Print

Print


Is this a minimal universal grammar? If not please accept this as an entertaining tangent.

My current mental model of how universal grammar works is that there is a biological core for syntax that sees the world in the form of words made of individually meaningless sounds, combined with meaningful chunks of words, collections of words that can be arranged in trees, then again collected into discourses of connected sentences.  But on top of that, in my mental model, is a learned cultural layer of tones, lexical classes, non-parametric/non-instinctive syntactical rules, learned templates that allow creating the mind boggling complex verbs of polysynthetic languages, etc all of which are learned like other non-instinctive skills, like calculus and bike riding.

If you can translate from English to some sort of Basic English/Globish to an arbitrary conlang, I would imagine you'd be translating the core, instinctive universal grammar. 

The other idea that came to mind is that this methodology would produce an pixelated, 8 bit English and a very English-y grammar. Definite articles are a peculiarity of languages around Europe, but they are so common in English, a corpus driven attempt to extract the "core" grammar would likely end up with some definite articles. And the conlang would have them too. Also, the conlang would be analytic an syntax driven like English, when if you start from a clean slate, it seems just as likely that a conlang could rely mostly on morphology (say, like Amerind polysynthetic langauges) or discourse to get the point across (say, like toki pona). 

But if the goal is to create a better English, this is an interesting methodology.

Matthew Martin