The source of the <em>de re</em> blocking effect

Abstract

This paper presents novel empirical data from English in which the de re blocking effect – that an obligatorily de se anaphor cannot be c-commanded by its de re counterpart – is attested. I argue that at least some instances of the de re blocking effect cannot be accounted for via syntactic accounts, contra Percus &amp; Sauerland (2003b) and Anand (2006). Via experimental evidence, it is shown that the de re blocking effect takes place with passives and across clause boundaries with certain predicates but not with others. I argue that a novel thematic theory of the de re blocking effect together with the aforementioned syntactic theories would be able to account for the full range of facts.