Re: Portuguese farpa "barb" < *bHardHa?

--- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@...> wrote:

> --- In cybalist@yahoogroups.com, "stlatos" <sean@> wrote:
> >
> > --- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@> wrote:
> > >
> > > --- In cybalist@yahoogroups.com, "stlatos" <sean@> wrote:
> > > >
> > > > --- In cybalist@yahoogroups.com, "Joao S. Lopes" <josimo70@> wrote:
> > > > >
> > > > > Portuguese <farpa> means "barb", as in <arame farpado>, "barbed wire". Could be <farpa> related to barba "beard"? Usual etymologies consider it as from Arabian or Germanic.
> > > > >
> > > > > JS Lopes
> > > > >
> > > > The E word barb << barba, but in L the words barba = beard, and forfex = scissors are irregular, as regular changes of:
> > > >
> > > > *
> > > > ...
> > > > bhordha:
> > > > phortha:
> > > > forTa:
> > > > forfa:
> > > > forva:
> > > > forba:
> > > >
> > > > would be expected, but isn't found.
> > > >
> > > > An assim. f-v > f-f or v-v appears to be the cause (sim. to kW-p > kW-kW , etc.), so *farfa besides barba is just as likely as forfex, with rf > rp being regular in (some?, or opt./dia./irregular?) Romance a good possibility.
> > > >
> > > Obviously Latin <forfex> is a Sabinism.
> >
> > There's nothing obvious about it.
>
> To those suffering from obsessive-compulsive optional-soundlaw disorder (OCOSD), the obvious becomes opaque.

So what regular sound law occurred for m>f-N in:


formi:ca L; bórma:x \ búrma:x \ múrma:x Dor G; maoiri:- Av; vamrá-s , valmí:- S;

or

hi:bernus L; kheimerinós G; jmeRn Ar;


versus nothing like m>f-N in words like:

membrum L; meNzdrica = membrane of egg R-CS

mi:luus L; mérmnos G;

Mone:ta L;

mu:nus, commu:nis L;


Is it rational to say m-m but not f-f could undergo dissimilation-at-a-distance (which is usually irregular even in standard theory)?

>
> > The expected *forbex is reflected in Tuscan <fo`rbice>.
> >
> > I didn't say anything excluding an additonal *forbex, or many others, above.
> >
> > >
> > This did not become *borbex, nor did <forbea> 'food' (Fest.) become *borbea, so simple assimilation cannot explain <barba>.
> >
> > I said it was irregular (just like forfex). The changes p-kW > p-p or kW-kW are both seen in L, also with no explanation. In a similar way to f-v, qui:nque beside prope can not be explained by anything but optionality, since borrowing from, say, P-Celt. or Q-Celt. for both (and every other word with p-kW) are unreasonably unlikely. Similarly, if I posit dis. in formica from m-m it doesn't mean every word with m-m HAD to undergo dis., too.
>
> Latin <forfex> is no more "irregular" than, say, Middle High German <wa:pen>. It is a straightforward borrowing from a neighboring related language.

Since barba is not a straightforward borrowing from O-U, no irreg. word from f-f should be considered obviously borrowed with no ev.

>
> Italic regularly has *kW...kW from PIE *p...kW

Wrong.

>

as we see in Lat. <qui:nque> (with long vowel from the ordinal), <quercus>, <coquus>, and <cu:nctus>. Borrowing from Celtic is a straw man.


I do not believe in borrowing from Celtic, which has a p-kW change that is sim. to but dif. from L., but I posited it to logically discredit it just as I attempted to do for forfex < O-U (I'm just clarifying this for anyone following along who might not be familiar enough w L/Celtic to understand my point).

>

It is silly to posit an optional reverse assimilation to *p...p in order to derive Lat. <prope> from *pro-kWe, which would not agree with the usual sense of *-kWe.

>

prope (adv) propius (com) proximus (sup) = near L;

propinquus (adj) = near L;


The "Lat. <prope> from *pro-kWe" was made long ago to explain -x- and -qu-, not by me, and though I don't agree with all parts of it, it's impossible to say < p-p with no irreg. or assim. There's nothing silly about what I wrote, or what anyone else has said about assim. in prope (from what you wrote I don't even think you knew or remembered about proximus , propinquus ).


Since propinquus is obv. contaminated by analogy w an older *propi, an older *pronkWos < *prokWnos < *prokWinos is likely (like *perkWinos > *perkWnos > *kWerkWnos > quernus ( < quercus < * ). Depending on timing, a regular change of p-kW in either direction is possible if pC- and kWC-blocking are part of the rule (as I already wrote long ago), but that's based on a small sample of each possible occurrence, and seems too complicated and much more unlikely considering other opt. changes for kW that I've seen in L. (and in many other IE). The existence of a tendency for these opt. changes, in some, but not others, very common to the point of possibly being regular within a lang., while obv. opt. in another, makes much more impossible to say.

> However, I do not share de Vaan's enthusiasm for Dunkel's ad-hoc derivation of <prope> from *pro-pro with dissimilative loss of -r-. I think instead that we are dealing with *pro-pi, a juxtaposition of two (prepositional) adverbs, with *pi being the zero-grade of <poi>, a preposition found in some Greek dialects (e.g. Locrian <poi ton waston> 'to the city'). Furthermore Sanskrit has prefixal <pi-> (e.g. <pinahyati> 'he ties (something) to') which cannot be regarded as zero-grade of <api->.
>

Since propinquus requires an older *propi < *prokWi, the der. of *prokW()i < *proxWi = in front of / facing is best (compare xW > kWH in keankH Arm; quick E; < *gWixWwos).