Print

Print


On Wed, 25 Feb 1998, Robin Gaskell wrote (excerpt):
 
>         As promised, here are the Universals from the Dobbs Ferry, New York,
> Conference of 1961. ["Universals of Language" (c) 1963,
> M.I.T., Lib. of Congress Cat. Card No.: 62-22020]
>
> ????How do these relate to conlangs and auxlangs?  Maybe you would like to
> check out your favourite auxlang against this list, to see that it is not
> contrary to the list, i.e. has no features outside Greenberg's findings.!!!!
 
    Is the claim that this list is merely descriptive of a sample, or
that it is somehow prescriptive in the sense that human languages
cannot function with other characteristics?  If the former, then what
difference does it make whether auxlang X does or does not conform to
this list?  If the latter, what is the proof other than a mere list
(which in and of itself is merely descriptive)?
 
>         Joseph H. Greenberg based his provisional list on thirty natlangs, which
> he found to vary significantly on three main paramaters.
 
    How were these thirty languages selected?  If the sample was not
quasi-random, then it could be biased, and thus the thirty might not
necessarily be very representative.  It almost seems to me to be
suspect a priori without further information on the sampling
methodology, considering that thirty is a puny fraction of the total
number of more or less distinguishable languages.
 
Paul                             <[log in to unmask]>
..........................................................
Paul O. Bartlett, P.O. Box 857, Vienna, VA 22183-0857, USA
Finger, keyserver, or WWW for PGP 2.6.2 public key
Home Page:  http://www.access.digex.net/~pobart