Re: Constructive linguistics

> But that's not an appropriate analogy.  A good weather analogy would
> be trying to seed the clouds to increase rainfall, or using windturbines
> to increase airflow, and then making conclusions about how weather in
> general works.  Your analog is equivalent to professional linguists
> using theoretical models, algorithms, to predict grammatical and
> ungrammatical strings.
Okay. Why not use your first version, then? Wouldn't need to have the
whole complex system to play with just one variable (or a few) - and
if you do, in some situations you could make an a posteriori conlang
with the minimum changes needed. Or am I missing something completely?

 - Sai