Speaker:Professor Stephen Crain
Department of Linguistics,
Macquarie University

Title:Principles and parameters of logic

Time: 2014/12/16 (Tuesday) 14:00-16:00
Venue: R101 Science Building 5, NCU (中央大學科五館 101 教室)

Abstract:
Across languages, young children map linguistic expressions onto the corresponding concepts of classical logic. For example, the meaning children assign to disjunction words across languages is inclusive-or, as in classical logic. Children also adhere to the patterns of classical logic in combining the meanings of logical expressions, including de Morgan’s laws of propositional logic. Despite the considerable direct overlap between human languages and classical logic, there are points of variation. Certain of the differences can be attributed to lexical (scope) parameters. These include parameters for disjunction OR, conjunction AND, and for the deontic modal MAY. These parameters encode binary scope relations between these expressions and negation. According to these parameters, the deontic modal undergoes reconstruction at Logical Form in some languages but not in others, and the expressions for disjunction and conjunction are Positive Polarity Items in some languages but not in others. The alternative values of these scope parameters are in a subset/superset relationship, so children would confront learnability problems if they initially adopted the superset value, in the absence of negative evidence in the primary linguistic data. Therefore, a subset principle is invoked to orchestrate the acquisition of scope assignments, directing children to initially adopt the scope assignments corresponding to the subset value of these parameters. Consequently, children’s initial scope assignments are expected to be the same as those of adult speakers of languages that adopt the subset value, but to differ from those of adult speakers of languages that adopt the alternative, superset value. This talk reviews some of the experimental evidence in favor of this account, and considers some of its theoretical and empirical consequences.