To shift or not to shift

There are two main competing views about the nature of sign language role shift within formal semantics today: Quer (2005) and Schlenker (2017a,b), following now standard analyses of indexical shift in spoken languages, analyze it as a so-called ‘monstrous operator’, while Davidson (2015) and Maier (2017), following more traditional and cognitive approaches, analyze it as a form of quotation. Examples of role shift in which some indexicals are shifted and some unshifted pose a prima facie problem for both approaches. In this paper, we propose a pragmatic principle of attraction to regulate the apparent unshifting/unquoting of indexicals in quotational role shift. The analysis is embedded in a systematic empirical investigation of the predictions of the attraction hypothesis for German Sign Language (DGS). Results for the first and second person pronouns (ix 1 and ix 2) support the attraction hypothesis, while results for here are inconclusive.