The aim of this collaborative research (between my grant #1424054 and Dominique Sportiche's grant #1424336) is to develop typologies and theories of exempt and non-exempt anaphora based on new criteria to distinguish them.
Anaphors (like English himself) sometimes obey condition A of the binding theory, sometimes are exempt from it, e.g. in some (syntactic or discourse) contexts. Given this dual behavior, a first task is to find independent criteria separating instances of exempt behavior from instances of non exempt behavior, and decide in each instance if an anaphor behaves in an exempt fashion, or not, or crucially, both ways.
The point of departure of this project is our research on French (see Charnavel & Sportiche 2016) which, using a simple and clear criterion ((in)animacy), reveals that, contrary to most dominant proposals on Condition A, (i) exempt and non exempt behaviors do not partition syntactic positions and (ii) exemption is in fact not a matter of syntactic positions at all. Rather exemption depends only on perspective or point of view related conditions on the antecedents of exempt anaphors.
These results make it necessary to systematically reassess all data grounding current theoretical proposals on Condition A and on exemption and formulate new theories on the basis of this reassessment, with a view to understanding crosslinguistic variation, if any.
Regarding Condition A, this means characterizing the notion of locality involved and questioning and exploring whether it is in fact anaphor dependent, as some literature suggests (viz. the reported difference between SELF anaphors and SE anaphors). Regarding exemption, as the notions of point of view or center of perspective are intuitive but not clearly defined and are in fact diversely used in the literature, it is necessary, once it is known which anaphors are exempt, to develop precise diagnostics to determine the properties of such exempt anaphors and determine whether exemption is a unitary phenomenon or not, that is build a typology of exempt anaphors.
We are exploring these questions in a small set of otherwise well described languages (e.g. French, English, Mandarin, Scandinavian) to develop these criteria and diagnostics both via traditional grammatical judgments and through larger scale web based magnitude estimation tasks. We are also in the process of extending the investigations to as broad a set of typologically diverse languages as possible (e.g. in Korean, Hebrew, Greek, Turkish and Uyghur).
See some of our published results here.
See the presentation of our past workshop here.