Towards a Biologically-Plausible Computational Model of Human Language Cognition

The biolinguistics approach aims to construct a coherent and biologically plausible model/theory of human language as a computational system coded in the brain that for each individual recursively generates an infinite array of hierarchically structured expressions interpreted at the interfaces for thought and externalization. Language is a recent development in human evolution, is acquired reflexively from impoverished data, and shares common properties through the species in spite of individual diversity. Universal Grammar, a genuine explanation of language, must meet these apparently contradictory requirements. The Strong Minimalist Thesis (SMT) proposes that all phenomena of language have a principled account rooted in efficient computation, which makes language a perfect solution to interface conditions. LLMs, albeit their remarkable performance, cannot achieve the explanatory adequacy necessary for a language competence model. We implemented a computer model assuming these challenges, only using language-specific operations, relations, and procedures satisfying SMT. As a plausible model of human language, the implementation can put to test cutting-edge syntactic theory within the generative enterprise. Successful derivations obtained through the model signal the feasibility of the minimalist framework, shed light on specific proposals on the processing of structural ambiguity, and help to explore fundamental questions about the nature of the Workspace.