another correspondence: Haida q'- : Proto-Miwok *t.-

[De Reuse, Willem]

Hi Geoffrey:

I have no problem with the q'-, t. correspondence.  Assuming these language groups are distantly related, anything could have happened.  But I am not sure why you say that this is a significant number, since you only have five correspondences.  Are you sure that such coincidences between consonants cannot be found anywhere, assuming one looks long enough?

I am of the old school of historical linguists, which believes that consonant correspondences by themselves do not prove anything, regardless of the statistics, but that they only mean anything when supported by striking morphological correspondences.  What do you with a language without morphology?  Well, there really are no languages without any morphology at all, and if the language has very little morphology, well, tough bananas, it is just harder to apply the comparative method convincingly in such cases.

I know there is a new school out there, which is convinced that consonant correspondences (or sound correspondences in general) are enough.  A very interesting example is the recent work of Gianfranco Forni, who has written intriguing articles arguing that Basque is Indo-European, and also that Etruscan in Indo-European.  Intriguing but not convincing, precisely because no convincing morphological correspondences can be found.

Willem de Reuse
________________________________
From: ATHAPBASCKAN-L [ATHAPBASCKAN-L at LISTSERV.LINGUISTLIST.ORG] on behalf of Geoffrey Caveney [geoffreycaveney at GMAIL.COM]
Sent: Sunday, August 11, 2013 9:38 AM
To: ATHAPBASCKAN-L at LISTSERV.LINGUISTLIST.ORG
Subject: another correspondence: Haida q'- : Proto-Miwok *t.-

I posted previously on this list about the correspondence in core lexicon items between Haida initial velar stops and Proto-Miwok initial *k- before a stressed vowel ('liver', 'bone', 'I/my'), and between Haida initial uvular stops and Proto-Miwok initial *k- before an unstressed vowel ('black', 'night', 'one').

I have found another striking sound correspondence between Haida and Proto-Miwok: Haida initial /q'-/ corresponds to Proto-Miwok initial /*t.-/. The Miwok languages have two distinct /t/ phonemes, and the dot after the t indicates the alveolar /t./ as opposed to the dental /t/. (In the standard orthography the dot is below the t, which is difficult to reproduce in computer typing.) I think it is plausible that the Proto-Miwok dental *t- corresponds to Haida d- and t- while the Proto-Miwok alveolar *t.- could correspond to Haida q'-.

Initial /*t-/ is relatively rare among Proto-Miwok and Proto-Eastern Miwok forms, with only 15 such forms reconstructed in Broadbent and Callaghan's definitive 1960 article, and initial alveolar /*t.-/ is even rarer, appearing only as an alternative phoneme in 2 such forms. But for both of these forms there is a corresponding Haida form with initial /q'-/:

Proto-Eastern Miwok *t/t.is<http://t.is>í:nay- 'ant'
Haida qaj q'iisdang 'ant'

Proto-Miwok *t/t.yy... 'to rest'
Haida q'a 'sleep'

Further, there are several other comparisons for which the Sierra Miwok languages have initial alveolar /t.-/ and Haida has initial /q'-/:

Proto-Eastern Miwok *tálka- 'skin'
Central Sierra Miwok t.álka- 'skin'
Southern Sierra Miwok t.olla 'skin'
Haida q'ál 'skin'

Proto-Miwok *ty... 'to shoot'
Central Sierra Miwok t.ýkky- 'to shoot'
Lake Miwok túw-en 'to shoot' (with dental /t/ but the rest of the form corresponds closely to the Haida form)
Haida q'wáan- 'to fire at'

Proto-Sierra Miwok *t.oʔ-nge- 'to sit down'
Haida q'awa 'sit'

Finally there are a couple more comparisons where the Miwok forms have only dental /t-/ but they otherwise appear to correspond well to Haida forms with /q'-/:

Proto-Eastern Miwok *tukú:n- 'salmon'
Haida q'áagw 'freshwater sockeye salmon'

Proto-Miwok *túmay 'stick'
Haida q'awáay 'plank, board, lumber'

It is striking that in all of these comparisons the Haida forms have the specific phoneme /q'-/, the uvular ejective stop, not just any uvular stop. With such a small number of Proto-Miwok forms with initial alveolar /*t.-/, the probability of such a significant number of them corresponding to Haida forms with initial /q'-/ by chance coincidence must be extremely small.

Geoffrey Caveney

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/athapbasckan-l/attachments/20130812/177a1bd7/attachment.htm>