Abstract
John Hawthorne and David Manley (The Reference Book, OUP 2012) endorse a unified treatment of the semantics of four kinds of expressions that can be said to have referential uses: specific indefinite descriptions, definite descriptions, demonstratives and proper names. The semantic theory the authors propose treats all these expressions as having a quantificational structure that achieves uniqueness of application via the presence of covert material contributing to the restriction of the domain of quantification. I argue that there are reasons to postulate disunity; reasons to set apart a class of purely referential expressions that do not just apply uniquely to an object via overt or covert restrictors, but that refer to it.