Re: [tied] tt/st/ss

On 2007-06-13 19:39, stlatos wrote:

> No syllabic C is possible for these, therefore those in which you
> say it could be are probably the result of met. (creating more common
> forms like nl, rl as *dwenlos > bellus, *sterla: > stella).
>
> pultlos
> pulslos
> pulsilos
> pusillos
>
> analogy:
>
> putlo+lo+ > putillus

The usual analysis is:

*putslo- > pullus
*putsle-lo- > *pu(t)sl.lo- > pusillus

> etc., so:
>
> wekslo+lo+ .. tegno+lo+ .. po:klo+lo+
> weksl+lo+ ... tegn+lo+ ... po:kl+lo+
> weksl+ilo+ .. tegn+ilo+ .. po:kl+ilo+
> wekslilo+ ... tegnilo+ ... po:klilo+
> weksillo+ ... teginlo+ ... po:killo+

That's my view too (and communis opinio, as far as I know), if one skips
stages III and IV as superfluous. What really matters is that stems with
-r- as the last consonant have regular diminutives in -ell- (like
capella), and those with -l- and -n- have regular diminutives in -ill-
(like vexillum and pugillus) if the vowel is epenthetic.

Piotr