Re: [tied] Seven, eight [was: gWerh3- "to devour"]

From: Miguel Carrasquer
Message: 15676
Date: 2002-09-22

On Sun, 22 Sep 2002 15:35:23 +0200, Piotr Gasiorowski
<piotr.gasiorowski@...> wrote:

>----- Original Message -----
>From: Miguel Carrasquer
>Sent: Sunday, September 22, 2002 3:21 PM
>Subject: Re: [tied] Seven, eight [was: gWerh3- "to devour"]
>>> ... the Slavic word shows no Winterian effect of *d (**se^dmU).
>> But Winter's law is blocked when a resonant (at least *n, *r, *l) follows.
>I don't believe it. We have *ablUko 'apple', *e^mI 'I eat' < *e^dmI (plus *e^dlU), *vydra 'otter' and *agne~t- 'lamb', to quote just the examples that occured to me at once.

The counterexamples to Winter's law with a resonant following *b/*d/*g as listed
in Rasmussen's article "Winter's Law of Balto-Slavic lengthening" are:

Slav. ognI, Lith ugnìs
Slav. o~glU, Lith. anglìs
Lith. gi~edras, gaidrùs
Latv. idrs (but Lith. ýda)
Lith. lùgnas
lith. slãbnas
Lith. vìglas
Slav. dUbrU, Lith. dubùs
BSl. *pigrás, Lith. pigùs

One might add (cf. dubùs, pigùs):
Slav. voda (if vowel from *wodr)
Lith. smagùs, Latv. smagrs

As to the counterexamples to the counterexamples you give above:
Slav. ablUko, but Baltic *a:bo:l-, with a vowel between *b and the resonant.
Slav. êmI, êdlU can easily be analogical (or the blocking did not work across
grammatical morphemes).

The "otter" and "lamb" words are indeed good counterexamples, which can oly be
explained away in ad-hoc manner.

Miguel Carrasquer Vidal