<html>
At 01:53 01.10.2002 +0200, Mikael Parkvall wrote:<br><br>
<blockquote type=cite class=cite cite>* Provided that we belive that
languages do split, which time span is most likely to produce a split --
five minutes or two millennia?<br><br>
* And once a split has occurred, what is most likely, that one single
isolate becomes extinct, or that all the dozens of daughters of a
proliferous mother dies out?<br><br>
I am not laying claim on any Absolute Truth, but put together, I think
the two rethorical questions above constitute -- an argument. And
therefore, until somebody has come up with a better argument, it makes
sense to me to assume that there is a correlation (albeit probably a weak
one) between age and the number of surviving daughters.<br><br>
</blockquote><br>
Of course there is a <i>statistical </i>correlation between age and the
number of surviving daughters. Such a correlation would be interesting to
linguists only to the extent that there is also <i>proportionate</i>
correlation, so that we could predict the number of daughter
languages from the time depth - or vice versa, with a reasonable degree
of likelihood. My claim is that that would be possible only of languages
were independent of the material condtions of their speakers. Since we
cannot predict the frequency of genocides, migrations, or meteors, we
cannot predict the proportionate correlaiton between splits and age. All
we can say is what Mikael says himself: "Two millennia is more
likely to produce a split than five minutes". I fail to see at what
level that is an interesting observation. <br><br>
Jan Terje<br><br>
<x-sigsep><p></x-sigsep>
Professor Jan Terje Faarlund<br>
Universitetet i Oslo<br>
Institutt for nordistikk og litteraturvitskap<br>
Postboks 1013 Blindern<br>
N-0315 Oslo<br>
Norway<br><br>
Phone<x-tab> </x-tab>+47 22 85 69 49 (office)<br>
<x-tab> </x-tab>+47 22 12
39 66 (home)<br>
Fax<x-tab> </x-tab>+47 22 85 71 00<br>
</html>