Principles, Parameters, and Schemata : A radically underspecified UG

Parametric models have been the only viable alternative to unsuccess- ful theories of the language faculty based on evaluation metrics for grammars. In this article I show that parametric analyses can attain a high degree of typological and historical adequacy, though they raise serious problems for explanatory and evolutionary adequacy. I propose to replace the Principles&Parameters theory by a simpli- ed model of the language faculty, which eliminates parameters altogether from the initial state of the mind, replacing them with few abstract variation schemata, and, in the absence of positive evidence in the primary corpora, eliminates them even as open questions in the course of acquisition in the absence of positive evidence. In this model, ‘parameters’ only arise as positive answers to yes/no questions of limited form. Attained I-languages can be represented as simple strings of positive and neutralized values of different lengths. The new research program (Principles&Schemata) is capable of retain- ing the advantages warranted by a system of heavily constrained binary choices for language acquisition, variation, and history, while underspecifying UG and simplifying the acquisition path and the representation of the steady state of each I-language: it promises to be able to return to a feasible question-based model of syntax acquisition triggered by positive evidence only, though without the shortcomings emerged from the classical Principles&Parameters theory.