Print

Print


On 4/25/07, Edgard Bikelis <[log in to unmask]> wrote:

>     Why not measure language development based on the size of its
> corpus? Provided it is meaningful, the language needs to be fairly
> developed to endure, say... twenty pages. But maybe I'm counting too
> much on the good senses of my hypothetical conlanger...

Corpus size was one of the criteria proposed for the draft
Wikipedia conlangs policy.  I would tend to agree that however
complete a language appears to be from its grammar description
and lexicon, its having a tiny corpus or no corpus at all would make
me hesitate to call it "complete" or "well developed".  But most of this
discussion has centered on how to count the size of a language's
vocabulary or lexicon, a different issue altogether, and perhaps
a more interesting one, because different decisions about how
to count words, morphemes, idiomatic phrases, etc. probably have
a more dramatic impact on the lexicon size than on the corpus
size.

-- 
Jim Henry
http://www.pobox.com/~jimhenry