The Iceman's Berries

[Douglas G Kilday]

Stanley Friesen (5 Jul 2001) wrote:

>Thus, there is a large class of roots in Pokorny, and elsewhere, that are
>only attested anciently from northern Europe and immediately adjacent
>areas.  These form a substantial corpus of words that, while locally
>shared, cannot be confidently reconstructed for PIE proper.  One
>interesting factoid about this corpus is that it includes most of the roots
>in Pokorny with a reconstructed *a that is not traceable to an a-coloring
>laryngeal.  (Note, this is not an identifying characteristic, it is an
>observation).

>Now, my *hypothesis* about many of these words is that they are borrowings
>from a non-IE language family formerly wide-spread in northern Europe, but
>now extinct.  I wish I could figure out how to test this idea.

I would look first at hydronyms. Much of the leg-work of collecting old
hydronyms in W Europe and determining their constituents has been done; the
results are mostly in the German literature. I am thinking particularly of
Hans Krahe's papers in the old series of Beitraege zur Namenforschung. Krahe
summarizes his system in "Indogermanisch und Alteuropaeisch" (Saeculum VIII,
1957, pp. 1-16). He identifies suffixes such as -na, -ma, -ra, -antia and
roots such as Alb-, Ar-, Av-, Var-. As a series of Old European hydronyms
based on the root Al-, for example, he gives Ala (Norway), Alma (Etruria),
Almana (mod. Alme, NW Germany), Alara (mod. Aller, W Prussia), Alantia (mod.
Elz, SW Germany), and *Almantia (mod. Aumance, France).

Krahe interprets his Old European system of hydronyms as the relic of an
early stage of IE, and assigns meanings to some of the roots on the basis of
"cognates": hence Al-, Av-, and Var- are supposed to mean simply 'water,
stream' vel sim.; Alb- is supposedly 'white'; the Neckar allegedly means
'black'. Krahe notes the problem that many of the roots and "unsuffixed"
names contain short [a] and resemble non-IE forms, but argues that the
suffixes are IE, so the entire system must be IE.

Hans Kuhn in "Ein zweites Alteuropa" (Namn och Bygd LIX, 1971, pp. 52-70)
presents a more sophisticated analysis. He notes that hydronyms and other
toponyms in -ur-, -ar-, and -ir- constitute a deep stratum found throughout
W Europe but best preserved in the upper Po valley, N Switzerland, and part
of S France. In much of NW-NC Europe, these roots commonly take the suffixes
-s-, -k/g-, -n-, and -apa. Kuhn doubts the validity of IE "cognates": the
river Suhre might seem to invite explanation in terms of *su:ro- (IEW 1039)
'sour; salty; bitter', but this etymology can hardly apply to the Norwegian
island Surno/y, which appears to have the same root. Kuhn regards his
-ur/ar- stratum as definitely non-IE, and Krahe's stratum as a later
Indo-Europeanizing intrusion which incorporated many of the pre-IE names,
commonly adding suffixes. Krahe considered -apa to be a form of *akwa:- (IEW
23), though it should be noted that Pokorny also required *ap- (IEW 51) to
handle some Indo-Iranian forms. Either way, it is plausible that IE-speakers
could add 'water' to pre-IE hydronyms.

Kuhn observes that representatives of his -ur/ar- stratum are almost
completely absent from E Denmark, where Megalithic monuments occur. It
appears that the intrusive Megalithic society (which is plausibly the source
of the Germanic Seewoerter like <ship>, <oar>, <keel>, <sword>, <swim>,
<tauchen>) obliterated the earlier stratum here in the early second
millennium BCE. Krahe's estimate of 1500 BCE for _his_ stratum fits this
chronology. The early IE-speakers must have absorbed the Megalithians.

On these matters W.P. Schmid in "Alteuropaeisch und Indogermanisch" (Mainz
1968) and "Baltische Gewaessernamen und das vorgeschichtliche Europa" (Idg.
Forsch. LXXVII, 1972, pp. 1-18) has positioned himself as the anti-Kuhn. He
regards all the "Old European" hydronyms as resulting from undifferentiated
old IE, and goes far beyond Krahe in supplying IE etymologies for the roots.
For example, he refers Atese~ and Atesy~s (Lithuania), Ata (Latvia),
Attersee (Austria), Odra (Ukraine), and Adria (i.e. the Adriatic!) to *at-
(IEW 69) "ein Verbum der Bewegung". Now I can imagine how Bewegung connects
the concept of 'year' (Lat. <annus> < *atnos, Goth. <athn>) to 'river', but
'lake' or 'sea'? Also an arbitrary "d/t-Wechsel" is introduced which thumbs
its nose at established comparative IE. Schmid's underlying philosophy of a
pan-IE Alteuropa leaves no room for determining the actual stratification,
and in my opinion Kuhn's model makes a lot more sense.

Schmid places the Old European "center of gravity" (Schwerpunkt) in the
Baltic region, and observes that the Slavic homeland (das Gebiet der
Urslaven) lacks Old European hydronyms, which he explains by the loss of the
pre-Slavic system. In my view, Schmid's "center of gravity" is a phantom
based on the preconception of a Baltic IE homeland which is unwarranted by
the facts. The logical place to put the IE homeland is where one does _not_
find pre-IE hydronyms, and that excludes the Baltic in favor of the Slavic
Urheimat.

Anyhow, what I would do (if I didn't have my hands full with S European
matters) is to compare the corpus of "Old European" hydronomastic elements
(found in Krahe's BzN papers) with that corpus of N European IE roots,
looking mainly for systematic phonetic similarity. The results would likely
suggest further directions of investigation.

DGK