Re: [tied] Tocharian dentals

I follow Winter's position as laid out in "Vertretung indogermanischer Dentale im Tocharishen" (IF 67, 1965) and summarised in "Armenian, Tocharian, and the 'glottalic' theory" (1992). I find his examples compelling. For the sake of those Cybalist members who know little about Tocharian, here's the gist of the regular changes (all the examples are from Tocharian B):

(1) *dH and *t merge as t, giving c before a palatalising vowel and ts when followed by *[j]

mit 'honey' < *medHu

trai 'three' < *trejes

tkacer 'daughter' < *dHugh2te:r

pa:cer 'father' < *ph2te:r

lac 'has left' < *h1leudH-e-t

petso 'lady' < *potja:- < *pot-i-h2-

(2) *d is lost before resonants, gives s' when palatalised, ts elsewhere

wi 'two' < *dw-, cf. twere 'door' < *dHw-, twe 'thou' < *tw-

s'ak 'ten' < *dek^m

tsa:ka- 'bite' < *dak- (Gk. dakno:)

(3) In the "Grassmann environment" (initially in a root containing another aspirate) *dH- > *d-

> ts-

tsäk- 'burn' < *dH..gWH- 'burn'

(This verb belongs to a conjugation [Krause's Class VIII] that shows no palatalisation of root-initial consonants, i.e. no inherited e-grade.)

Piotr

----- Original Message -----
From: Miguel Carrasquer Vidal

To: cybalist@egroups.com

Sent: Thursday, December 14, 2000 4:59 PM

Subject: Re: [tied] Tocharian dentals

On Thu, 14 Dec 2000 15:06:08 +0100, "Piotr Gasiorowski" <gpiotr@...> wrote: >I think that's wrong. Before PIE *e we should get Proto-Tocharian palatalised *s' (as in the numeral "ten", TB s'ak). I'd expect the following development (more or less): > >*deiwo- > *s'@ywV- > TA *s'ew, TB *s'aiwe You're right about the development PToch *äywe > A <ew>, B. <aiwe> (I had mistakenly looked at the development of *new(y)os > ñu/ñuwe). For the initial consonant, I was following Adams, more or less, but the question is a thorny one. Adams posits a First Palatalization, which gives *d > ts, as in TochAB tsär- "to separate" < *der-. There is, however, a Second Palatalization, which also affects the result of the First Palatalization, giving PToch *ts', which subsequently gives TochA. <s'>, TochB. <ts> (TochA tsäm- ~ s'äm- "to cause to grow" < *dem-, TochA tsär- ~ s'är- "to separate"). Sometimes, however, the effect of the First Palatalization was undone analogically, restoring *t, which under the Second Palatalization gives *c, and then, according to Adams, then sometimes shows up "under certain, not always clear, conditions" as <s'>, as in the word for "10" (TochA. <s'äk>, B. <s'ak>). On the other hand, Werner Winter's position is that there are no two separate palatalizations, but that *t and *dh give <t>, <c> (palatalized) [except that *ty > ts], while *d (including *dh before voiced aspirate, in a Tocharian version of Grassmann's Law) gives unpalatalized <ts> (or zero), palatalized <s'> (but often <ts> in Toch.B). Adams, in a footnote, also summarizes the positions of v. Windekens (*d palatalizes as ts, *t and *dh as c), and Anreiter (*ty and *dhy give ts, *t, *d and *dh before front vowel always give c). There is also Evangelisti's position, discussed by Werner, that *t palatalizes as c, *d as s' and *dh as ts. ======================= Miguel Carrasquer Vidal mcv@...