[Lexicog] Turkey

[Koontz John E]

On Sun, 28 Aug 2005, David Kaufman wrote:
> Aha.  The French, Italian, Portuguese, and Rumanian cognates of pavo are
> paon, pavone, pava~o, and paun, respectively, all meaning "peacock."

Well, probably not cognates, strictly speaking, i.e., the product of
regular and parallel evolution from Vulgar Latin, but learned terms,
borrowed from Classical Latin pavo/pavonis (Nom. sg., Gen sg.) into the
various descendents at some later date.  Historical linguistics gets
complicated when the possibility of borrowing from a co-existing classical
language exists, and especially when an ongoing tradition of interaction
between the classical language and the daughter languages exists.  Here, I
think the treatment of the ending is regular - analogous with what would
happen in an inherited form, but I'm not so sure about the medial /avo/
sequences, at least not in Spanish.  Think of the paradigm for habeo.

A somewhat similar situation exists with loans in any case.  A lot of the
acculturation vocabulary that Rory Larson was working with exiswts in
multiple Siouan languages, ditto things like place names or ethnonyms.
These often look rather like cognates, but a certain amount of outside -
non-linguistic - information makes it clear thgat they can't be.

Examples:

'watermelon'  Santee sak[h]ayutapi, OP sakka dhide, Wi wic^aNwaNsake

A very early European introduction that spread like wildfire, I believe.
Regular enough in form, though this would seem to be the only OP relict of
*(r)ut- 'eat', which is suspicious.

'Platte River'  OP niN bdhaska, IO <ny>iN bra<th>ke

Perfectly regular, but probably a calque from Pawnee kickatus 'water
flat'.

'Ioway' OP maxude, IO baxoj^e, Os baxoce, ppaxoce

The second Osage version is irregular.

Similar things happen with terms that sem like they ought to be native,
e.g.,

'tobacco' Dakota c^haNli, OP niNniN, Wi taaniN

Tantalizingly similar, but not a regular correspondence.  In fact, no two
Siouan branches seem to agree.