Objects and the Grammar of Countability

This book proposes an object-centred and contextualist account of the count/mass distinction. The main finding is that there is a connection between variation in count/mass lexicalization patterns and the ways in which nouns can diverge from manifesting the canonical grammatical reflexes expected of the count/mass distinction in a given language.

This book proposes an object-centred and contextualist account of the count/mass distinction. The main empirical finding is that there is a connection between variation in count/mass lexicalization patterns across and within languages and the ways in which mass or count nouns can diverge from manifesting the canonical grammatical reflexes expected of the count/mass distinction in a given language. Assuming compositional semantics enriched with Classical Extensional Mereology, Peter Sutton and Hana Filip propose that mass-count pairs of nouns (across and within languages) lexicalize the same number-neutral core property that is satisfied by objects, which comprises not only one discrete entity separate from other entities, but also a discrete entity based on its affordances and topological properties. For a concrete common noun to be grammatically count, as needed in counting and quantificational constructions, it must specify a quantized set of objects relative to a context of individuation. Context plays a crucial role, since it can afford different ways of viewing the entities in a noun''s extension, and so what counts as ''one'' for that noun in that context. The book outlines a system of three perceptual-interactive constraints, based upon how we perceive and interact with objects, that predicts the propensity for a number-neutral core property that is satisfied by objects to nonetheless be lexicalized as mass. These constraints are also key to explaining variation in the count/mass lexicalization patterns of such properties. The authors also show that the extent to which a core property can satisfy these constraints can be estimated from a corpus, and how these corpus measures can be used to model variation in count/mass lexicalization patterns. The main data relies on concrete common nouns in number-marking languages, namely Czech, English, Finnish, and German, but the account is also extended to abstract nouns, and to two non-number-marking languages, namely Mandarin and Yudja.