Passive and passive-like constructions in English and Polish

This dissertation deals with a family of grammatical constructions which overlap with the meaning of the basic passive and which can be seen as resulting from alterations in grammatical voice. In particular the thesis describes and analyses several related types of construction including the passive, the impersonal, and the anticausative. The main objective of this dissertation is a study of variation within the rich inventory of Polish passive-like constructions and a comparison with the corresponding constructions in English. It presents a detailed account of how the constructions vary within each language and between the languages. The main result of this research is a clear demarcation of the relevant constructions which are related in form. In Polish, the participial form is used by both the passive and the impersonal, and the reflexive form is used by the impersonal and the anticausative. Thus, a single form class subsumes different morpholexically derived constructions. Due to this fact, as well as to a serious descriptive bias in favour of the passive, all formal syntactic frameworks hitherto developed have misanalysed and misclassified the Polish -no/-to impersonal as an ‘ill-behaved’ impersonal passive. The present dissertation attempts to counter the prevailing bias and uses a detailed analysis of the morphosyntactic behaviour of the constructions in question as the basis for their demarcation. Furthermore, morphosyntactic criteria also separate valency-changing constructions from non-derived constructions which have only a conventional interpretation coinciding with the interpretation of the passive. The distinctive properties of the morpholexically derived constructions are attributed to operations in the argument structure which affect the meaning component of the predicate, the assignment of grammatical functions to the predicate’s arguments, or the realisation of the arguments in surface syntax. These hypotheses are modelled using a formalism derived from LFG’s Lexical Mapping Theory. The contrast between Polish and English, highlighted throughout the work, also illuminates certain peculiar properties of English constructions which have had a disproportionate influence on the analysis of the passive and valency-changing operations. The distinction between active and passive provides an example of a verbal category which seems designed to confuse our usual habits of thought: it appears necessary, and yet many languages do not have it; simple, and yet we have great difficulty in interpreting it; symmetrical, and yet it abounds in conflicting expressions. . . . Emile Benveniste 1950