Dolgopolsky's new book

Thu Apr 2 17:00:52 UTC 1998

----------------------------Original message----------------------------
This really belongs on the Nostratic list, but the Nostratic list no
longer exists.

Dolgopolsky's new book on Nostratic has just been published in
Britain.  Here are the details:

Aharon Dolgopolsky (1998), The Nostratic Macrofamily and Linguistic
Palaeontology, Cambridge: The McDonald Institute for Archaeological
Research, ISBN 0-9519420-7-7 (pb), price unknown, 116 + xxii pp.

The book is distributed in the UK by Oxbow Books, Park End Place,
Oxford OX1 1HN, UK; tel (0) or (+44) (1865)-241249; fax (0) or (+44)
(1865)-794449.  In the USA, it is distributed by The David Brown Book
Company, P.O. Box 511, Oakville, CT 06779, USA; tel 860-945-9329; fax
860-945-9468.

The book presents a reconstructed phoneme system for Proto-Nostratic,
the reflexes of the consonants (but not the vowels) in ten major
branches of Nostratic, and a sample of 125 proposed cognate sets.

As the title suggests, the author is largely interested here in
linguistic palaeontology, and he focuses his attention therefore on
putative etyma pertaining to habitat, social organization and material
culture.  He can find no PN words pertaining to agriculture, animal
husbandry, fishing, pottery, or maritime activity, and apparently also
none pertaining to metal use or to any social unit larger than a clan.
He therefore concludes that PN was spoken in the Mesolithic or late
Palaeolithic period -- in other words, no later than 15,000 BC, or
17,000 BP.

Given the words he thinks he finds for animal names, plant names, and
weather phenomena, he further concludes that the PA homeland must
have been subtropical and most likely located in southwest Asia.

The proposed cognate sets await the attention of specialists in the
relevant languages.  Certainly some of the PIE comparanda cited are
new to me, such as PIE *<bhel-> `marten' and *<kat-> `wickerwork,
wattle-fence', but I'm no IEist, and I may just be ignorant.

Dolgopolsky reconstructs 50 consonant phonemes for PN; he doesn't
provide a vowel system, but there appear to be seven distinct vowels
present in his PA reconstructions.  Among the 50 consonants are 35
obstruents, 14 resonants, and /h/, which you can classify to suit
yourself.  There are five contrasting nasals and no fewer than eight
contrasting coronal laterals, which strikes me as rather a lot of
coronal laterals.  There are three series of obstruents: voiced,
voiceless and emphatic.  There are four orders of obstruents (labial,
dental, velar, uvular), plus a further four of coronal affricates and
fricatives (plain, palato-alveolar, alveolo-palatal and lateralized).
All this makes for one hell of an obstruent system.

The book is typographically a little challenging, in spite of
obviously painstaking efforts, and it has been poorly proofread:
typos are frequent.

This is clearly only a taster toward the author's projected (and
mammoth) Nostratic Dictionary, which we are told will contain around
2000 sets of comparisons.  But it's enough to give the flavor of the
author's Nostratic work, and I'd be interested to hear any comments
on the quality of the reconstructions offered for the six families
assigned to Nostratic -- especially since a number of them appear to
be based on "reaching down" into the daughter languages to find
comparanda.  That is, there is a good deal of inverted reconstruction
(top-down reconstruction) along the following lines: "this word I
found in one Cushitic language must be assignable to Proto-
Afro-Asiatic, because a suitable PAA reconstruction for it would
match the PN reconstruction I've arrived at on the basis of IE and
Tungusic".  Inverted reconstruction is all very well in the case of a
secure family, but it makes me nervous seeing it employed to set up a
family in the first place.

Just to cite an example, item number 2, for `hyena', claims reflexes
in Afro-Asiatic, Altaic and Dravidian, but the claimed Afro-Asiatic
form is found nowhere but in Semitic, while the claimed Altaic form is
found nowhere but in a single Tungusic language.  Not overwhelming.

Moreover, I am a little disturbed by the very monotony of the reflexes
of the PN obstruents (and indeed most other consonants) claimed for
the assorted daughter languages.  In spite of the truly vast
time-depth claimed by Dolgopolsky, labial plosives just remain labial
plosives practically everywhere, sibilants remain sibilants, */l/
remains /l/, */n/ remains /n/, and so on.  It would appear that the
daughters of Proto-Nostratic exhibited a singular reluctance to
undergo any interesting phonological changes of the sort found in most
other languages, even though we are talking about more than ten
millennia from PN to (say) PIE.

On the plus side, Dolgopolsky tries to be scrupulous in setting up
systematic phoneme correspondences, though it's a little disconcerting
that the first entry in the book, detected only in Afro-Asiatic and in
Dravidian, fails to exhibit the claimed Dravidian reflex of *<-b->.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
England

larryt at cogs.susx.ac.uk