LINGUIST List 5.992

Fri 16 Sep 1994

Disc: Altaic

Editor for this issue: <>


  1. Matthew Dryer, Binary Comparison
  2. , Altaic (End of the Controversy)

Message 1: Binary Comparison

Date: Thu, 08 Sep 1994 06:30:42 Binary Comparison
From: Matthew Dryer <>
Subject: Binary Comparison

Last week, Richard Alderson asked the following questions, which I think have
only been partially answered in the discussion on LINGUIST: "Who are these
linguists whom G & R see as the enemy and the anti-Altaicists see as supports?
And how have they concluded that a binary comparison is reasonable?"

Unlike the situation with Altaic, where some people have specifically advocated
examining only two languages at a time, the situation in the Americas is simply
one of actual practice rather than ideology. Namely, the earlier literature on
the classification of languages in the Americas is full of instances of binary
comparisons, either between two languages (cf. Silver 1964 on Shasta and Karok;
Jacobsen 1958 on Washo and Karok), or between one language and a postulated
group (cf. Haas 1965 on Kutenai and Algonquian; Newman 1965 on Zuni and
California Penutian; Radin 1916 on Huave and Mixe), or between two postulated
groups (cf. Haas 1958 on Algonquian and Gulf; Freeland 1930 on Mixe and
Penutian; Olson 1964, 1965 on Chipaya and Mayan) (all references here from the
bibliography to Greenberg's book). In none of these cases were people
advocating binary comparison; rather, it was the case that they simply had
access to or expertise in the pairs in question. It is this binary tradition
in the Americas that Greenberg (and Ruhlen) have been critical of. Their
argument is that such comparisons are analogous to comparing Swedish and
Bulgarian, or Swedish and South Slavic, or North Germanic and South Slavic: the
relevance of such pairs only makes sense within the context of comparison of
Indo-European languages as a whole. From the perspective of Greenberg's
classification, a number of these comparisons have led to classificatory
errors, such as Chipayan as a branch of Penutian, or pseudo-groupings, like
so-called Macro-Algonquian (consisting of Algonquian-Ritwan and Gulf), both of
which made their way into Voegelin and Voegelin (1977) and the latter of which
is still cited occasionally in typological studies as the family to which some
language belongs. Greenberg's argument is that the resemblances noted in these
binary comparisons may be valid but they may only reflect higher level
groupings to which the items compared ultimately belong but that in at least
the two cases just noted do not reflect genetic groups.

While arguments can be given that the comparisons in question are not quite
analogous to hypothetical cases like Swedish and South Slavic, the main point
is that the practice of binary comparisons in the Americas that Greenberg is
critical of is quite distinct from the ideology of binary comparison apparently
espoused by some anti-Altaicists. In fact, I doubt that few if any of
Greenberg's detractors today would argue with Greenberg's point about binary
comparison, particularly as it applies to Chipayan as a branch of Penutian or
"Macro-Algonquian" since neither of these are hypotheses that anybody that I am
aware of still believes. Their primary argument with multilateral comparison
is not whether it is necessary, but whether it is sufficient.

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

Message 2: Altaic (End of the Controversy)

Date: Thu, 8 Sep 94 21:24:23 EDTAltaic (End of the Controversy)
From: <>
Subject: Altaic (End of the Controversy)

In reviewing the discussion on Altaic, I just noticed that
no one seems to have responded to a question posed early
on by John Koontz, who points out that Comrie supposedly says
somewhere that Turkic, Mongolic, and (Manchu-)Tungusic do
not form a unit but may perhaps belong together with IE,
Uralic, etc., in a grouping larger than Altaic, namely, Nostratic.

I have not read the relevant passage of Comrie's, but if he
meant thereby to deny the validity of the Altaic reconstruction,
then I think he was wrong. On the other hand, as noted by Vovin
in our discussion, the true Altaic situation seems to be that
Turkic, Mongolic, and (Manchu-)Tungusic do nOT form a unit
on their own but only with Japanese and Korean.

As to the relation to Nostratic, the main advocate of this
theory, Vladimir M. Illich-Svitych, did express some doubts
about Altaic as a unit himself, suggesting that perhaps
Turkic, Mongolic, and (Manchu-)Tungusic were daugthers of
Nostratic. This was because his reconstrtructions of
Altaiuc and Nostratic were very similar.

However, if we accept the new look of Nostratic proposed
in Manaster Ramer (in press), then we ipso facto must
assume that Nostratic and Altaic phonologies were different
(Nostratic would have had initial clusters which survived in
Proto-IE and Proto-Kartvelian but were reduced elsewhere,
incl. Proto-Altaic), and I think other arguments for Altaic
as a valid intermediate level between Nostratic and the
more recent languages can be found. There seem, for example,
to be numerous Altaic etyma which probably do not goback to

However, it is perfectly possible for an older stage in
a linguistic prehistory to be more certain than a more
recent one, and so I would not be unwilling to say that
Nostratic is a more certain grouping than Altaic, although
in reality I don't think that this is the case. (In case
this sounds paradoxical, consider that Indo-European is
a more certain grouping than are most of the proposed
groupings immediately below it, such as Balto-Slavic).

Having said this, I think that there is little more to
say about the Altaic hypothesis. While much remains
to be done, I was pleased that no one (on LINGUIST anyway)
seems to want to defend the tired old pseudo-arguments
against it, so the discussion would seem to be over. I
do believe, as I pointed out some time ago and as John Koonts
(sorry, Koontz) also seems to suggest, that there is a
connection with Nostratic in the sense that the same
people tend to object to both theories, and I of course
would be open to a discussion of Nostratic--about which
even stranger tales are told than about Altaic.
Manaster Ramer, A. in press. Clusters or affricates in
Nostratic and Kartvelian? Diachronica.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue