Person and Binding (a Partial Survey)

We provide a partial survey of three approaches to two problems in the semantics of person, which we call the ‘Bindability Problem’ and the ‘Shiftability Problem’. On the standard (Kaplanian) view of indexical pronouns, the semantic value of I and you is determined by a context rather than by an assignment function [=Separation Thesis], and furthermore this context must be that of the actual speech act [=Fixity Thesis]. The ‘Bindability Problem’ originates in the observation that indexical pronouns sometimes behave as bound variables, as in Heim’s Only I did my homework (... therefore John didn’t do his). This problem suggests that the Separation Thesis is too strong. The ‘Shiftability Problem’ consists in the observation that some indexicals may be evaluated with respect to the context of a reported speech act, so that what is literally John says that I be a hero may mean (in Amharic) that John says that he, John, is a hero. This problem suggests that the Fixity Thesis is too strong. We lay out and compare three approaches to these problems. According to THEORY I, pronouns are concealed definite descriptions whose restrictors may include predicates of the form speaker(x, c) or hearer(x, c), which are endowed with (i) an individual variable (which solves the Bindability Problem), and (ii) a context variable (which solves the Shiftability Problem, with the auxiliary assumption that attitude verbs manipulate context variables). According to THEORY II, first person pronouns are variables that are always bound, be it by (i) another first person pronoun (which solves for the Bindability Problem), or (ii) an attitude verb (which solves for the Shiftability Problem), or (iii) a (bare) λoperator. Rules of feature percolation account for the appearance of the right morphological features on the pronoun. According to THEORY III, a first person pronoun is a variable like any other variable, except that it must denote an individual with a specified role (‘author’ of a context). This requires an enriched notion of sequence of evaluation, one on which a sequence includes not just objects but also their roles with respect to the speech act. In order to solve the Shiftability Problem (as well as more complex instances of the Bindability Problem), Theory III posits that certain operators can introduce roles in the sequence that they manipulate. Somewhat surprisingly, a version of Theory III can naturally be stated in a way that derives important aspects of the syntactician’s Binding Theory (especially Condition C and Condition B).