No Proto-Celtic?

Gabor Sandi g_sandi at
Fri Apr 27 17:08:03 UTC 2001

This is an excellent topic to raise, Larry. I have also encountered the
"crystallization" idea in my readings (always from archaeologists) and for
the life of me I can't imagine what the term means in linguistic terms.

One example you have not mentioned is Bosch-Gimpera (1980): Les
Indo-Europiens - problhmes archiologiques. Crystallization is a big topic
with him, as a convenient process to back up his views of IE origins: it all
goes back to the mesolithic. Europe has pretty much always been IE according
to this view, it is thus unnecessary to look for migrations, let alone
conquest as an explanation for the widespread presence of the family on the

I think that the crux of the matter is the preconceived notion is that
migration / conquest is not a major feature of prehistory. The fact that the
linguistic map of Europe has been several times redone in historical times
through migration / conquest does not seem to be a useful argument in this
context: neolithic and bronze age peoples behaved differently from their
descendants according to anti-migrationists.

In my view we linguists are themselves to blame to a certain extent: we have
tied language too closely to ethnicity. We might be better off saying: look,
we don't really care about the process by which Europe became
Indo-Europeanized. But our evidence for PIE points at bronze-age speakers
for whom animal husbandry was much more important than cereal growing, and
who were familiar with horses and wheeled vehicles of some sort. You,
archaeologists, make of this information what you will, but we linguists are
not going into contortions to explain IE words for the horse, the wheel and
its parts, copper (or bronze), silver and gold through any means than simple
derivation from the PIE vocabulary.

The same kind of argument could be used for Proto-Celtic (PC). Rather than
try to identify some iron-age population as the PC speakers, why not use the
following working hypothesis:

(1) There was a language ancestral to Gaulish, Welsh, Irish etc, which was
derived from PIE, but which was not ancestral to, say, the Germanic, Italic
or Baltic languages we know.

(2) This PC language had undergone certain changes, so that the word for
'father' was *(h)atnr [I am not sure about the *h], the word for 'hundred'
*kNtom [N standing for a syllabic n], for 'I carry' they said *ber{ and for
a demonstrative they used *sindos, -b, -on. Etc. etc. for those aspects of
Celtic that we can take back to the hypothetical common language.

(3) Any language satisfying these criteria is by definition part of PC, any
language contradicting them is not.

(4) If someone does not believe in the existence of PC, than it is for
him/her to explain how the features listed in (2) could conceivably exist in
different languages.

If we put it like this, it would be just as difficult to deny the existence
of PC as it would be to deny the reality of, say, Schwitzerd|tch or
Australian English.


An interesting view of why some archaeologists (and linguists) seem to have
an almost gut-felt dislike of migrationist theories is on the ff. web site:

With my best wishes,

Gabor Sandi
g_sandi at

More information about the Indo-european mailing list