Consider the structure in (1) from the Dravidian language Tamil below:

Abstract The paper focuses on an interesting form of (person) indexical shift in the Dravidian language Tamil which surfaces as 1SG agreement marking in a clause embedded under a speech predicate. I show that this agreement is an instance of indexical shift and label it “monstrous agreement”. However, I demonstrate that its full range of empirical properties cannot be adequately explained by the major analyses of indexical shift in the literature. The bulk of these, I argue, in addition to being predominantly semantic in spirit, and thus ill-equipped to deal with a morphosyntactic phenomenon like agreement, also involve two core misconceptions regarding indexicality vs. logophoricity on the one hand and speech vs. attitude predicates on the other. I propose that these core assumptions be strongly re-evaluated from first principles and that syntactic and typological clues on the subject be paid more heed. I propose a new analysis of the Tamil paradigms which derives indexical shift within an enriched grammatical model involving contextual features instantiated in a structurally articulated cartographic left periphery