From: dgkilday57
Message: 63582
Date: 2009-03-05
>The root of the 'icon' word must be *weik- or *weik^-, since the Cypriot acc. sg. is <weikona>, and Homeric scansion requires a lost digamma; thus <eoika> must have been *wewoika. Frisk says there are no secure cognates outside Greek; if some questionable Baltic forms like Lith. <pave'ikslas> 'example' are in fact related, they would point to a velar stop. I don't think we can get <aequus> from *ai-weikos, but I wouldn't be surprised if this *weik- did turn up somewhere in Italic.
> G&P:
>
> >> I had an idea a couple of years ago that <aeque> "equally, evenly"
> >> should be parsed as <ae>+<que> in the same way as <susque> and <deque>.
> >
> > Unfortunately, (a) the final que has a long vowel; (b) we're stuck with
> > the adjective aequus. It seems unavoidable that the root is *aequ-
>
> Thanks. Perhaps I'll try again -- I'm never as eloquent as I imagine. I
> had several ideas, confusing them then but even more in retrospect. The
> first was that the ending <-i:cus> itself could stem from a -- uh --
> cliticization of the "icon"-word, say something like *Hei-k- "alike".
> The next was that the root element of that word could be <ae(u)> from
> <aeuom> "even(ly)" or "continuous" or some such. Then I thought that
> <aeque> could be from <ae(u)-> and <-i:cus>, with <anti:que> as a
> parallel, but also that the ending could be <que> as in <deque>, using
> the supposed root <ae(u)->. But at least the latter would need
> reanalysis to yield the adjective and the verb ... It all ended in my
> usual level of confusion.