Constraints on Resultative and Depictive Modification

According to Halliday (1967), there is a ban on simultaneous resultative and depictive modification in English. It is here observed that such a ban is not general, but instead restricted to certain examples of modification with abnormal theta relations, which are commonly found in the resultative construction. Otherwise, simultaneous resultative/depictive modification is possible. An account of this asymmetry is given in terms of the theory of complex predicates (Neeleman 1994). Complex predicate formation is necessary just in case the resultative predicate lacks an argument position for its theta role. This situation obtains in resultative constructions with fake reflexives, for example. In such cases, the main verb must license the direct internal argument of the resultative construction through the indirect assignment of the result predicateÕs theta role. For this purpose, the result predicate must be base generated as sister to the main verb. The theta role of a depictive predicate however, must be assigned to the external argument. For this purpose, the depictive predicate must also be base generated as sister to the main verb. Therefore, given that complex predicate formation is necessary, simultaneous depictive/resultative modification is impossible. This account assumes strict binary branching.