A note on impersonal passives

Chomsky (I98I: 124) remarks that 'the unique property of passive morphology is that it in effect "absorbs" Case'. The idea here is that a passive verb lacks the Case-assigning ability of the related active verb, and that as a result when a passive verb appears in a structure, the NP which would be assigned Case by the active verb must move into subject position to acquire Case. The impersonal passives that occur in many languages appear to pose a problem for this conception since they involve a passive verb but no movement of an NP into subject position and thus suggest that a passive verb can sometimes assign Case just like the related active verb. Jaworska (I986) sketches an analysis of impersonal passives in terms of socalled inherent Case, which does not necessitate this conclusion. In this note, I will argue that this analysis is unsatisfactory for the Polish impersonal passives which Jaworska considers. It seems, then, that passive verbs can sometimes assign Case, and thus that passive morphology does not necessarily remove the ability to assign Case. Jaworska analyzes inherent Case as case which is specified in the subcategorization frame of a lexical item and hence present in D-structure.2 Given this conception of inherent Case, a verb with an NP complement can have the ability to assign Case but not utilize it because the complement has inherent Case. When such a verb bears passive morphology, there will be no need for the complement to move into subject position. The result will be an impersonal passive. This, then, is how Jaworska suggests that impersonal passives should be analysed. A very similar analysis is proposed in Belletti (I988) in connection with German impersonal passives.3 I will show that an