This contribution presents arguments for a split C-domain on the basis of Dutch and German dialectal data and shows parallels in head-final Asian languages which lack a left-peripheral C-domain. In question complements C can split up into a disjunctive head and a subordinative head, or even into the wh-part, a disjunctive head and a subordinative head. This argues in favor of a close link between disjunctive questions and constituent questions and is in fact supported by the semantics of questions, in particular by the partition approach. The last part of the paper explores how illocutionary force features are established. It is shown that verb-second (I-to-C movement) is a central but not the only device to prepare the clause for the acquisition of force features (which are absent in dependent V-final clauses), but does not determine them. Additional mechanisms govern the ultimate pragmatic interpretation.