A Classification of Right Node Raising. Are some instances of Right Node Raising parentheticals?

Abstract

The term Right Node Raising (RNR) refers to a parallel construction in which a sentence-peripheral element seems to be shared by both conjuncts but is only realized in the second (e.g. Mary likes, but Sue hates the TV show). The enormous flexibility of this phenomenon has occupied researchers’ interest for some decades now. The aim of this thesis is twofold. First, I am going to provide an extensive overview of the phenomenon and its properties. This classification will serve as the basis for the evaluation of current syntactic approaches to RNR. Along with other researchers, I will come to the conclusion that an eclectic approach is needed. Secondly, I will investigate the question whether RNR might also receive a parenthetical interpretation in some instances (referred to as parenthetical RNR). I argue that a parenthetical interpretation makes fundamentally different predictions with respect to the overall syntax and the status of the target. Instead of assuming a coordinate structure between two conjuncts of the same type, a parenthetical approach postulates the existence of a non-coordinated host sentence which is interpolated by an elliptical string of words introduced by some kind of connective (the parenthetical). To support this claim, I present the results of a BNC corpus study, searching for examples where the punctuation (brackets or dashes) marks the construction as a parenthetical. The analysis of these examples provides the opportunity to gain valuable insights about what kinds of instances are possible candidates for a parenthetical interpretation when punctuation is inconclusive (commas).