An alternative model of indexical shift: Variation and selection without context-overwriting
Sandhya Sundaresan
July 2018

Indexical shift is the phenomenon where an indexical is interpreted, not against the utterance context, but against the index associated with an attitude. Context-overwriting and quantifier binding are two of the main strategies adopted in the literature to derive this phenomenon. In the former, all indexicals are fundamentally shiftable; indexical shift obtains when a contextual operator (the "monster") overwrites the utterance-context with the intensional index. In the latter, indexical shift is effected by a contextual quantifier (the monster) which binds free context variables associated with indexicals in its scope. But some indexicals are simply specified not to be bindable, thus never shift. Others can optionally shift, while yet others must always be bound, thus always shift. Since the utterance-context is entirely overwritten, context-overwriting approaches predict that it should never be possible for some indexicals alone to shift, while others remain unshifted, under a monster. Thus, all indexicals in an intensional domain should shift, or none should --- a generalisation termed the Shift Together restriction. In contrast, under the quantifier binding approach, an indexical may "decide for itself" whether it can be bound by the contextual quantifier or not. As such, Shift Together is never predicted. The goal of this paper is to present a range of evidence showing that neither approach is valid as it stands, and to develop a new model of indexical shift that addresses these issues. To this end, I present three types of evidence: (I) Systematic exceptions to Shift Together in Tamil, varieties of Zazaki and Turkish, and potentially also Late Egyptian; (II) novel evidence from imperatives in Korean and supporting secondary data from imperatives in Slovenian, showing that the utterance context continues to be instantiated even in putatively shifted environments; and (III) results from personal fieldwork in Tamil dialects and secondary data from 26 languages (from 19 distinct language families) showing that there is a one-way, implicational selectional variation in the intensional environments where indexical shift obtains. The following desiderata emerge: 1. Shift Together holds whenever possible, but systematic exceptions may obtain when it cannot; 2. the utterance-context is never overwritten; 3. indexical shift is an embedded root phenomenon that privileges speech predicates. To capture these, I develop an alternative model of indexical shift which combines insights from both context-overwriting and quantifier binding approaches, and adds new insights to these. I propose that all intensionality is executed by a monster, which comes in different contextual shapes. This new type of monster is a contextual quantifier, like under the quantifier binding approach, but like in the context-overwriting model, it is nevertheless severed from the attitude verb. I present further evidence to argue that this monster is encoded on structurally distinct types of C head along the clausal spine, each under the scope of a different class of attitude verb.
Format: [ pdf ]
Reference: lingbuzz/004115
(please use that when you cite this article)
Published in:
keywords: indexical shift, attitude, attitude predicates, context, complementizer, monstrous agreement, semantics, syntax
previous versions: v1 [July 2018]
Downloaded:389 times


[ edit this article | back to article list ]