On Wed, 24 Jan 2001 16:47:51 , "Glen Gordon"
<
glengordon01@...> wrote:
>Whoa! Stop right there. You're honestly saying that Illich-Svitych "sticks
>rigourously to his regularities" and obtains results which "seem rather
>convincing"? Why don't you give us an example of a convincing root by IS?
>The only reason why IS sticks to his "regularities" is because he freely
>puts phonemic variables into his reconstructions as if they were part of a
>mathematical formula. In the end, IS attempted to make regularities out of
>truely irregular connections. Granted he may have identified some
>unavoidable connections involving pronouns, demonstratives and the like but
>he cheated a bunch. I like Bomhard because he doesn't use these variables.
>He just reconstructs the damn root and that's it. No trickery. That way, the
>flaws of a particular reconstruction become a little more obvious.
I'm not sure what you mean here: which "phonemic variables"?
The notation used in Illich-Svitych's Nostratic dictionary is rather
more exact than what is customary in etymological dictionaries. The
conventions are as follows:
? before a reconstruction: in some way doubtful, e.g. difficult
semantics, irregular phonetics, attested in only two branches of
Nostratic not counting attestations in only one subgroup of a branch.
(deskr[iptivnyj].) before a reconstruction: the root may be
onomatopoeic or sound-symbolic.
Segments between / /: this part of the root contains irregularities
(including transposition of segments).
Reconstructed phoneme in capitals: the exact reconstruction of this
segment is not possible, usually because the word is not attested in
the branch or branches that would disambiguate the reflex. For
instance, most fortis velar stops in Illich-Svitych's dictionary are
notated as *K, because those roots are not attested in
Proto-Kartvelian (which would disambiguate velar *k from uvular *q).
Vowels in second syllables are also often notated as A (=*a or *ä), E
(=*ä or *e), I (=*e or *i), O (=*o or *u) or U (=*u or *ü), or even V
(any vowel), because Illich-Svitych's "West-Nostratic" has mostly
ablauted them out of existance, and "East-Nostratic" has
vowel-harmonized away the distinctions. Note that the phonemes in
capitals are *not* irregular: *K is regular (just not attested in PK).
If there are irregular reflexes, the notation is /k/ (or /K/).
No trickery. In this way, the flaws of a particular reconstruction
become a little more obvious.
Bomhard indeed just "reconstructs the damn root". If the PIE reflex
is just in Germanic, this is not marked. If the equation only
involves PIE and PAA, so be it. If the reconstruction is possibly
sound-symbolic, this is not marked. If the root is not reflected in
Kartvelian, Bomhard just reconstructs *k. And Bomhard doesn't have
the problem of second syllables, because his roots are almost
exclusively *CVC-.
=======================
Miguel Carrasquer Vidal
mcv@...