Dolgopolsky's new book

[Larry Trask]

----------------------------Original message----------------------------
Alexis M R writes:
 
[My first attempt at replying to this was interrupted by two crashes
in rapid succession: the first of thunder, the second of our system.
Here goes again.]
 
> The PIE *kat 'wattlework' which Larry says Dolgopolsky has is
> clearly the same etymon as Illich-Svitych posited in the form *ket,
> and I do not understand how the vowel could be given as PIE *-a-.
 
Yes, I'm now satisfied that PIE *<ket-> is the intended form here.
Possibly this is a typo (recall that the book is full of typos), but I
don't think so.  The PN reconstruction is *<kadV->, with an emphatic
/k/, and all the comparanda except the Kartvelian one have /a/.
 
Elsewhere, PN */e/ is continued as PIE */e/, but in fact there is only
one relevant form in the book.  The small number of comparisons on
offer here makes it difficult for the reader to check the consistency
of proposed developments in the daughter languages.
 
Also a problem is that only ten of the 125 PN reconstructions on offer
are of definite form.  All the others contain one or more instances of
generic segments, optional segments, or fluctuation between segments.
 
> This is in fact supposed to be one of the many cases where Nostratic
> *a gives PIE *e (alternating with *o).  However, this is a PIE
> etymon which is not traditional with IEnists, and was put together
> this way by Illich-Svitych himself, I believe.  This does not mean
> that it is wrong, of course, and IS makes it clear that this is a
> new proposal.
 
Dolgopolsky gives no clues as to which of his proposals follow I-S and
which are new with him.
 
> In general, although etyma that are not widely attested are
> quite often posited, I tend to agree with Larry's skepticism,
> and as I have argued elsewhere I think that AT BEST a very
> small percentage of the etyma posited for Nostratic by IS and
> Dolgopolsky will stand the test of time.  Of course,
> as it happens *ket is fairly widely attested in IE, and I
> tend to think this was one of IS's better ideas.
 
While I'm here, I might add that Dolgopolsky's PIE *<bhel-> `marten'
is supported only by the Welsh word cited by Richard Coates and Latin
<fe:le:s> `wild cat, marten, polecat>.  Buck is not certain these
words are related, and I have yet to check Pokorny.
 
Also, I note that Dolgopolsky cites Sanskrit <simha-> `lion' and
Armenian <inj> `leopard', plus Tocharian <sisak> `lion', to justify a
supposed PIE *<singho-> `lion, leopard'.  It's news to me that the
Sanskrit/Armenian link is generally accepted, though I have certainly
seen it mooted, and the Tocharian form is new to me (and not obviously
convincing).
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk