Re: "u" versus "a"

--- In cybalist@yahoogroups.com, "alex" <alxmoeller@...> wrote:

> I got intrigued by the "u" in the Latin word "insulto" which is

considered

> to be a derivat from IE *sel, as well as "salire", "saltare".
>
> How is in fact to explain the change of the vowel in the root here? My
> Walde-Hoffmann dictionary does not give any explanation and when

one take a

> look at "salire" there is no mention about "insulto". My little dictionary
> gives the "i:nsulto:" but there is no explanation about the change

of "a" to

> "u". Or do we have to consider the mention that "l" was influenced

by "e"

> which followed and thus the previous "e" or"a" became an "o" or "u"?
> At least that should be the explanation cf. my little dictionary for
> explaining the "u" in "-sulto".
> The examples given there are Latin "oliva" from Greek "elai(f)a" >

olaiva >

> oliva or the one of "homo" which originaly was an "hemo", bonus

versus bene

> or famulos versus familia.
>
> The "change" appears not very clear at all; the explanation here

appears a

> bit forced to me specialy when we do know about the alternance "e"

versus

> "o" in IE roots. Can it be in fact we have to deal with reflexes of "o"
> grade here where we have "o" or "u" instead of usual "e", "a", "i"

which are

> supposed to ve reflexes of "e" grade of the root?
>
> Anyway, it appears curious this alternance in IE of e/o with a _more_

later

> alternance e/o this time considered just as a phonetic evolution.

The conditioned changes el > ol > ul can be seen in the single verb
volo: 'wish':

Infinitive velle < *vel-se

Pres. ind. 1s: volo:
Pres. ind. 2s: vi:s (this seems totally irregular to me!)
Pres. ind. 3s: vult
Pres. ind. 1pl: volumus
Pres. ind. 2pl: vultis
Pres. ind. 3pl: volunt

Impf ind. 1s: vole:bam

Pres subj. 1s: velim
Pres subj. 2s: veli:s

As I recall, the rules are:
1. el > ol except before l, i, i: (Possibly also not before short /e/)
2. ol > ul before any consonant but /l/.

As PIE l. > ol in Latin, the ablaut is very much obscured before /l/.

AS to PIE *sel, I must admit I don't see much evidence. Apart from
Latin sulta:re and OCS vúsle^pljo, all I see is evidence for PIE *sal.
The Slavonic could be secondary ablaut, but I don't know any
explanation for Latin sulta:re from PIE *sal, or Latin sali:re and the
Celtic forms in -a- from PIE *sel. I too will be interested to see the
explanations.

Richard.