Anaphora: Insights from Sign Language (Summary)*

Sign language anaphora is realized very differently from its spoken language counterpart. In simple cases, an antecedent is associated with a locus in signing space, and an anaphoric link is obtained by pointing towards that locus to recover its semantic value. It has been argued that this mechanism is an overt realization of the device of coindexation in formal syntax and semantics. Summarizing some of our recent work, we discuss two kinds of insights that sign language research can for this reason bring to the foundations of anaphora. First, in some cases the overt nature of indices in sign language makes it possible to bring overt evidence to bear on classic debates in semantics. We consider two: the first concerns the availability of situationdenoting variables in natural language (we argue that pointing signs can be situation-denoting in addition to being object-denoting); the other concerns the availability of binding without ccommand (we suggest that sign language provides overt evidence for such a possibility). Second, in some cases sign language pronouns raise new challenges for formal semantics. We argue that the iconic component of sign language loci is a case in point; in the end, we suggest that loci may be simultaneously formal variables and simplified depictions of what they denote – which requires the construction of a formal semantics with iconicity to analyze their properties. Note: The present piece aims at summarizing the results of Schlenker 2011a, b, c, 2012, to appear, Schlenker and Lamberton, to appear, and Schlenker et al., to appear; it heavily draws on these earlier articles in form and as well as substance. In this summary of some of our recent research on sign language, we discuss three cases in which sign languages can bring new insights into the foundations of anaphora. These cases are of two sorts. –First, we re-visit two classic semantic debates. One concerns reference to times and worlds (or situations); the other concerns the nature of variable binding. We suggest that in both cases the realization of sign language pronouns provides direct and overt evidence in favor of certain facts that must be inferred rather indirectly in most spoken languages. –Second, we suggest that some iconic properties of sign language pronouns suggest a renewed view of semantic interpretation: in some cases, some geometric properties of signs appear to directly reflect properties of their denotations, which argues for a condition of 'formal iconicity' at the very core of a formal semantic account. 1 Background: Pronouns in Sign Language Pronouns in sign language have been the object of considerable interest, in part because their realization seems to be so different from that of their spoken language counterparts: in simple cases, sign language pronouns are realized by way of pointing. For present purposes, one typical mechanism involves two steps: (i) first, an antecedent establishes a position in signing space, called a ‘locus’; (ii) second, the pronoun is realized by pointing back towards this locus (but see for instance Schlenker 2011c for discussion of alternative realizations of anaphora). In (1), we see examples of locus assignment to expressions of direct reference (here proper names), definite descriptions and indefinite descriptions (only meaning can distinguish between the latter two). * The present work was supported by an NSF grant (BCS 0902671) and by a Euryi grant from the European Science Foundation (‘Presupposition: A Formal Pragmatic Approach’). Neither foundation is responsible for the claims made here. The research reported in this piece also contributes to the COST Action IS1006.