Pseudosamordningar i svenska särskilt sådana med verben sitta, ligga och stå

Sammanfattning: This is a study of pseudo-coordination in Swedish, a common phenomenon in Swedish and other Scandinavian languages, formally a regular verb (phrase) coordination [V1 CONJ V2] but with some special characteristics: the verb in V1 is a) taken from a restricted set of verbs; b) semantically weak; and c) typically contributes aspectual or modal properties to the construction as a whole. An example is Anton sitter och läser, lit. ‘A. sits and reads’, i.e. ‘Anton is reading’.Various criteria have been suggested to identify these constructions as a category distinct from other verb coordinations: single event interpretation; a single phrasal accent; different behaviour of object extraction and the scope of negation, etc. These criteria prove less useful when applied to authentic text, most importantly as they do not unambiguously tell pseudo-coordinations from other verb coordinations. In this thesis, it is suggested that it is the relation between the two verbs that is crucial. This hypothesis is tested in Study 1.In Study 1, 15310 verb coordinations in a tagged corpus (Parole) were investigated. 4281 of these had a potentially pseudo-coordinating verb in V1, and were classified for the relation between V1 and V2. Among the 3881 cases of actual pseudo-coordinations, the verb relations were BACKGROUND and GOAL DIRECTION. These relations were not found in regular coordinations with potentially pseudo-coordinating verbs in V1, which provides us with a criterion to identify pseudo-coordination in text.In Study 2, the relation criterion was applied to the verbs sitta ’sit’, ligga ’lie’, and stå ’stand’, studied in a corpus of news text (Press 97). Constructions were classified with respect to animacy of subject, type of V2 verbs, and relative frequency and position of adverbials. The V1 verbs were found to form a coherent category but also exhibited interesting differences, both from each other and from their uses outside pseudo-coordination, along all parameters investigated.