Re: Can relationships between languages be determined after 80,000 y

One of the most rugged survivors on the linguistic scene is -yV, which seems
to make not love but adjectives all over the world.

Patrick

----- Original Message -----
From: "Piotr Gasiorowski" <gpiotr@...>
To: <cybalist@yahoogroups.com>
Sent: Monday, January 28, 2008 7:32 PM
Subject: Re: [tied] Can relationships between languages be determined after
80,000 years?


> On 2008-01-28 18:52, mkelkar2003 wrote:
>
> > http://itre.cis.upenn.edu/~myl/languagelog/archives/001037.html
>
> The Andamanese languages were assigned to the Indo-Pacific superfamily
> by Joseph Greenberg (together with Tasmanian, for example). I, for one,
> find such classifications unacceptable. The Andamanese peoples have
> probably lived there for tens of thousands of years and speak languages
> belonging to at least two endemic families. Not all their pronoun
> systems are similar to that of Juwoi, so the argument involves some
> cherrypicking (BTW, Kusunda has <gina>, not <gida> for 'he', which makes
> the similarity to Juwoi kitE slightly less striking, and I'm sorry to
> report that the similarity does not extend to the plural).
>
> We have no proof that the Kusunda have lived just where they still live
> for time immemorial; their language may once have had relatives all over
> India. With all these reservation the idea that a small segment of the
> basic vocabulary, such as the pronoun system, can remain in existence
> for tens of millennia, is not absurd. I mentioned this very possibility
> a few days ago, commenting on the word-evolution study by Pagel et al.
> However, to DETERMINE relationships between distant languages which have
> pronoun systems that look as if they might be related one would have to
> look for further matches following the same pattern. After all, striking
> as the similarity is, it MAY be accidental. A correspondence like tui :
> tsi, gen. tii-ye : tsi-yi seems to indicate that both languages are
> incredibly conservative -- almost no change over millennia upon
> millennia. It looks closer than the agreement between, say, Greek and
> Gothic! If so, cognates should be very easy to find, and if they haven't
> been found outside the pronoun system -- well, we may be dealing with
> another BD/BD case (barely detectable but beyond demonstration), or
> merely a ghost relationship.
>
> Piotr
>
>