Grammar is NOT a Computer of the Human Mind/Brain
Prakash Mondal
January 2020 

This paper will attempt to debunk the idea that human language grammar as part of the Faculty of Language(FoL) is intrinsically a computing device. The central argument here is that grammar does not compute. One way of demonstrating this is to show that the operations of grammar in the Generative model do not have the character typical of computations. Thus, the central operation of grammar Merge, which combines lexical items to produce larger expressions, can be defined as a recursive function, but it does not share the inductive properties of recursive functions in mathematics in view of the consideration that recursive functions define computability. On the other hand, if the language faculty is a computing system, the language faculty must inherit the halting problem as well. It is easy to impose the halting problem on the selection of lexical items from the lexicon in such a manner that FoL may or may not terminate over the selection of lexical items. We can say: there is no FoL way of telling if FoL will ever terminate on x or not when x is a selection from the lexicon. The halting problem for FoL is disastrous for the view that grammar is a computing system of the brain/mind since it detracts from the deterministic character of FoL. This has significant repercussions not just for grammar that cannot be restricted to any limited view of mental computation but also for the nature of the cognitive system as a whole since any cognitive domain that is (supposed to be) language-like cannot be said to compute as well.

Format:	[ pdf ]
Reference:	lingbuzz/003984 (please use that when you cite this article)
Published in:	Journal of Cognition and Neuroethics
keywords:	grammar; syntax; language faculty; merge; computation; cognition; semantics
previous versions:	v1 [February 2018]
Downloaded:	610 times

 

[ edit this article | back to article list ]