--- In
cybalist@yahoogroups.com, enlil@... wrote:
> Miguel:
> > The EA 1st and 2nd p. sg. markers, as well as the reflexive,
> > are voiceless 1. *-k > -ng, 2. *-t > -n, 3R. *-c > -ñ, while
> > other suffixes, such as the dual and plural are voiced (du.
> > *-g > -k, pl. *-d > -t).
>
> My spidey senses are telling me that this supposed EA
reconstruction
> is flawed. The "voiceless" markers would appear to be in reality
> nasals while the "voiced" markers are actually voiceless to begin
> with. In that way we have Boreal 1ps *-uN equating properly with
> former *-ux elsewhere, 2ps *-un with the ProtoSteppe 2ps
intransitive
> *-un and then the plural, being in reality *-t, is from Steppe *-
it.
> Finally, Boreal languages gained a dual suffix using the numeral
> 'two' (presumably *kaN) in a similar way to Icelandic /-t/ but this
> is not found outside of Boreal so let's just accept it and move on.
>
> The whole is that there was _no_ voicing in Boreal.
Miguel is of course fully capable of replying by himself, but I
would like to add a comment. The Eskimo Auslautgesetze were long a
matter of some tacit controversy. I say "tacit" because the matter
was for a long time not really addressed, and most scholars just
went on using the word-final values as if they were underlying. The
correct analysis was shown by Swadesh in 1951 and soon adopted by
Bergsland, but it is only by being used by the Comparative
Dictionary by Fortescue-Jacobson-Kaplan (Fairbanks 1994) that it has
suddenly been canonicized. In the meantime most scholars thought up
phony rules for "q-deletion" for all of Eskimo and "k-deletion" for
the East, rules that were found to operate nowhere else in the
language. That all disappears when the underlying stem-finals are
posited as spirants, for there are rules governing loss of spirant R
and g (gamma) already. Since the far East, meaning Greenland, has no
opposition between nasals and stops in final position, it did not
much matter that the 2sg possessive -t was -n farther to the West.
That probably spared it the fate of having ever been made underlying
(as you are now apparently the first to propose). Still, the
internal morphophonemics is very clear, showing that, in the
interior of words where oppositions are many, the word-final stops
behave as (voiced) spirants, while *some* cases of word-final nasals
are oral (voiceless) stops when they appear inside a word. I would
guess that the genitive/ergative marker *-m really is a nasal, but
it is no so easy to show that is also nasal when word-internal.
However, it is now a generally recognized analysis that the person
markers *-ng, *-n are from *-k, *-t, and that the number markers *-
k, *-t are from spirantic *-g, *-d (gamma and delta). These changes
are demanded by the morphophonemics of the language itself, and you
get large parts of the grammar regular if you accept them. That's
why I did that in my 1979 book on the matter, adding the derivation
of the reflexive *-ni from *-ñ from an older palatal stop *-c. Stems
that end in *-g or consonant clusters insert an anaptyctic vowel
before a consonantal ending, as *iñug-¤t > West Eskimo iug&t
(singular iuk, yuk), East Eskimo inuit 'people'. The extra-vowel
inflections have spread to all nouns in most of the Canadian
territory, a fact that fooled some for a long time, so that the
endings were posited as *-&m, *-&t rather than plain *-m, *-t (older
*-d). My book was in the main an attack on that extra vowel which
was associated with some very high-prestige names of the field.
Nobody has expressed his acceptance of my analysis, but I find it
used in the Comparative Dictionary. Torsten told me recently that
scholars' reactions vary from "you're silly, that's plain wrong"
over "you're silly, we've been over that already" to "you're silly,
we know that already". I find comfort in those words, I've been
through it all.
Jens