On Fri, 20 Nov 1998, Nik Taylor wrote:

> Mathias M. Lassailly wrote:
> > maybe all languages need a certain optimal number of specific
> > *concepts* that may be transferred on as differentiated words as
> > possible ?
> I suspect that that's closer.  English, for example, has seperate words
> "die" and "kill" (as do most, if not all, IE languages).  Turkish, on
> the other hand, uses a causative prefix (suffix?) on the word for "die"
> to make "kill".  Languages with more complex derivational/inflectional
> morphology can probably have fewer roots.  But even a certain number of
> *concepts* I suspect is a range, tho smaller than the range of roots.

Well then, a lower bound would be 214 or so, the number of
Chinese radicals ... On AUXLANG and GLOSALIST there has been
discussion of the size of a basic vocabulary, and many lists
have been cited, ranging from 850 (Basic English) up to
6000 (the Glosa full list), with the concensus being
that 2000 to 4000 is about right.

This is only a rough measure of how many roots are needed,
not words per se. Or maybe it is a measure of how much
patience an average adult has for learning new arbitrary
terms for the sake of international communication.

I think 3 lists are needed: one for grammatical particles
and derivational affixes, around 100 to 200 items;
another for basic vocabulary, around 3000; and the open
class of terms, mostly noun roots, 10000 to infinity.
Of course it all depends ...