Dear friends,
I would just copy letters send by Ante Aikio to another IE mailing list
(
Indo-European@...). Please read these before answering the poll.
I asked him permition to copy it on this list.
[Ante Aikio :]
I'm interested in knowing how widely accepted is the theory that the
Indo-European "Urheimat" was located in Eastern Europe. And how much
support do such theories as e.g. Colin Renfrew's idea of the Anatolian
origin of IE languages have?
I ask this beacause one central piece of evidence in support for the
East-European origin comes from outside the field of IE studies, namely
Uralic linguistics. But it seems to me (correct me if I'm wrong!) that
among many IE-ists, there's a tradition of uninterest in diachronic
linguistics done outside the IE language family. So, I'd also like to
ask
how well is the recent progress in Uralic linguistics known inside your
field of research? And especially, I'd be interested in hearing comments
on what is presented below from those who do NOT support the
East-European
original home.
During the last ten years it has been discovered that Uralic languages
possess extremely ancient IE loan words: they were loaned from proto-IE
to
proto-U[ralic], which has been dated approximately 4000 bc or before. In
order to provide some substance for discussion, I will give some
examples on the loan etymologies. All etymologies derive from the
Germanist Jorma Koivulehto and most can be found in his book "Uralische
Evidenz für die Laryngaltheorie" (1991). The etymologies are meant to
serve only as an illustration. Thus, only a fraction of the credible
loan
etymologies put forward are presented here.
proto-U *pel(x)i- 'fear' < proto-IE *pelH- 'grau, fahl; schreckig'
p-U *toxi- 'bring, give, sell' < p-IE *doH- 'give'
p-U *koki- 'see, find' < p-IE *Hokw- 'see'
p-U *kulki- 'move, flow, walk' < p-IE *kwelH- 'drehen, sich drehen
usw.'
p-U *mos´ki- 'wash' < p-IE *mozg(-eye)- 'untertauchen'
p-U *s´alkaw- 'pole, rod' < p-IE *g´halgho- id.
p-U *weti- 'water' < p-IE *wed- id.
The criteria by which the loans must be judged proto-Uralic are the
following:
1) The phonological shape of their cognates in present-day U languages
does not warrant one to assume that they were loaned separately into
(and
between) already differentiated U languages / dialects. The distribution
suggests the same: all the etymologies above have cognates in at least
one
U language in the Baltic Sea area and one in Siberia.
2) the proto-U form requires a specifically proto-IE loan original. Many
even show proto-U *k or *x as a substituent of an IE laryngal.
It is undeniable that the contacts between speakers of U and IE
languages
date back to the earliest stages recovered by the comparative method.
Thus
the speakers of proto-U and proto-IE must have been geographical
neighbors. As a result, theories such as Renfrew's Anatolian "Urheimat"
must obviously be discarded (it is of course impossible to assume that
proto-U spekers would have occupied an area south of the Black Sea). It
seems that the only logical option is to place proto-IE in Eastern
Europe
north of the Black Sea. This area is just about south from area where
current research usually places the center of the Uralic expansion.
[another letter :]
> As for the position of Anatolian, there are very few things that
> distinguish Proto-IE as it was before the break-off of pre-Anatolian
from
> what it became during the time up to its next split when the second
branch
> (Tocharian?) left the remaining stock. Consequently, we cannot tell
from
> the form of an IE loanword in Uralic whether it was borrowed before or
> after the separation of Anatolian from the rest. That being so, we
cannot
> quite exclude that the IE homeland was in Anatolia and that the rest
had
> moved to the north of the Black Sea and there met the
(pre-)Proto-Urals
> and handed them a bag of loanwords. Note that laryngeals are not
retained
> only in Anatolian; there are enough laryngeal-sensitive phonetic
changes
> in the individual branches to guarantee that laryngeals survived as
> segmental units well into the separate lifelines of the other
subbranches
> also.
This is an interesting point of view. Is there strong evidence for the
claim that the first split in IE was between Anatolian and the rest? If
only very few things distinguish P-IE before and after the break-off of
Anatolian, are these differences substantial enough to warrant the
reconstruction of a separate, intermediate proto-language for the rest
of
the IE branches?
> We need very specific evidence to tell whether Anatolian had
separated
> from the rest or not by the time of the oldest loans in Uralic (the
words
> for "earth", "bear" or "fire" would be interesting)
It appears that these words were not borrowed to Proto-Uralic: P-U has
*myxi- 'earth', *tuli- 'fire', *elä- 'carry, lift' and *kan-ta- 'carry',
which appear to be native Uralic words. The last one is actually a
causative derivative of P-U *kani- 'go'.
There are some later loans, however. Samoyedic has *pura- 'burn' < IE
*pur- 'fire', and Hungarian and Ob-Ugric have replaced P-U *tuli- 'fire'
with *tüwV-s (perhaps from IE *dhew-, although the vowel causes some
difficulties).
> Flaunting my ignorance, I may perhaps ask the silly question: How
can
> you exclude that the Uralic homeland had a prehistory south of the
Black
> Sea? If Proto-Uralic was only one language when it split into the many
> that have become known, can one really exclude that the speakers of
that
> language had earlier lived somewhere else? I'm not advocating that one
> should make up all sorts of fanciful scenarios, far from it, but it
just
> could be playing a trick on us.
In principle, it cannot be totally excluded. However, the idea of a P-U
homeland south of the Black Sea would not be a very fruitful hypothesis,
since it would only create a new, very difficult question to answer: why
and how would the P-U speakers have migrated north to become
hunter-gatherers in the taiga/tundra zone of northern Eurasia? There is
no evidence suggesting that the P-U Urheimat would have been -outside-
the
area where U languages are spoken today.
[another letter :]
One thing worth noting is that dates have gone back conciderably. The
Uralic expansion (for whatever reason it happened) may have been earlier
than the IE one. Some descendant of proto-U (which would later evolve
into
Finnic and Saamic) was spoken in the Baltic Sea / Scandinavia area
already
3200 bc, when the Indo-European battle axe culture arrived in
southwestern
Finland. There are some Finnic-Saamic loan words with proto-IE
characteristics (e.g., laryngal reflexes) which are probably connected
with the battle axe culture, such as Finnic-Saamic *kas´a- 'tip, end' <
IE
*Hak´-, *suki- 'family, kin' < IE *suH-. These words are unknown in U
languages outside the Baltic Sea / Scandianavian area. There are also
independent proto-IE loans in Saamic (e.g. *s´uki- 'sharpen' < IE *k´uH-
'pointed, sharp'), which possibly points towards a very early IE /
pre-Saamic contact zone in mid-Scandinavia. At any rate, a uniform
proto-U
language cannot be assumed to have existed after 4000 bc, and 4500-5000
bc
seems more likely. I am not sure how this correlates with IE dates - is
4000-5000 bc too early a date for proto-IE?
A question that should be thoroughly researched is the possibility
of U loan words in IE. Since there are massive amounts of IE loan words
of
varying age in the U languages, it seems overwhelmingly unlikely to me
that there would be (almost) none in (proto-)IE. However, I know only
two
debatable cases that have been proposed, Germanic-Celtic *yeg- 'ice' <
p-U
*jäNi id. (N = velar nasal), and IE *(s)kwalo- '?whale, ?fish species' <
p-U *kala 'fish' (the first case seems convincing enough to me). Has
anyone written anything on this?
[another letter :]
ords can be borrowed for many reasons, not only because there is a
"need" for them. If the contact is intensive enough and/or the prestige
great enough, basic vocabulary can be borrowed. I see no reason why
proto-U could not have borrowed these words from
Indo-Europeans. Similarly, Finnish has borrowed words for e.g. 'mother'
and 'food' from Germanic, and 'tooth' and 'neck' from Baltic, and North
Saami has borrowed words for e.g. 'son', 'meat', 'man' and 'moon' from
proto-Scandinavian. These examples are not exhaustive; there are lots of
relatively late "basic vocabulary" loans from IE languages in Finnic and
Saamic.
Furthermore, there is internal evidence in Uralic supporting the loan
origin of p-U *weti 'water'. This item does not have cognates in Saamic
and Khanty. Instead, these two languages share a lexeme p-U *s´äcä,
whose
reflexes mean 'water' in Saamic and 'flood water' in Khanty. This was
probably the original U word for 'water', which was replaced in the
other
languages by the IE loan word *weti. The word *s´äcä has no cognates in
languages that show reflexes of *weti. (The Khanty meaning 'flood water'
is probably secondary, since the Khanty word for 'water' derives from
the p-U word for 'ice').
[another letter :]
As for common origin, it is not possible to
discredit semantically and phonologically natural loan etymologies by
replacing them with Nostratic speculations. There have been multiple
attempts to relate IE and U genetically, and all of them have failed. Of
course, it is impossible to -prove- (in the strict sense of the word)
that
the lexical similarities are not due to common genetic origin. But
then, it is impossible to disprove -any- proposed genetic
relationship. Because of this, the task of proving belongs to those who
propose a genetic relationship, and this has not been succesful for
proto-Indo-Uralic or Nostratic.
Chance correspondeces are naturally another matter. However, given that
there are only about 300 semantically and phonologically credible P-U
etymologies, and these contain at least 30 proto-IE loans, I'd say that
chance correspondence is ruled out. One also has to bear in mind that
the
U languages contain substantial amounts of later loan words of varying
age (eg. proto-/pre-Baltic, proto/pre-Germanic, proto/pre-Aryan,
proto-Iranian); this makes it natural to assume that contacts between U
and IE were possible at a still earlier date (i.e., when at least the
former still was a proto-language).