Non-de dicto construals as a unified phenomenon

The interpretation of expressions in the scope of intensional operators μ is not always relativized to the quantificational domain of μ. The literature has identified several such phenomena, subsumed here under ‘non-de dicto (NDD)’- construals. Based on novel data, we argue contra most existing accounts that (i) different kinds of NDD-construals (der re and transparent construals of predicates) must be derived by the same mechanism, (ii) this mechanism should involve replace- ment of meanings of the sub-constituents of the clause embedded by the intensional operator, (iii) replacement is constrained in that different replacements cannot lead to different answers to the QUD, and (iv) NDD-construals are not grammatically individuated, i.e., they do not ‘live off’ a structure distinct from the one that yields us meanings without NDD. We then sketch a proposal that captures these properties.