The logic of Person markedness: Evidence from pronominal competition

In some languages, such as Hebrew and German, a d-pronoun (a pronominal demonstrative form) may refer to a human. When it does, the use of the d-pronoun may be associated with a pejorative effect, implying a negative evaluation of the denoted individual (henceforth N(egative)-effect). The N-effect is triggered, however, only under certain conditions. For example, when the d-pronoun is modified, no N-effect arises. This article examines the syntactic and pragmatic conditions under which this meaning emerges, and develops an account that integrates pronominal markedness and competition into the fold of conversational implicatures. The study addresses two questions: (i) What is the distribution of the N-effect? (ii) How is it linguistically encoded? Regarding (i), it is shown that the N-effect is triggered only when a personal pronoun could also have been used. This suggests that, everything else being equal, a personal pronoun is preferred over a d-pronoun; it also suggests that the N-effect is not intrinsically, or lexically, encoded. Regarding (ii), the N-effect must derive from the nonuse of a personal pronoun, and in this sense, it is related to markedness, and to systems that derive conversational implicatures. We argue that the use of a d-pronoun when a personal pronoun could also have been used gives rise to an implicature that the d-pronoun is associated with [−person], and we substantiate a theory of person as a contentful category that marks discourse participation.