Iberia

[Larry Trask]

Ed Selleslagh writes:

>  -In Antiquity both Georgia and Mediterranean SE Spain were already called
>  Iberia, puzzling the writers of the time. In Spain the name seems to be
>  related to 'ibar', Basque (Iberian?) for 'river (valley)'(cf. the river
>  Ebro, Lat. Hiberus, another example of a river 'River'?). Note that ancient
>  Georgia also consisted essentially of two major river valleys parallel to
>  the Caucasus, so it could have been called 'ibar' by non-Kartvelians.

The idea of deriving the name 'Iberia' from a Basque word has long been
popular, but there are big problems with it.

The central Basque word here is <ibar>.  Today this word means 'valley',
and it is not systematically distinguished from the other 'valley' word,
<haran>.  However, there is some textual evidence pointing to an
earlier distinction.

Azkue concludes that the earlier sense of <ibar> was 'water meadow',
'piece of low-lying fertile land' (possibly next to a river), and this
is accepted by the great Vasconist Michelena.  Indeed,
Azkue translates the word as Spanish <vega>, which means exactly this,
and which itself (it is of unknown origin) has at times been suspected
of being derived from an unrecorded Basque derivative *<ibaika>.
In the mountainous Basque Country, of course, such patches of fertile
low-lying land are found nowhere but in valleys.

In contrast, Sarasola notes that <haran> sometimes sems to denote
specifically a *high* valley, a valley lying high in the mountains.

It is therefore at least possible that <ibar> and <haran> were once
distinguished along these lines, and that their modern synonymy results
from a loss of the distinction.

The most widespread Basque word for 'river' is <ibai>, or sometimes
<hibai> on the French side.  This is pretty clearly a derivative of
<ibar>, whose regular combining form is <iba->.  (Compare, for
example, <iztei> 'groin', from <izter> 'thigh'.)

But this is not the only recorded Basque word for 'river'.  In much
of the center and east we find instead <uhalde> ~ <ugalde> ~ <ubalde>.
This is a transparent compound of <ur> 'water', whose regular combining
form is <u->, and <alde> 'side'.  The original sense was therefore
'waterside', and the word is well attested in this sense, as well as
in the sense of 'river'.   We also have the northern word <ibaso>,
another derivative of <ibar>, with an opaque second element, but this
is not recorded before 1852.  In the Zuberoan dialect on the French
side, we find yet another word for 'river', <uhaitz>, obviously a
compound of <ur> 'water' with a puzzling second element.

Now, <ibai> looks very much like a reasonably ancient formation, but
perhaps not all that ancient.  The problem is that this word is
almost entirely absent from toponyms.  Basque toponyms are overwhelmingly
based on local geographical features, and <ibar> in particular is
exceedingly frequent in toponyms, and yet <ibai> is almost absent.
The only toponym I can think of which obviously contains it is
<Ibaizabal>, the name of the biggest river lying within the Basque
Country.  This is transparently <ibai> + <zabal> 'wide', and the
wholly unreduced form of the name suggests that it is of no great
antiquity.  Two other toponyms, both town-names, may possibly
contain it: <Baiona> 'Bayonne' and <Baigorri>.  But such an analysis
of the first is largely dismissed by the eminent philologist Irigoyen,
who prefers a personal name <Baio> as the first element.  For the
second, an *<ibai-gorri> 'red river' does not seem impossible but
cannot be regarded as certain.  I don't know if there is anything reddish
about its river, with the very non-Basque name <Aldude>, but I'm inclined
to doubt it, since this river comes rushing down on the town from
limestone peaks nearly 1500 meters high.

On the other hand, <ibai> is not uncommon in surnames, which were
established only in the late Middle Ages, also mostly from local
geographical features.  So, we can't conclude that words for 'river'
were systematically disfavored by the earlier Basques for coining names.
I note also that <ur> 'water' is very frequent in toponyms.

All this suggests to me that <ibai> is perhaps not especially ancient
in Basque, and therefore perhaps not available as an etymon for any
names recorded more than 2000 years ago.

This leaves the seemingly more ancient <ibar>, which, from the point
of view of <Iberus>, the Roman name for the Ebro, at least has the
advantage of containing the needed /r/.  But Basque has no long
vowels, and no long vowels can be reconstructed for the Pre-Basque
of some 2000 years ago.  And Latin <Iberus> has a long /e/ in it.
How could <ibar>, with a short /a/, be the source of <Iberus>,
with its long /e/?

I'm afraid a Basque source for the name of the Ebro does not look
especially plausible to me.

By the way, I have seen it suggested that the source might be a
supposed Celtic *<iber>, but I have no idea if any such Celtic item
can be substantiated.  However, for what it's worth, there were
certainly plenty of Celtic-speakers in the Ebro valley when the
Romans arrived, and Celtic toponyms are not rare even within the
Basque Country.  A good example is <Deba>, the name of a river in
Gipuzkoa and of the town lying at its mouth.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: 01273-678693 (from UK); +44-1273-678693 (from abroad)
Fax: 01273-671320 (from UK); +44-1273-671320 (from abroad)