Summary

The principal aim of the present study, called Cause and Agent, is the description of passives in Dutch grammar.

It is stated in an introductory chapter that the proposals from Chomsky (1957) till Fiengo (1974) to solve the problems of English passives fail to account for certain phenomena in Dutch. These phenomena are connected with action-sentences and adverbials of cause beginning with door.

In the second chapter a survey is given of the syntactic and semantic properties of verbs which can occur in passive sentences. Among them we find the sub-class of the so-called pseudo-activity verbs, like duperen (injure), overtuigen (convince), creëren (create), verhinderen (hinder), verkleinen (diminish, reduce), voorkomen (prevent) etc. Those verbs are at times to be interpreted as real activity verbs, at other times the verbs are stative. They can be combined in their activity sense with door-adverbials of a limited scope. Although a causal adverbial commonly has the rest of the sentence as its scope, denoting the cause of the state or happening described in the other part of the sentence, the subject of the sentence Door hun maatregelen voorkwamen de doktoren een ramp (By their measures the doctors prevented a disaster), is to be interpreted outside the scope of the adverbial. The measures are not the cause of the doctors' preventing a disaster.

In the first part of the third chapter a proposal is made for a projection rule that can interpret causal door-adverbials, accounting for their hierarchy if there are two or more of them in one clause. In the second part we try to account for passive phenomena in Dutch, taking Fiengo's proposal as a starting point. Fiengo's frame-work turns out to be insufficient to account for the passive sentences with intransitive verbs, for the passive sentences with door-phrases of a limited scope and for the passive structures (complements and nominals) without the passive auxiliary worden or zijn (be). These phenomena coerce us into an investigation of action sentences in the fourth chapter. An attempt is made to show that Dutch grammar has a base-rule NP → VP. A NP dominating a VP represents the underlying structure of activity gerunds, the meaning of which is called ‘act-type’ (Rescher 1967; 1970). According to Ross (1972) the pro-verb doen (do) is present in the underlying structure of action sentences. Unlike Ross, however, who could not describe relative sentences like the dancing she did, we argue that the object of doen (do) has the structure of a NP dominating a VP. The analysis is completed with a transformation T-handeling in substitution of the rule Do-gobbling of Ross. The doen-analysis also applies to sentences of a certain type in which the subject is non-human. An example is De spons zoog het water op (The sponge sucked up the

water). The joint factor in the action sentences and the sentences of this type is called ‘verrichting’ (operation). An interpretive rule, the ‘verrichtingsinterpretatie-regel’ (operation interpretation rule), is formulated delimiting the set of sentences and gerunds which get an operation interpretation.

The remaining problems of the third chapter are partly solved in the fifth. The solutions are based on the analysis of action sentences in the previous chapter. As an extra we find a plausible explanation for the fact that intransitive verbs like komen (come), gaan (go), i.e. verbs denoting a change of place, do not passivize, whereas other intransitives do. The former verbs share properties with intransitives like breken (break), smelten (melt), i.e. verbs denoting a change of state, and not passivizable either. We get suitable results in these cases by means of the interference of ‘property interpretation’ (Fiengo 1974) and the operation interpretation rule.

In the last section of this chapter new problems arise concerning the status of the passive door-phrase in nominal constructions like Shakespeare's Macbeth door the Old Vic (Shakespeare's Macbeth by the Old Vic). Therefore we seek in the sixth chapter to develop the passive proposal by Bresnan (1972). Combining the results of the analysis of the causal door-phrases with and without a limited scope, we try to state interpretive rules to account for the meaning of passives and for the synonymy of active and passive sentences. The difference between passive door-phrases and causal door-phrases is described as a difference of scope. Their meaning is described in terms of direct and indirect causation. The rules make Agent-postposing and lexical subject selection restrictions superfluous. However, we meet difficulties when the results of this analysis are applied to passive nominals. Too little is known of the semantics of these structures to solve satisfactorily the problems connected with the occurrence of a passive door-phrase in the absence of the passive auxiliary.