At 4:12:36 AM on Friday, May 4, 2012, Tavi wrote:
> Don't forget this PIE-speranto (as well as your own "PIE")
> are actually *conlangs*, not real languages.
No, the result of linguistic reconstruction is not a
conlang. It is our best (albeit limited) approximation to a
real language, ontologically distinct from a conlang. The
further extension of a reconstruction to a working language,
à la dnghu.org, arguably does turn it into a conlang, albeit
one with a basis in reality. Linguistically informed
hypothetical projections based on counterfactual assumptions
(e.g., what might English look like today if it had been
much more resistant to borrowing) are yet another category
distinct from conlangs per se, though they certainly have
some commonalities.
Brian