Re: [tied] kinship systems

--- Glen Gordon <glengordon01@...> wrote:

> For instance, Bomhard has four
> reconstructions for the first person alone - *mi, *na, *?a and *wa!
> And my
> attempts at bringing order to the chaos by proposing an
> ergative/absolutive
> suppletive pronominal system, which would reduce things to only the
> ergative
> 1ps *nu (yeilding later forms in both *m- and *n-) and the absolutive
> 1ps *u
> (yielding forms in both *w- and *?-), have been ignored so far
> (Boohoo...)

Four reconstructions from six families equals major doubt.

> As for kinship terms, basically Bomhard has:
>
> *?ab- father
> *?at(t)- father
> *?am(m)- mother
> *?an'- mother, aunt
> *?ay(y)- mother, female relative
> *?ak(k)- older female relative
> *xaw- a maternal relative

Bear in mind that some of these are "nursery words". In particular,
any of the possible "double consonant" roots. I'd eliminate these as
Nostratic cognates since nursery words are pretty universal. (And
vague. For instance, Georgian has _mama_ for "father"!)

> Looking at *?an'-, something looks fishy to me. It's based on:
>
> Uralic *an'a "mother, aunt"
> Dravidian *an.n.- "a woman, mother"
> Altaic:Turkish ana "mother"
> AfroAs:PSC *?aN- "father's sister" (N = ing)

Depends on how stable words for relatives are. (And what kind of
social structure the member families constitute. You have matrilinear
and patrilinear societies mixed together in these.)

> Uralic, Dravidian and Altaic are part of the Eurasiatic subbranch
> while
> AfroAsiatic would only be remotely related. The fact that only one
> branch of
> this group is attested makes me suspicious.

Some are rethinking Afro-Asiatic altogether. (I consider it a member
family, but like you said, one in similar relation as Hittite is to
Indo-European.)

By the way -- what is the status of Nilo-Saharan? I've heard it linked
with Niger-Congo, and less commonly, Afro-Asiatic. (The Nubians live
south of Egypt and north of Ethiopia, after all.)

> Perhaps we could reconstruct the following diagram until somebody
> thinks up
> something better:
>
> *aba === *aka *aba === *aka
> (g-fa) | (g-mo) (g-fa) | (g-mo)
> | |
> |-----------| |-----------|
> | | | |
> *ahwi *ata === *ama *aya
> (fa-br/fa-si) (fa) | (mo) (mo-br/mo-si)
> |
> *u
> (EGO)

Hey, I like that! I'll note it.

> ...which would look kinda Eskimo-ish or something. I don't there's
> any Omaha
> pattern here, but I could be wrong.

Some link Eskimo-Aleut to Eurasian, don't they?

117.

__________________________________________________
Do You Yahoo!?
Get Yahoo! Mail - Free email you can access from anywhere!
http://mail.yahoo.com/