Re: Trask on Basque-IE comparison

From: Tavi
Message: 69341
Date: 2012-04-16

--- In cybalist@yahoogroups.com, "stlatos" <stlatos@...> wrote:
>
> http://tech.groups.yahoo.com/group/Nostratic-L/message/432
>
> --- In Nostratic-L@yahoogroups.com, Miguel Carrasquer mcv@ wrote:
> >
> > Larry Trask's "The History of Basque" discusses a number of words for which
> > possible IE connections exist:
> >
> > adar "horn, branch" ~ OIr. adarc "horn"
> >
This vould be a Vasconic -> Celtic loanword.

> > angi(o) "fenced-off pasture" ~ OIr. aingid "protect"
> >
> > Dismissed by Trask, correctly.
> >
A better comparanda would be Germanic *angjo: 'meadow', so this is probably a non-Celtic IE loanword.

> > aran ~ Aragonese arañón ~ Ir. áirne "plum"
> >
> > Trask objects that the Romance forms are from *agranio:, while the Irish
> > word points to *agrinja.
> >
But the source is surely Gaulish *agranio: 'sloe, fruit of the blackthorn', which differs from Insular Celtic.

> > argi "light, bright" ~ PIE *h2arg- "to shine"
> >
> > Trask's objection that no IE form shows *h2argi- with final -i is incorrect
> > (*h2argi- is the Caland variant found in compositis for this word, e.g.
> > Grk. argi-, Skt. r.ji-).
> >
IMHO Basque -i would be the outcome of the IE thematic vowel *-o-.

> > bost "five" ~ Celtic *bost- "palm of the hand"
> >
> > Trask rejects this: the Basque word is bortz (> bost in the W. dialects).
>
> The connection of bost "five" ~ Celtic *bost- "palm of the hand"
being denied because of dif. in Basque dialects is ridiculous; he has no way of knowing how far back such alt. goes or even what the (more) original form or C-order was.
>
Trask is right here because the shift -rtz- > -st- in some Basque dialects is well attested. However, I think likely a more remote connection with a root meaning 'hand' found in Berber *fus 'hand', Chadic *pVs 'hand; arm', and from which Celtic *bos-ta: 'palm, fist' derives.

> > hogei "20" ~ MWe. ugeint "20"
> >
> > Rejected by Trask. I'd agree.
> >
The serious incovenience of lacking Proto-Celtic forms (as the ones reconstructed by Matasovic) for comparison becomes evident in this case, as PCeltic *wikanti: is readily compable to Basque *bogeCi.

Basque, like Celtic, has traces of a vigesimal count system where 40 = 2 x 20, 60 = 3 x 20 and 80 = 4 x 20.

> > izoki(n) "salmon" ~ OIr. éo, Gallo-Latin esox
> >
> > Trask prefers a borrowing from Latin/Romance ESOCINA.
>
> Especially odd is his supposed esox >> izoki(n) borrowing.
The word has no known IE et., was borrowed from Celtic into Latin, and all ev. makes it likely it was borrowed into Celtic from another unknown language, most likely Basque or a relative. At the very least, it would look like an unknown language >> both Basque and Celtic; def. not L >> Bq.
>
Asturian esquín 'young salmon' is surely a Celtic loanword parallel to Basque izoki(n). Trask was merely following Mitxelena here, who had plenty of abusrd etymologies such as azeri 'fox' from Asena:rius or luki (Biscayan) 'fox' from Lu:cius.

> >  zulo ~ zilo "hole" ~ Celtic *si:lon
> >
> > Not possible (*si:lon would have given *ziro, not zulo). My idea is that
> > it's connected to PIE *dheub-lo-/*dhoub-lo- "hole", which would have given
> > exactly Basque *zullo > zulo.
> >
The Celtic comparanda is bad. The right one is *tullo- 'pierced, perforated'.

> A lot here looks less like scholarly consideration and more like Trask dismissing obvious connections for no apparent reason (leaving open the pos. he did have an agenda or was less competent than you say).
>
I think you're right on both respects: he had a agenda and he also was less competent than Rick thinks.

I'd also like to comment some of Carrasquer's proposals:

> > xahal "calf" < *sanal, *s^anal < *s(^)amal-, can be connected to Scandinavian words for "reindeer".
> >
This is connected to Celtic *damo-/*damato- 'bull', Greek damále:s 'young bull', etc.

> > orein "deer" < *oleni, cf. Slavic elenI ~ olenI etc.
> >
This is from Celtic *elan(t)i: 'doe, hind' with a similar evolution than hogei '20'.

> > zaldi "horse" < *daldi. Pliny mentions a breed of Asturian ambling horses "thieldones" (*�ieldones <
> > *deld-), PIE *del- "to amble".
This is a satemized form of IE *g(W)old- 'foal, young of an ass' (Germanic *kulta- 'colt', Sanskirt gardabhá- 'ass'). Also found in Iberian saldu and Berber a-serdun 'mule'.