LINGUIST List 5.910

Sat 20 Aug 1994

Disc: Altaic

Editor for this issue: <>


  1. , Altaic
  2. John E. Koontz, Re: 5.905 Altaic
  3. Scott C DeLancey, Re: 5.908 Altaic
  4. Ecological Linguistics,Anderson,PRT, Re: 5.908 Altaic

Message 1: Altaic

Date: Fri, 19 Aug 1994 10:21:24 Altaic
Subject: Altaic

A.M.Ramis' recent posting concerning Altaic has piqued my interest. Most
of the what I have read about Altaic is at least 20 years old. I am looking
for references to the proposed Japanese/Korean relationship to Altaic. Poppe's
work was I thought particularly convincing for the body of Altaic but the
little that I had seen for these two was less so. I am also interested in
work that has been done on external relations for Ainu, other than Patrie
(Roger) Brad Coon "Lions are basically
 Hans Kruuk

Kill a lion, save a hyaena!
Boycott Disney lionist propaganda!
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 5.905 Altaic

Date: Fri, 19 Aug 1994 09:57:10 Re: 5.905 Altaic
From: John E. Koontz <>
Subject: Re: 5.905 Altaic

> From:
> Subject: Altaic

> Thus, I think we need to discuss two separate questions. One is what is
> the true state of the scholarship (both pro and con) on the Altaic
> question (as opposed to the nonsense that has been going around).

I believe that Bernard Comrie, in his <Languages of the USSR>, suggested that
Altaic was not a family, but that its proposed components might well belong
with, e.g., IE, Uralic, etc., in some higher entity. While this is
certainly one logical subgrouping, and different per se from a hypothesis of
Altaic unity within such a higher entity, it struck me as a way of denying
the Altaic hypothesis without denying it, by making it dependent on a more
controversial hypothesis. You might want to address this approach in your
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Re: 5.908 Altaic

Date: Fri, 19 Aug 1994 15:00:28 Re: 5.908 Altaic
From: Scott C DeLancey <>
Subject: Re: 5.908 Altaic

Reinhard (Ron) F. Hahn <> lists as among the
perhaps inappropriate arguments that have been raised against the
Altaic hypothesis the "Absence or scarcity of cognates among numeral",
and notes:

> This poses questions regarding the universal validity of
> numerals as mandatory indicators of genealogical affinity. The
> fact that numerals seem to supply evidence for the
> Indo-European hypothesis ought not lead us to assume that this
> applies universally.

This an important point. It simply is not true that numerals are
particularly stable vocabulary items around the globe. In North
America, for example, they seem to be readily subject to both replacement
and borrowing.

Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 4: Re: 5.908 Altaic

Date: 20 Aug 94 14:25 GMT
From: Ecological Linguistics,Anderson,PRT <>
Subject: Re: 5.908 Altaic

Thanks now to Rheinhard (Ron) F. Hahn and to Harold Schiffman for further
commentary on Altaic.

I do not assume the family-tree method is the norm for historical linguistics.
It tends to be more valid when the descendents are few and the time depth is
great. But if **either** of those criteria is not met, then the descendents
may more often than is believed form a network reflecting an original language
**area** as it first was differentiated by waves of innovations spreading from
different centers and overlapping in complex ways, thus leaving a dialect chain
or more appropriately even, a dialect network.

Now this kind of data distribution is conceptually not too far from the
opposite pole, waves of borrowings, except that borrowing implies languages
which were originally unrelated (or at least unrelated back to a very
substantial time depth before the borrowings took place), while the dialect
network model implies an original sameness later differentiated regionally.
These are in principle distinguishable.

The bottom line for me is that the family-tree is usually much too strong a
kind of claim to make, recklessly so. However, I find many historical
linguists who think that if one is not making such an explicit tree claim, one
is not doing proper historical linguistics. There are just more subtle ways of
handling data now.

For this reason I would doubt the need of setting up a framework for discussion
which assumes from the outset that Korean is at a more distant remove, to be
related to Altaic itself (Mongolic-Tungusic-Turkic) rather than as a branch
within Altaic. Also, although "Tungusic" makes sense, the older Manchu-Tungus
did indirectly and incompletely recognize a regional grouping into southern
Tungusic and northern Tungusic? I do not care about the terminology (much),
but want to avoid artificially isolating groups (Tungusic?) when perhaps Korean
might be a much changed distant outlier of southern Tungusic?

For those interested, I have about 8 copies of a 43-page paper, "Sound change
and pitch-accent systems in Korean dialects: two results of original
differences in *stress-accents". It appeared in Chin-Wu Kim, ed.: Papers in
Korean Lingusitics (1978). Available free for the mailing cost, send $1.90 in
stamps along with a stick-on address label to the address below.

On the pronominal system, this note from Ron Hahn:
>The Turkic, Mongolic and Tungusic pronominal systems bear astonishing
>resemblance, which enables us to reconstruct a proto-Altaic system.
>This has been dismissed as irrelevant, since this proto-Altaic system
>bears great resemblance not only to the proto-Uralic one but also to
>the Indo-European one. (Obviously, the anti-Altaists are among those
>least likely to accept the possibility of even earlier genealogical
>affinity, "Nostratic" or otherwise.)

When I have asked some honest linguists about this pronominal question, they
have essentially said "the field is not yet ready to deal with" these
excessively strong resemblances. I guess implying is it sound symbolism or
chance or historical relationship and we can't figure out which?

For a small amount of analysis of Uralic pronominal systems which seems to
imply Indo-European relations to Indo-European and Altaic, please see my 1975
article "Grammar-meaning universals and proto-language reconstruction" in
Papers from the Eleventh Regional Meeting, Chicago Linguistic Society. Section
6 of that paper, examples (25) and (26) and (24). I still believe some portion
of what is in that article, including much of the northern Eurasian pronominal

Especially core to the article is the idea that historical recontruction
depends crucially on what are typologically normal paths of historical change,
and the accumulation of typological universals of paths of historical change
equally depends crucially on solid cases of historical reconstruction. The two
are built together. (That is not to deny that abnormal cases of change do
occur, of course.)

Because so many historical linguistics have a visceral aversion to typology
(because it has been misused to assert historical relation based merely on
typological similarity), they find it I think difficult to remember and to
acknowledge that much of the plausibility of hypotheses of historical
relatedness comes from the plausibility (i.e. the typological normalcy) of the
paths of change and intermediate stages hypothesized to get from posited
ancestors to each descendent. The other basis for the plausibility of
hypotheses of historical relatedness comes from having actual data on
intermediate stages, when such is available. At greater time depths, less such
data is available, for which reason those who focus exclusively on it claim
reconstruction becomes less possible, while those of us who recognize the
double base of historical hypotheses then rely more on typologically normal
paths of change.


For some mention of Altaic numerals (Mongolic and Tungusic, I think I did not
attempt to link Turkic numerals!) in the context of a much fuller analysis of
the numeral systems of Northeastern Asia (Luorawetlan, Chukchi-Koryak-Alyutor,
Kamchadal, Chuvan, etc.), please see my 1983 "Number Words in Northeastern
Asia", Proceedings of the Second Conference on the Non-Slavic Languages of the
USSR, pp.27-65.

Copies of this paper are available to anyone who requests it and sends a
self-addressed stamped envelope (52 cents stamps) to

Ecological Linguistics
P.O. Box 15156
Washington, D.C., 20003

I should mention the most iffy part of the numerals paper is the attempt to get
information out of Pallas' 1787 recordings of old Japanese, because some of the
variation may be matters of faulty interpretation of original handwriting when
the material was transcribed in older times from handwriting into cyrillic
print (so Sam Martin's view if I remember correctly).

Lloyd Anderson

Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue