Talskubilos wrote: ↑Sat Oct 24, 2020 3:33 pm
WeepingElf wrote: ↑Sat Oct 24, 2020 11:31 amNo, the problem with those "lost IE languages" lies somewhere else entirely: you can
almost always make up a fictional IE (or, for that matter, Semitic, Vasconic or whatever) language that "just happens" to show the sound changes you need to explain the word(s) you are trying to find an etymology for. That way,
nothing can be established; this dragon bites its own tail, so riding it gets you nowhere.
This isn't my way, though.
It indeed isn't. But I feel that your proposed languages are
weaker still. The fictional IE languages I have criticized above at least not only presuppose the conventional model of PIE you reject, but are based on the assumption of
regular sound correspondences, which I don't see in anything you have posted here - in fact, you have always been
dodging our questions for regular sound correspondences.
That said, I wouldn't claim that there are
no lost IE languages; in fact, one of my hypotheses (but one I am not sure of yet, though I use it in my conlangs) is that there was a lost branch of IE I call "Aquan" in western Europe, associated with the Bell Beaker culture and leaving traces in western loanwords and the Old European Hydronymy (you surely remember). But one must always be aware of the danger of circularity with such ideas!
I also wouldn't say that Pelasgian and Temematic are certainly wrong - just that the evidence I have seen has not seen sufficient to convince me. In my personal hypothetical model, there was an IE language related to Anatolian on the Balkan Peninsula once, and that one would have left loanwords in Greek. That one would, however, not be a
satem language (nor a
centum one, as Anatolian stands outside this division), nor do I see a reason to assume an Armenian-like stop shift (while you have been away, I have abandoned the glottalic theory, after finding that the root syllable constraints can be made sense of by simply assuming that the PIE voiceless stops were aspirated in an early stage of the language).
But it is IMHO unlikely that the Cimmerians spoke Temematic, as Holzer suggested. IE branches with such sweeping consonant shifts are found on the
outer edges of the ancient IE dialect continuum (such as Germanic, Armenian or Tocharian), not in its centre. If the Cimmerians didn't speak an outright Iranian language, they probably spoke something intermediate between Iranian and Slavic, such that any innovation of these two branches would also have been found in Cimmerian - i.e., Cimmerian would have been a
satem language with the widespread *D-*Dh merger.
Talskubilos wrote: ↑Sat Oct 24, 2020 3:33 pm
WeepingElf wrote: ↑Sat Oct 24, 2020 11:31 amAs an example of how this may fail, consider Greek
theos, where you could invent a "lost IE language" which has changed
*d to
th and thereby relate the word to Latin
deus. Of course, we know that such an etymology is spurious as there is the PIE word
*dhesos, with reflexes in other IE languages.
I've just found this example of a fictional lost IE language:
A shared substrate between Greek and Italic
I have taken a brief look at that paper; I shall read it again tomorrow. What tells us that these words are not just substratum loanwords
within Greek that were later borrowed into Italic? After all, there were
many Greeks in ancient Italy, and they had a
great cultural influence on the other peoples of that region, so one would expect Italic languages to be full of Greek loanwords - including ones that Greek had borrowed from some other languages, IE or not.
