Experimental Evidence for the Syntax of Phrasal Comparatives in Polish

We first review the data behind the generalization and the formal model that has been proposed as its explanation. In section 2 we discuss the gradient nature of the effect and the resulting need for eliciting acceptability judgments in controlled experimental conditions allowing statistical analysis. In section 3 and 4 we discuss two acceptability-rating studies on Polish testing the validity of the empirical observation in (1a) and the theoretical explanation in (1b). The results support (1). We note the implications of our findings for theories of subject islands and conclude with a summary.