Hallo!

Andreas Johansson wrote:

> [...]
> 
> It can be read here:
> 
> http://www.psych.auckland.ac.nz/psych/research/Evolution/Gray&Atkinson2003.pdf
> 
> Turns out I was misremembering - it shows Germanic+Italic as the sister of
> Celtic. Next group out is Balto-Slavic.
> 
> An odd feature is that it had Sinhalese+Romani as the outgroup of all other
> Indic languages bar Kashmiri. It's very tempting to suspect a linguistic
> equivalent of lateral gene transfer (ie. borrowing) at work here.

I finally found the time to read it, and I noticed a number of problems.

First, the whole family tree approach is somewhat controversial.
Languages simply do not bifurcate as cleanly as genetic lineages do.
Before a language breaks up into several different languages, it
differentiates into dialects, a process that is more aptly described
by the wave model (changes spread through the dialect continuum
in different directions, resulting in a complex pattern of
intersecting isoglosses).  You cannot meaningfully draw up a binary
tree for, say, the dialects of German.  The mere fact that the
High German consonant shift doesn't have a clear-cut boundary but
peters out towards the north with virtually every word having its
own "shifted/unshifted" boundary, demonstrates this.

Second, and this is a much bigger problem: what Gray and Atkinson do
is glottochronology.  And glottochronology is no longer accepted
by the vast majority of historical linguists.  This is because
glottochronology is based on the assumption that the rate of
lexical replacement is constant across time and languages
- in reality, it is neither.  For example, languages in intense
contact with other languages (e.g., Middle English) replace much
more vocabulary than geographically isolated languages (e.g.,
Icelandic).  Gray and Atkinson address some "minor" problems with
glottochronology which they claim to have overcome with some
advanced mathematics they call a "Bayesian Markov chain Monte Carlo
model"; but I don't see how that can remedy the problem that the
basic assumption of glottochronology - that lexical replacement rate
was constant - is false.  It is still glottochronology.

Third, the age they assign to Proto-Indo-European is impossible.
Any archaeologist will tell you that the wheel wasn't invented
yet 8000 years ago.  Yet, a PIE word for `wheel' is reconstructed
with as much certainty as is possible in this discipline.  And also
words for `yoke', `wagon', `carry by wagon', etc.  This means that
Proto-Indo-European can hardly be older than 6000 years.

Fourth, it is a leap of logic concluding that PIE was spoken in
Anatolia just because it is "at least 8000 years old".  Even if it
was indeed that old, it could have been spoken somewhere else,
for example in the area where now are the Bay of Odessa and the
Sea of Azov, which was dry land before 5500 BC.  And with the
maximum age of PIE at 6000 years (see above), the Anatolian
scenario collapses entirely.

Fifth, they base their study on modern languages - with the
exception of Hittite, Tocharian A and Tocharian B.  (I agree that
it is not easy to find a modern Anatolian or Tocharian language.)
All the historical linguistics textbooks I have read (and I have
read at least half a dozen of them) strongly advise to start
your reconstructive explorations with the oldest attested languages
you can find.  But compared with the other problems I have laid out
above, this is a minor quibble.

With so many problems, it is better to discard the whole thing.

Greetings,

Jörg.