phonetic resemblances, etc.

[Alexis Manaster-Ramer]

----------------------------Original message----------------------------
I so rarely agree with anything Johanna says and vice versa
(in THIS area at least, for I like most of her work in
other areas of linguistics very much and indeed admire it)
that I am tickled that she has made some of the very points
I was going to make later, but much better than I would have
(of course, we may disagree about some issues, as I will
point out below).  I esp. appreciate Johanna's support
given that I have been pretty ruthless (and in her view
unfair) in my critiques of her on various lists and in
print.  Alexis

On Sat, 30 Jan 1999, Johanna Nichols wrote:

> ----------------------------Original message----------------------------
> An important statistical point.  If this list
>
> >
> >>          Mamulique   Garza      Comecrudo
> >>
> >>sun       atl        ai         al
> >>moon      kan        an         eskan
> >>water     aha(?)     axe        apanekla
> >>road      --         aie        aaul
> >>man       (kessem)   knarxe     na
> >>woman     kem        kem        kem
> >>sky       --         apiero     apel
> >>
>
> were the entirety of our data on those three languages, we would be
> justified in considering relatedness to be probable.  For each of the seven
> glosses, at least two and often all three of the languages have resemblant
> forms; for each of the languages, each word resembles one or both of those
> of the other languages.
>
This is THE most important point, although absolute numbers cannot
be completely ignored (one of our small disagreements).

> Even if we reduce the phonetic resemblance to the binary distinction
> between initial "k" vs. initial "a", or even initial C vs. initial V, I
> suspect that the number of matches would exceed what is expected by chance.
>
The problem is that the chance vs. greater-than-expected-by-chance
business is not all that helpful, because (a) borrowings also
exhibit patterns of correspondence that could not have arisen by
chance, and (b) I continue to maintain that it makes no sense
to use a priori probabilities.  I have written on these points
in various places.

> Problem is (if I recall the article correctly), this isn't the entirety of
> our data. It's the seven most nearly resemblant glosses that could be
> picked out of a larger list.  That is, it's all and only the positive
> evidence.
True, but that the total amount of data for Mamulqieu and Garza is
maybeonly two or three times what we have here.  For Comecrudo, we have
LOTS iof data. But still we have so little data for M and G that this
is surely more than enough to make a strong case, just like Ives says.

Please note that the same thing happens all the time with lgs that
are represented by a few words in some newly dug-up or deciphered
inscription.

>
> (Do I recall correctly that only these three languages were compared in
> Goddard's original article?

Yes and no. He also tries to show that Comecrudo is NOT related to
Cotoname, although I claim they are, and so on.

> In that case the evidence, though weak because
> selected from a longer list of forms, is still stronger than the evidence
> usually offered in multilateral comparison...

This is really a side issue, but it seems to me that Johanna is just
repeating the usual disinformation about so-called multilaterla
comparison.  Please see Baxter and MR's revier of Ringe 1992 in
Diachronica and my forthcoming "Uses and abuses of math in lx"
(I can email either or both to anyone who wants them) on the
correct mathematical treatment of so-called multilateral
comparison.

> From wordlists as small and unreliable as the three above we know so little
> about the languages that there's little point in debating whether the
> resemblances are phonetically close, whether correspondences are regular,
> etc.
>Words consist of more than sounds.
[snip]
>  In that case the phonetics is immaterial; it's the morphemes that
> yield the resemblance.
[snip]
>  That would be an
> example of relatedness justifiably hypothesized from lexical material on
> the basis of something other than phonetic resemblance.
>
Precisely the point I made several times here and elsewhere.
When people classify languages, they often compare morphological
patterns without establishing REGULAR sound correspondences, just on
impressionistic ideas that the sounds look similar enough.
IE was established in this way, more or less, as a matter
of fact.  I think Niger-Kordofanian is an even clearer
example.

> For all we know, these could be
> gender-prefixing languages in which inanimates have /a-/ and animates
have
> /k-/.
Actually, we DO know, because Comecrudo is quite well known
as are more distnatly related lgs.  And this is not so.  But the
general point is exactly right. Thank you, Johanna.