Re: bidet

--- In cybalist@yahoogroups.com, "Tavi" <oalexandre@...> wrote:

>
> --- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@> wrote:
> >
> > Basque dialects also have <gipulla> and <dipula>. We are dealing with
> a culinary Wanderwort,
> >
> IMHO this is a cultural loanword from Latin/Romance like 'grape',
> 'lentil', 'oats', 'hen/chicken' and many others imported by Paleo-Basque
> speakers, mostly agro-pastoralists previously unacquainted with these
> items.

But the point I make is that given the variety of <kipula>, <gipulla>, <tipula>, <dipula>, we have several different borrowing routes, and it is rash to assert that palatalization of *k- occurred WITHIN any particular (Paleo-)Basque dialect.

> > Your own theory involves unspecified other languages with their own
> soundlaws before the
> > High Middle Ages.
> >
> That's right. Linguistic data points to several varieties at the
> Paleo-Basque stage. For example, Asturian and dialectal Basque basa
> 'mud' (with an apical sibilant /s'/) became pats 'pomace' with syncope
> of the final vowel and strengthtening of the sibilant at word-final.
> Also the initial labial stop /b/ hardened to /p/, subsequently lost by
> Martinet's Law in ats 'muddy'. But in the "Proto-Basque" variety
> reconstructed by Mitxelena we've got batz 'pomace; dregs; slush' (with a
> laminal affricate sibilant /ts/) . Interestingly, French névasse
> 'slush' has a similar element -assa usually regarded as a despective
> suffix.

Are you suggesting the French word might be a compound? That seems very unlikely.

> The thing is Mitxelena's reconstruction is mostly based on the evolution
> of Latin loanwords in Basque such as causa > gauza.

And Martinet's depends heavily on structuralist considerations, putting philosophy before facts. I really think he was at a loss to explain why fortes should have been lost in word-initial position (usually the most marked) and kept word-internally, while of course lenes (by definition WEAKER than fortes) were kept word-initially.

But philosophy aside, the most glaring defect of Martinet's treatment is his neglect of word-internal mediae geminatae, both in borrowings (abbas > Guip. apaiz, Lab. apez, Low Nav./Soul. aphez) and in compounds (*beg-buru 'eye-head' > *bebburu > bep(h)uru 'eyebrow'; *beg-gaitz 'eye-evil' > bekaitz 'envy').

My working hypothesis is that the rarity of word-initial tenues in Basque is not due to attrition by Martinet's Law (which indeed leaves inexplicable exceptions). Instead I think the tenues arose, as a rule, from old fortes, in turn from old longae, in turn from clusters of stops. These did not arise word-initially in native Basque, but could occur in borrowings from Celtic, or by redivision of compounds.

> > > You seem to ignore that "Aquitanian" is an epigraphic corpus
> > > representing more than just one linguistic variety (probably a
> dialectal
> > > continuum), as shown by lexical isoglosses like this one. Another
> one is
> > > t-/h-, a particular case of Martinet's Law by which fortis plosives
> at
> > > word-initial got aspirated into /h/ or zero.
> >
> > If Aquitanian is so heterogeneous, there is even LESS reason for
> making these equations.
> >
> Actually, the Aquitanian evidence is largely outnumbered by Basque
> itself, where we can find doublets with and without Martinet's Law.

Only for initial k- are these convincingly numerous. "Uhlenbeck nous offre une longue liste." For initial t- and p- only a handful of examples can be given, and the zero-initial forms may have arisen by analogy. Martinet feels compelled to cite highly speculative Vasco-Georgian correspondences as supplementary evidence. (Since you have hitched your Basque wagon to Starostin's Proto-NORTH Caucasian, I doubt you take Vasco-Kartvelian any more seriously than I do.)

> > Anyhow, whatever century it comes from, the fact remains that
> <borondate> has initial bo-.
> >
> > > Comming from an earlier *p- as in poz, ponte, putre, palatu, pago,
> etc.
> >
> > I doubt it, and <ponte> 'tufa' is from Latin <fontem>.
> >
> Actually 'baptismal font'. Romance /f-/ became regularly /p-/ as in fago
> > pago.

Basque <flakatu> 'to get weak' appears to come from Romance, and <farisau> 'Pharisee' already from Medieval Latin (with <ph-> pronounced [f-]). Since Bq. <praka(k)> 'breech(es)' is from Celtic, I do not rule out the same source for <pago> 'beech tree', specifically the Gaulish accusative *ba:gon. The Lat. acc. <fa:gum> should have given Bq. *pagu, and Romance either *fago or (if it came through Gascon) *(h)ago.

> > Martinet said that ancient Basque opposed initial [pH]
> >
> > > = fortis /p:/
> >
> > to [b],
> >
> > > = lenis /p/. I think you didn't read him well.
> >
> > He used a special sign [b.] with a circle subscript.
> >
> But [] indicates phonetic realization, while // is *phonemic*.

I am not arguing against you here, but if we use fortis/lenis opposition in phonemic notation, perhaps we should use /P/ and /P:/ to avoid confusion with longa/brevis opposition.

DGK