This paper explores the role that the scalar properties and presuppositions of even play in creating polarity sensitive even meanings crosslinguistically (henceforth EVEN). I discuss the behavior of three lexically distinct Greek counterparts of even in positive, negative, subjunctive sentences, and polar questions. These items are shown to be polarity sensitive, and a three-way distinction is posited between a positive polarity (akomi ke), a negative polarity (oute), and a 'flexible scale'even (esto) which does not introduce likelihood, but is associated with scales made salient by the context. The analysis is a refinement of Rooth's original idea that negative polarity is involved in the interpretation of English even, and establishes further that the "negative" polarity domain of EVEN includes a sensitivity that is not strictly speaking negative (flexible scale esto). The distributional restrictions of EVEN items are shown to follow from distinct presuppositions (positive polarity and flexible scale EVEN), or from their lexical featural specification (negative polarity EVEN), a result that squares neatly with the fact that ill-formedness is systematic pragmatic deviance in the former case but robust ungrammaticality in the latter. This result supports the by now widely accepted view that polarity dependencies are not of uniform nature, and that we need to distinguish presupposition failures (which are weaker and possibly fixable in some contexts) from cases of ungrammaticality which are robust and cannot be fixed in any context (Giannakidou, 2001).
Natural Language & Linguistic Theory provides a forum for the discussion of theoretical research that pays close attention to natural language data, so as to provide a channel of communication between researchers of a variety of points of view. The journal actively seeks to bridge the gap between descriptive work and work of a highly theoretical, less empirically oriented nature.
Springer is one of the leading international scientific publishing companies, publishing over 1,200 journals and more than 3,000 new books annually, covering a wide range of subjects including biomedicine and the life sciences, clinical medicine, physics, engineering, mathematics, computer sciences, and economics.
This item is part of a JSTOR Collection.
For terms and use, please refer to our Terms and Conditions
Natural Language & Linguistic Theory
© 2007 Springer
Request Permissions