Non-maximality and vagueness: Revisiting the plural Sorites paradox

This paper is an attempt at a synthesis of two superficially conflicting approaches to non-maximality: the issue-based approach (Malamud 2012, Križ 2015, Križ & Spector 2021 a.o.), which generates clear-cut truth conditions once the issue parameter has been fixed, and the strict/tolerant approach (Burnett 2017 a.o.), on which non-maximal construals involve vagueness. I argue that there are two classes of contexts that license non-maximality. One of them gives rise to the Sorites paradox once the non-embeddability of non-maximality is controlled for. The other class does not license vagueness at all. To model this distinction, I introduce a formal framework that combines the issue-based approach with the notion of strict and tolerant truth conditions (Cobreros, Egré, Ripley & van Rooij 2012a), which are defined via super-/subvaluation over different issues. This system provides two sources of non-maximality, only one of which involves vagueness.