On Thu, 30 Jan 2003 13:39:13 -0000, "Richard Wordingham
<richard.wordingham@...>"
<richard.wordingham@...> wrote:

>If the concordance has any genuine connection with Antonio Tovar, it
>will be a typographical error or misreading of something like 5.0%!
>Using the data at www.rosettaproject.org , I did a quick comparison
>of the first Basque and Tamil 100-word lists I found. Only 12 items
>seemed worth taking a second look at, and that was being charitable.
>
>Isn't the claim that A. Tovar found such a high concordance a libel?

The claim is that Tovar "asked Dr. Lahovary to try this method on
Dravidian". I haven't seen the 1961 publication (A. Tovar, L.
Michelena, K. Bouda, R, Lafon, W. Vycichl & M. Swadesh, "El método
léxico-estadístico y su aplicación a las relaciones del vascuence"),
but the only numbers cited by Trask ("The History of Basque", p. 359)
are 10% for Basque-Berber, and 7.5% for Basque-Kartvelian (presumably
Edo Nyland's "Circaskian/Kirrukaskan").

Michelena, Tovar and Lafon have made important contributions to the
study of the Basque language, but I don't doubt Larry Trask's
conclusion about this 1961 article is justified: "this is the height
of silliness".

=======================
Miguel Carrasquer Vidal
mcv@...