--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...> wrote:
>
>
> > >
> > Point for consideration (?): If /a/ is always of foreign origin
> > (since the original PIE didn't have /a/),
>
> Not always, cf eg. Latin schwa secundum.
>
> > why then didn't /a/ of foreign languages become borrowed as /e/ or
/o/
> > or some other native sound? As an example, after English lost the
> > phonemes /y/ and /y:/, Old French /y:/ was borrowed as /eu/ (> /iu/ >
> > /ju:/)(as in <pure>) and Old French /y/ was borrowed as /u/ (as in
> > <punish>). So if Latin, being a descendant of PIE, didn't have any
> > /a/ inherited from PIE, why then did they adopt an entirely new
> > unfamiliar sound?
>
> Because /a/ is a more 'natural' sound than /ΓΌ/. Only few vowel
> inventories have the latter, almost all have the former; it is a
> naturally given extreme of the vowel triangle. English took a road
> comparable to that of PPIE: some /a/'s became /e/, some became /o/,
> and the empty space of /a/ was filled with foreign loans, eg. 'spa'.
>
>
I guess my point is that if Latin really is derived from a PIE that
had no original /a/ phoneme, okay yes maybe words such as <lacus> and
<mare> were borrowed from a coastal people who had the /a/ sound, thus
increasing the frequency of /a/ (which by your criteria would arise in
Latin only from schwa and schwa secundum, both zero grades), but how
do you explain verbs like <sapere>, which has a Germanic cognate in OS
<bisebbian, afsebbian>, or <rapere>, or any of the duratives with
/e:/ as the formant like <patere>, <manere>, <habere>, <tacere>, etc.,
leaving aside <capere> (and also <habere>) where the /a/ could be
explained as developing from earlier /e/ under the influence of a
preceding uvular consonant (*qap-). These are all verbs of primary
(non-derived) declension, not transferred into one of the secondary
declensions as one would expect of verbs borrowed from outside PIE, or
no? I'm just somewhat skeptical of the idea that the majority of
/a/'s in Latin have to be accounted for as having come from outside
rather than having been inherited -- and therefore of the idea that
PIE _absolutely_ had no /a/. But all this has been discussed very
often before, I'm aware, yet I've never been fully convinced of the
idea that /a/ did not exist. Just my personal feeling.
Andrew
>