared innovations. Now,
how Formosan languages could possibly represent a single subgroup,
having, e.g.,  different reflexes of PAN *C, and *S? Etc., etc. Also,
how you are going to argue for Atayalic deriving from the same subgroup
as other languages? Etc., etc. The argumenys against Formosan languages
as a single subgroup are really endless -- but we can discuss it at a
greater length, if you wish.


> Malayo-Polynesian, also known as extra-Formosan, is quite a ways
> down on
> the tree in his scheme, and is not even necessary for reconstructing
> proto-Austronesian.

This is an apparent exaggeration: it is not possible to reconstruct a
lot of stuff in PAN without PMP evidence.

>

>
> One more thought on the possible correlation between number of
> branches and
> language family age. If Ainu is the descendant of the language
> originallyspoken in the Japanese archipelago, just as the people
> themselves appear to
> be the descendants of the original inhabitants, then it is a single-
> memberfamily that goes back at least 10,000 years. (Sasha, please
> correct me if
> I'm wrong about the number of members in Ainu--I don't have your
> reconstruction handy.)

Good example, Blaine. It is probably binary: Hokkaidoo-Kuril Ainu vs.
Sakhalin Ainu, but the split again probably occured no earlier than
12th c. C.E. or so, which is 11 thousand years later than the Joomon
culture that is normally associated with Ainu-speaking population. So,
the number of splits really has nothing to do with the age od a given
family.

Cheers,

Sasha


>
> Best wishes,
>
> Blaine Erickson
> erickson at piercingsuit.com

=========================
Alexander Vovin
Associate Professor
East Asian Languages & Literatures
University of Hawaii at Manoa
vovin at hawaii.edu

----- Original Message -----
From: Blaine Erickson <erickson at piercingsuit.com>
Date: Monday, October 7, 2002 1:50 am
Subject: Re: Age of various language families

> ----------------------------Original message------------------------
> ----
> Sasha Vovin brought up some very good points about the non-
correlation
> between primary branches and language family age. I'd just like to
> pointout that although Robert Blust is an outstanding scholar and
> an expert in
> Austronesian linguistics, his account of the branching of
> Austronesian is
> not the only one out there. The late Stanley Starosta, another true
> expertin the field, held that Austronesian had just two primary
> branches, and
> that all the Formosan languages split of from one or another of
these.
> Malayo-Polynesian, also known as extra-Formosan, is quite a ways
> down on
> the tree in his scheme, and is not even necessary for reconstructing
> proto-Austronesian.
>
> I attended a presentation of his on this a few years ago, so my
> memory is a
> little fuzzy on details, but I do remember his data and analysis
> were quite
> convincing. Two branches, or four, or ten, I am in no position to
say.
> Perhaps Austronesian is a family for which we don't yet have a
> consensus on
> the number of primary branches. The age, however, is generally
> agreed to be
> about 6000 years BP.
>
> One more thought on the possible correlation between number of
> branches and
> language family age. If Ainu is the descendant of the language
> originallyspoken in the Japanese archipelago, just as the people
> themselves appear to
> be the descendants of the original inhabitants, then it is a single-
> memberfamily that goes back at least 10,000 years. (Sasha, please
> correct me if
> I'm wrong about the number of members in Ainu--I don't have your
> reconstruction handy.)
>
> Best wishes,
>
> Blaine Erickson
> erickson at piercingsuit.com
>
> At 12:00 AM -0400 02.10.6, vovin at hawaii.edu wrote:
> >----------------------------Original message-----------------------
> -----
> >I couldn't agree more with Jens Rasmussen, Scott DeLancey, and
others
> >who voiced their opposition to the connection between the number of
> >primary branches in a family and its age. First of all, I am sorry
to
> >say, I believe that Johanna Nichols calculations of a number of
> primary>branches in many cases are either fundamentally flawed, or
> are based on
> >some outdated material. Thus, e.g., Japanese (or rather Japonic as
we
> >call it nowadays), certainly is not just one primary branch, but
two,
> >with quite obvious split between Japanese proper and Ryukyuan.
Having
> >just two primary branches, this family is *definitely* much older
> than>Slavic that has three. Austronesian does not have just 4
> branches, as
> >JN asserts, there are 9 *primary* branches on Taiwan alone as
> cogently>demonstrated by Blust 1997, + the Malayo-Polynesian
> branch, altogether
> >totaling 10 primary branches.
>