Dating the final IE unity, in particular the word for "horse"

[Gábor Sándi]

----- Original Message -----
From: <X99Lynx at aol.com>
Sent: Monday, 06 March, 2000 4:34 PM

I would like to comment on the following statement by Steve Long:

> You see, my guess has always been that there is NOTHING in comparative
> linguistics that in some scientific way clearly DISPROVES Renfrew's dates.
> And although professional impressions have weight, they cannot be dispositive
> on scientific issues.  Nothing so far has scientifically disproven Renfrew's
> dates.

It is very hard to see what kind of evidence you would accept as
"disproving" Renfrew's dates (and, consequently, Renfrew's theory). After
all, any word in a language can be a loanword, therefore - in principle -
the words for horse, wheel, the cart and its parts, yoke, copper/bronze,
various trees and wild animals, can all be borrowed in a specific IE
language. Alternatively, they can all be derived independently from native
roots. The main question is: how likely is it that such borrowings /
derivations happened independently from each other, often in languages far
removed from each other?

Let me concentrate on the word for horse in Latin: equus.

Where do you think it comes from? Let me assume that you accept the reality
of PIE - if you don't, it is difficult to continue any kind of debate. So -
I can think of four possibilities:

------------------------------------------------------
POSSIBLE EXPLANATION ONE:

Equus is derived linearly, with one generation of speakers following
another, from a PIE word reconstructed by linguists as *ek^uos.

Here, k^ stands for a velar or palatal voiceless stop that was subject to
satemization in the "satem" languages; I am ignoring the possibility of an
initial laryngeal, which - not being reflected in any daughter language - is
not relevant here. Now, if this etymological explanation is correct, the
word *ek^uos or its linear descendants all the way to 'equus' must have
existed at all stages in the development of PIE > Latin, as - before the
invention of writing and written traditions - there was no way for words to
disappear and then reappear in a language, unless by borrowing (for which
see later).

If this was the case, the word must have referred to something, presumably
to a horse-like creature. The problem is that there is no evidence of
horses, wild or domesticated, in Italy before the Polada culture (ca. 2600
BC, if I am not mistaken). If Renfrew's Indo-Europeans arrived in Italy
around 5000 BC, or even earlier, they presumably encountered no horses
there. Actually, there is no evidence for horses at the time in the areas
that they had come from according to Renfrew, namely Greece or the Balkan
peninsula. So: what did *ek^uos refer to? Either a mythical beast, or the
donkey (were there actually donkeys around?), or maybe the sheep or the cow?

I don't like any of these explanations, and I find it hard to believe that
when the pre-Latins encountered the horse after 2600 BC, they pointed at it
and said: aha, here is this mythical beast that our ancestors in Anatolia
called 'equos', nice to find it again after so many millennia. Or,
alternatively, they said: let's call it 'equos' after the donkey, which it
resembles somewhat, now all we have to do is invent a word for the donkey,
not to confuse the two. Good thing that the word 'asinos' is around, applied
to - say - the deer, now we have to find a word for the 'deer'. A nice chain
reaction, there.  [You'll notice that I presume that the ending -os had not
yet become -us, because this is a historically attested phonetic change that
occurred during the 1st millennium BC].

The only weakness in this argument is that maybe horses were present in
Italy at the time, we have just found no evidence for them. I ask
archaeologists: is this likely? Italy is well covered by archaeological
sites from many eras - if horses were there before 2600 BC (at first only in
the extreme north), is it likely that their bones would be consistently
missing from before 2600 BC, while present in ever increasing numbers
afterwards?

----------------------------------------------------------------

POSSIBLE EXPLANATION TWO

The word "equus" in Latin is an independent creation, possibly from the root
that gave also 'ocior' (faster). The fact that Old English had "eoh",
Sanskrit "asvas" is irrelevant - by chance these languages have relied on
their own derivational resources to end up with a similar word for horse.

I will not spend time refuting this possibility - it is too absurd.

----------------------------------------------------------------

POSSIBLE EXPLANATION THREE

The word "equus" is ultimately a loanword in Latin - it was borrowed from an
IE language spoken north of the Alps at the same time (or later than) the
animal itself was introduced into Italy.

This explanation is not impossible, and it would not contradict Renfrew. The
main problem is that there are far too many words associated with the early
Bronze Age and Sherratt's Secondary Products Revolution that would have to
have been borrowed around the same time. Not impossible, but unlikely - at
least in my opinion.

A modification of this theory would have other IE languages borrowing the
word from Italy - this strikes me as extremely unlikely, seeing that the
domestication of the horse is in any case associated with the area north of
the Black Sea.

---------------------------------------------------------------

POSSIBLE EXPLANATION FOUR

I can think of a fourth explanation, one that would not completely destroy
Renfrew's theory, but modify it: Renfrew could be right as far as PIE
(Anatolia) > PIE (Balkans), but then the Italic languages would have been
introduced later into Italy from the north or the northeast, say as
derivatives of the Baden culture. These people would have had the horse,
which would have came to them by diffusion from the east together with its
name.

This explanation would not exclude an earlier strain of IE speakers in Italy
(with no name for the horse), who would have come in as Renfrew indicates,
and who would have supplied Latin and the other Italic languages with an IE
substratum. Such a substratum might well have supplied Latin with the many
words with "unetymological" a: quattuor, canis, manere etc.

--------------------------------------------------------------

A nice hypothesis, this last one - and, maybe, one pointing in the direction
of eventually reconciling Renfrew and Gimbutas. After all, the Kurgan theory
has no very convincing hypothesis for the prehistory of the PIE speakers -
they might have come from the Balkans, in which case the neolithic
revolution in Europe would have introduced languages closely related to, or
even ancestral to, what we know as PIE.

But I leave all this to another posting.

Best wishes to everyone,

Gabor Sandi