Reducing Long-Distance Binding to Logophoric Exemption

This chapter re-examines the hypothesis that some anaphors can be long-distance bound independently of their discursive conditions. All analyses of long-distance anaphora, whether they assume binding domain parameterization or covert movement, rely on the existence of a specific type of anaphors that can be bound out of the Condition A domain and are further characterized by monomorphemicity, subject orientation, sloppy readings, and blocking effects. The goal of this chapter is to question this empirical claim and examine the hypothesis that such purported long-distance anaphors can in fact be reduced to exempt anaphors subject to logophoric conditions. Some tests are proposed and applied to Icelandic sig, Mandarin ziji, French soi, and Norwegian seg/sin using online questionnaires. The results suggest that the hypothesis that long-distance binding should be eliminated from the theory and reduced to logophoric exemption is viable—pending further cross-linguistic studies.

How to Identify Exempt Anaphors

The goal of this chapter is to design tools for reliably identifying instances of exempt anaphors. In particular, the inanimacy-based strategy distinguishes between plain and exempt anaphors independently of the definition of Condition A: since inanimates cannot be logophoric, an anaphor occurring in a configuration disallowing an inanimate anaphor is necessarily exempt. This strategy allows us to re-examine the distributional properties of exempt anaphors: unlike plain anaphors, they need not be locally or exhaustively bound and can give rise to strict readings; complementarity with pronouns, however, is not a robust property of anaphors. Furthermore, potential confounds are identified, which may obscure the logophoric conditions on exempt anaphors. First, exempt anaphors may be unacceptable for reasons independent of logophoricity: strong anaphors compete with weak elements; anaphors are unacceptable in positions construed with agreement. Second, other elements like intensifiers, which are not subject to logophoric conditions, may be mistaken for exempt anaphors.

The Logophoric A-Binder Hypothesis

This chapter defends a new analysis of exempt anaphors based on logophoric A-binding. The reduction of exempt to plain anaphors results from the presence of a logophoric operator in their spellout domain, which provides an A-binder: the subject of this operator, a covert logophoric pronoun syntactically representing the local logophoric center, locally binds exempt anaphors, which are thereby interpreted as logophoric. This hypothesis is supported by exhaustive coreference constraints on co-occurring exempt anaphors, which shows that they must be exhaustively bound by a local binder; the illusion of split or partial antecedence is due to the fact that their implicit binder need not be exhaustively bound. Although it is inspired by previous hypotheses, the logophoric A-binder hypothesis improves upon them by locating the source of difference between plain and exempt anaphors in their binders: it thus derives all properties distinguishing exempt from plain anaphors while reducing exempt to plain anaphors.

The Logophoric Properties of Exempt Anaphors

The goal of this chapter is to determine the interpretive constraints on exempt anaphors. After a review of the literature about the role of perspective and logophoricity on anaphora, the proposal is presented, based mainly on the behavior of French anaphors. First, several diagnostics show that an anaphor can be exempt if it occurs in a phrase expressing the de se attitude of its antecedent. Second, further diagnostics reveal that outside attitude contexts, an anaphor can be exempt if it refers to the empathy locus of its context. Third, it is shown that attitude holders and empathy loci are the only two types of logophoric centers that can antecede exempt anaphors; in particular, deictic centers do not create the relevant conditions for logophoric exemption. This implies that the notion of logophoricity relevant to exemption should be restricted to mental, first-personal perspective. This notion further promises to be relevant beyond anaphora.

Introduction

This chapter introduces the gist of the exempt anaphora issue. Anaphors are considered to obey locality requirements defined by Condition A of Binding Theory. But some instances of anaphors (exempt anaphors) seem to escape them and further differ from plain anaphors by licensing split antecedents, triggering strict readings and freely alternating with pronouns. Four types of solutions have been proposed: Condition A must be redefined to incorporate exempt anaphors; exempt anaphors undergo LF-movement to the domain of their antecedent, thus covertly obeying Condition A; exempt anaphors are in fact pronouns with specific interpretive properties; Condition A only concerns anaphors in co-argumental positions. But these solutions face methodological and theoretical issues: how to empirically distinguish between plain and exempt anaphors; and how to uniformly account for plain and exempt anaphors without postulating homophony. This book proposes an inanimacy-based strategy and the logophoric A-binder hypothesis to solve these issues.

Conclusion

This very brief last chapter concludes by summarizing the main results of the book and offering methodological suggestions for further research.