Is there a principled difference between entailments in natural language
that are valid solely in virtue of their form or structure and those that are not? This
paper advances an affirmative answer to this question one that takes as its starting
point Gareth Evans’s suggestion that semantic theory aims to carve reality at the
joints by uncovering the semantic natural kinds of the language I sketch an Evansinspired
account of semantic kinds and show how it supports a principled account of
structural entailment I illustrate the account by application to a case study involving
the entailment properties of adverbs this involves developing a novel proposal
about the semantics for adverbs like quickly’ and slowly In the course of the
discussion I explore some implications of the account for the place of modeltheoretic
tools in natural language semantics.

EB00000002750K

Available

• Structural entailment and semantic natural kinds