An idea re nonnasality & Puget CJ; cf. Tlingit CJ discussion

[Sally Thomason]

Dave's idea is interesting and quite possibly correct.
My own analysis of those m's in loanwords from CJ in
languages of the region has been a bit different:
often, when borrowing words from another language,
sophisticated multilinguals will apply "correspondence
rules" (or, as Jeff Heath has called them, "borrowing
routines").  Roughly, though I certainly wouldn't
claim that speakers are always conscious of doing so,
the process would be this: Those guys' sound X
corresponds to sound Y in my language, so I'll replace
X with Y in this word I'm borrowing from their
language.  Sometimes speakers actually do articulate
this: well, it's X in their language, so it ought to
be Y in ours.

I think that's what happened to produce loanwords
like latam.  That is, I think languages that have
latam instead of latab (or latap) have m because that's
their sound that corresponds to the nasalless languages'
b -- in other words, they borrowed the CJ word not
directly from CJ, but from one of the nasalless languages.
(Indirect borrowings are common throughout the Americas,
and no doubt elsewhere too: language A borrows from, say,
Spanish; language B borrows the word from A, C borrows
from B, etc.)

Tlingit is beyond the main CJ area, which might be
relevant in considering why Tlingit speakers nativized
CJ phonology: maybe there was a distinctive northern
variant of the pidgin, even.  I'd be astonished to find
that *any* group of speakers were unable to reproduce the
CJ sounds they heard; but it's quite common for speech
communities to have their own habits of reproducing words
borrowed from other languages, and even of speaking other
languages.  Another possibility is that Johnson was
mistaken about Tlingit-CJ; unless I'm remembering wrong,
there isn't much data on it.

  -- Sally