Re: Nostratic, Eurasiatic, Mitian, ...
Posted: Tue Aug 22, 2017 7:32 pm
Do they?Richard W wrote:Of the 20 most conservative Uralic words, as measured by retention within Uralic, 10 have apparently related, well-established Indo-European correspondents.
The best-retained Uralic words, with unambiguous reflexes in every branch, are the following 25:
– pronouns: *ku- 'who', *me- '1PP', *mi- 'what', *to 'that'
– basic verbs: *kalə- 'to die', *kad₂a- 'to disappear', *ńalə- 'to lick', *ńälə- 'to swallow', *puna- 'to plait', *ujə- 'to swim'
– body parts: *mëksa 'liver', *sënə 'vein', *sülə 'bosom', *śilmä 'eye', *śɜd₁äm(ə) 'heart'
– other nouns: *ëla 'low', *nimə 'name', *ńëlə 'arrow', *ńërə 'moist/wet', *pälä 'half', *pesä 'nest', #wäśkä 'metal'
– etc: *kuma- 'overturned', #kɜk/ttA '2', #wittə '5'
The bolded ones have well-established IE counterparts (in English: who, me, that, spin, sinew, name). But I can only get together six, half of them pronouns (even considering that the 1st person pronouns *mi(n(ä)), *ti(n(ä)) do not even make the cut). One could think up some sort of comparisons for some of the others too —I mentioned 'heart' above — but they're not similarly obvious.
If we allow words that fail to cover one branch, and/or have uncertain reflexes (I have a collection of these, down to 6 out of 9, assembled here: http://www.frathwiki.com/Proto-Uralic/distribution), the track record doesn't look much better, plus we actually also rake in two clear Indo-Iranian loanwords (*śëta '100', *śarwə 'horn').
Don't get me wrong, this is decently good evidence still, but it's not case-obviously-settled-everyone-go-home good.
Also, let's assemble a similar list on the IE side. How many of those will find good Uralic counterparts? Mallory & Adams give 15 words found in "all major IE groups" (— counting Baltic separately from Slavic and Iranian separately from Indic, but that doesn't seem to affect the big picture). Their list scores only three clear hits. Four if we count the suppletive oblique stem of 'I', but I don't know if that's included in the count here:
*wódr̥ 'water', *gʷōws 'cow', *pōds 'foot', *dʰwōr 'door', *tréyes '3', *pénkʷe '5', *septḿ̥ '7', *h₁néwh₁m̥ '9', *swep- 'to sleep', *h₁nómn̥ 'name', *h₁eǵ- '1PS', *wey '1PP', *tuH '2PS', *yuHs '2PP', *so 'that'
They also have a distribution-ranked Swadesh list. Interestingly, in case there's only one branch missing, the branch scoring the most misses is not Anatolian (the supposed earliest branch)… it's Albanian. (Maybe this and Armenian should be not counted as full-fledged "major groups".)
Hardly: the Korean three-phonation system only arose in Middle Korean through consonant cluster reductions (e.g. ps- sp- > *ph- *pp- > pʰ- pʼ-).WeepingElf wrote:I think the main reason for Starostin, Dybo and Mudrak to reconstruct three grades of stops for Proto-Altaic is the inclusion of Korean, which of course has a three-grade stop system.