When agents disappear: on the morpho-syntax, semantics and informational value of passive, middle and active impersonal constructions

The project investigates passive and impersonal sentences in Finnish and Swedish, using the methodology of Magnitude Estimation. The focus is on borderline cases: if a sentence displays the key properties of passive sentences, is that enough to call it a passive? What are these key properties and how many of them need to be present? How well-formed are these sentences in native speakers' minds?

Although passivization has been a core object of study in many languages and within various theoretical frameworks, no consensus exists on what should count as a passive in each language or framework. The boundaries between passives and many passive-like constructions, such as impersonal actives and middles, are also difficult to draw: it is therefore not uncommon that a sentence is called a passive by one author, but a non-passive by another. The purpose of this project is, first, to look data from various languages and linguistic frameworks, and define a key cluster of properties for passives/non-passives. With these criteria in mind, we look at some borderline cases in more detail, and try to decide if a passive, a non-passive, or both analyses are available. A case in point is the status of the periphrastic passive, which in many languages looks (almost) identical to the copular (adjectival) construction. This has sometimes been used as an argument for saying that there is no periphrastic passive in the language, and that all such sentences are, in fact, examples of the copular construction. A further aim of the project is to collect data and judgments using the methodology of Magnitude Estimation. As many of the borderline cases are sentences which are, prescriptively, claimed to be ungrammatical, the purpose is to establish exactly how well-/ill-formed such sentences are in native speakers' minds, and how they compare with normal passive and active sentences.