Re: [tied] tt/st/ss

On 2007-06-07 22:56, Sean Whalen wrote:

>> <ra:dula> and <pendulum> are, in my opinion, new
>> formations, parallel to
>> <te:gula, re:gula> from <tego:, rego:>.
>
> The long V is explained by ?

Lachmann's Law. Influenced by <te:ctum, re:ctum>, with which they were
semantically connected. Note that there is no length in <tragula> (:
<traho:, tractum>, which is not a Lachmann verb. There's an interesting
article by Benedicte Nielsen (2004), where more Latin instrument nouns
in -(u)lum/-(u)la are analysed.

>> In old
>> instrumental nouns the
>> variant *-tro- was preferred after roots containing
>> a liquid,
>
> PIE *trom and *tlos are not variants of one
> morpheme; they show dif. accent and variants of
> e/0-grade in different cases. Since they are similar
> to each other, and rules create irregular changes,
> they often are mixed or one gains full productivity.

This is a strange thing to say. Those accent differences and e/zero
grades can easily result from the fact that *-tlom neuters had *-tlah2
colectives with an accent shift.

> Latin shows late rules changing l-l > l-r, etc., so
> I don't believe you can use this sort of thing to
> prove ra:strum is older than ra:dula.

But in instrument nouns we see the /r/ variant after _both_ /l/ _and_
/r/ in the root, which is different from the usual pattern of liquid
dissimilation in Latin suffixes (but similar to what we see in the
distribution of -t(H)lo- and -t(H)ro- in Greek). <ra:dula> is
transparent, <ra:strum> is not, so the latter is more likely to be older
anyway.

> I believe this is regular from many stems, not just
> dentals (since tl>kl, *spektlo+ > -kkl- > -kl- >
> speculum).

I'll write more about that tomorrow (it's long past bedtime in my time
zone ;-)).

Piotr

Piotr