Vagaries of Swadesh Lists (was: Automatic clustering of languages)

From: Richard Wordingham
Message: 48273
Date: 2007-04-07

--- In cybalist@yahoogroups.com, "Roger Mills" <rfmilly@...> wrote:

> Yes, "bad" Ml. jahat, BI buruk. The Bahasa Indonesia list at
> http://language.psy.auckland.ac.nz/austronesian/language.php?id=233
> gives BI jahat.

> The two languages IMO are just marginally more diverse than US/UK
English;
> minor differences in pronunciation and vocab; they differ of course
as to
> sources of loanwords (English vs. Dutch), and BI has many more
borrowings
> from regional languages that aren't likely to show up in a Swadesh-type
> list.

Talking of loans, isn't _nama_ 'name' in the Indonesian Swadesh list
an Indic loan?

I'd have expected to see more Indic loans, as Malay is presumed to be
the source of Indic loans in Tagalog. There's only one identified,
_kepala_ 'head' < Skt _kapa:la_ 'skull'.

The Javanese list has one identified Indic loan (_mega_ 'cloud' - the
Thais have no doubt that their normal word for 'cloud', _me:k_, comes
from Sanskrit/Pali _me:gha_) and one loan whose source is not obvious
to me - _tlaga_ 'lake'. But I could see another two apparent loans
from Sanskrit - _dina_ 'day' (Skt. _dina_) and _sirah_ 'head' (Skt.
_çiras-_ or _çira_) in the Swadesh list, and they are not labelled as
loans. (Thai has borrowed the form _çi_rs.a_ for 'head', but though
common it has not displaced the native word.)

Of course, plain coincidence can occur - Austronesian _mati_ 'die'
looks a lot lile Pali _mata_ 'dead', but Maori _mate_ 'die' indicates
that this is a coincidence.

Richard.