On 05-02-28 17:27, tgpedersen wrote:
> Kuhn has an article on the alternation -tt- ~ -ss- in Nordwestblock,
> cf hatt- ~ Latin cassis "helmet", Chatti ~ Hessen. The
> -t- + -t > -ss- in Germanic, Italic and Celtic ppp.'s could be the
> result of a rule -tt- > -ss- in a substrate language, and -tt- the
> unchanged variant, hence the mixture of forms. Cf Latin matta and
> massa (with -a-), and Dutch mes "knife".
Any intervocalic -tt-'s that appear in early Germanic in words that are
not obviously loanwords represent either irregular expressive gemination
(as in Goth. atta 'father' or in truncated personal names like Otto <
*auda-[whatever]), or the result of nasal assimilation after short
vowels, where *-tn-, *-dn- and *-dHn- all end up as -tt- (similarly for
other stops). The precise etymology of tribal names like that of the
Chatti is inherently uncertain, but supposingthat they are "the Hats"
and that, as Pokorny has it, we are dealing with derivatives of
*ka(:)dH- (perh. *k(a)h2-dH-, cf. <hood>) 'cover', a *-dHn- should be
assumed here (while Latin cassis < *kadH-ti-). Hessen is a re-Germanised
Latin word (<populus Hassiorum> in the 8th century). Isn't <hass->
simply a Latin rendering of Old Rhein Franconian *hazz- < *xatt-?
Piotr