From: Sean Whalen
Message: 45161
Date: 2006-06-28
> Sean Whalen <stlatos@...> wrote:It only dissimilates before w or u; otherwise
>
> --- Andrew Jarrette <anjarrette@...> wrote:
>
> > If /a/ in <tardus> is not derived
> > from an original *e, then what is the origin of
> /a/
> > in this Latin word?
>
> There are cases of *wr. > ar before w or u at a
> certain stage:
>
> arduus "steep"
> u:rdhva- "upright" Skt
>
> arvum "arable land"
> urvara:- "field" Skt
> ________________________
> Although the evidence is undeniable, what a weird
> change this is. You'd think the *w preceding the
> *r. would cause rounding and raising, not spreading
> and lowering.
> Also odd to me that Latin has <ar>Not every case of lengthening is caused by a
> in <arduus> and not *<ra:duus>, as I would expect to
> correspond to Sanskrit < u:rdhvam>.
> If tardus is from PIE, a very speculative derivationSo far I've been very speculative; but to go even
> could be
>
> turdu-
> twr.du-
> tardu-
> _______________________
> Then what root is this from? *twer-? *twerd-?
> What meaning would this root have? Could it be
> related to the *twer- root that has meanings such as
> "enclose" and "shut off" in Slavic, Baltic, and
> possibly Latin (<paries>)? Just wondering.