Re: [tied] Re: The trouble with *h3

On 04-10-19 16:53, Rob wrote:

> 1. Which IE dialects have the *abon- word?

Celtic and Italic (Lat. amnis < *(h2)ab-(h3)n-), and the
(Italo-Celtoid?) hydronymic substrate in Germany.

> 2. Aside from (Vedic?) Sanskrit (pibati) and Latin (bibet), which IE
> dialects have voicing in their forms for *peh3- 'drink'?

Celtic has it too, cf. OIr. ibid < *pib(h3)-e-ti.

> That said, I think that it seems likely for the *-h3on suffix to be
> related to the root *h3en(h2)- 'carry (a burden)'.

Or maybe 'charge, load, burden (with sth.)'. The verb itself is only
weakly attested, and the most common derivative is the -es-stem noun
*h3onh2-os 'load, burden' (Lat. onus, Skt. anas-)

> If this was
> indeed the case, then we have a quite nice compound for a 'river'
> word: *h2ap-h3on- 'water-carrier' > 'river'.

This particular one seems to have been an endocentric compound in which
*-h3on- means 'a load, a bundle, a lot of', i.e. *h2ap-h3ón- 'a lot of
water' or the like. According to Olsen, this type contrasts with
substantivised adjectives such as *h2jú-h3on- 'having a lot of vital
force'. But the "suffix" (in fact, a root noun) is the same in both cases.

> Since noun-verb
> compounds (I forget their technical term, something from Sanskrit)
> were productive in IE, such a compound seems likely.
>
> On another note, what do you think the *phonetic* realization of /h3/
> was? If the voicing effects noted above were in fact the result
> of /h3/, then it was indeed voiced. However, this does not say what
> its articulatory position was. I think, based on the fact that it
> backed an adjacent vowel, that /h3/ was a pharyngeal fricative.

It could be. I remain agnostic as regards the exact point of
articulation (which may have varied diachronically anyway).

PIotr