The standard Hintikka semantics of clausal embedding treats the
complementizer "that" as vacuous, and takes the meaning of the CP to simply
inherit the meaning of the TP. We present data from CP conjunction and
disjunction in Russian and Hebrew to show that this is wrong. In Russian
and Hebrew CP con/disjunction doesn't have the same interpretation as the
corresponding TP con/disjunction under a single C. Adopting the
decompositional approach to attitude reports initiated by Kratzer (2006), we
develop an analysis that makes crucial use of the **equality** hypothesis
(Moulton 2015, Elliott 2017): the content of the event expressed by an
attitude predicate equals the proposition expressed by the embedded
sentence. Complementizers, on our analysis, contribute the "equality" part.
We show how the Russian and Hebrew data fall out from this analysis.
English and Italian behave differently from Hebrew and Russian with respect
to CP conjunction. We hypothesize why that might be based on the
observation that in English and Italian, unlike Russian and Hebrew, CPs
have a more flexible syntactic distribution: they can appear bare in
subject position. We therefore propose that CPs in English and Italian are
more type-flexible, \ being able to denote type-e elements and be conjoined
by a non-Boolean *and*.
*Tanya Bondarenko and Itai Bassi*