THE ACQUISITION AND DEVELOPMENT OF SWEDISH AS A FIRST AND AS A SECOND LANGUAGE: THE CASE OF CLAUSE COMBINING AND SENTENTIAL CONNECTORS. Åke Viberg ABSTRACT. This paper is based on data from three related projects on Swedish as a second language, in which the learners represent different ages (from five-year-olds to adults). Since even native controls have been recorded, it is also possible to shed light on age-related differences in language development. Two examples of developmental trends will be discussed, both of which are concerned with different aspects of clause combining. The first one is the development of the system of sequential markers corresponding to (and) then in English. In Swedish, there are three frequently used equivalents of then with regard to this function: sen (sequential 'then'), då (simultaneous 'then') and så (sequential 'so'), the latter of which in Swedish is more or less freely interchangeable with sen. Learners have a strong tendency to simplify the system by favouring the most salient marker, namely sen. The development towards the native pattern is in general a protracted process due to language-specific characteristics of the target language. In addition, there seem to be age-related differences in the usage of sequential markers. Children (both NS and NNS) have a stronger tendency than adults to use such markers. The second example concerns the role of lexical factors in syntactic acquisition, in particular the role of nuclear verbs as syntactic prototypes. Sentential complements have a strong tendency to emerge after a small range of verbs. This tendency is strongest for WH-S complements, which tend to be used with veta 'know'. That-S complements are primarily used after tro 'think' as an expression of propositional attitudes and after säga 'say' in indirect reports. 1 Introduction In Sweden, as in other Western European countries, migrants and their children acquire the majority language as a second language under many different conditions with respect to variables such as the type of contact with native speakers, instruction and the age at which the learning commences. The achievement of levels of proficiency approaching the upper end (near-)native proficiency is typically extended over a long period of time and it is obvious that the Swedish of both native and non-native speakers continues to develop and change across the entire lifespan (see Gleason ed. 1989 for a lifespan perspective on first language development and Hyltenstam & Obler eds. 1989 for a similar perspective on bilingual development.) This is perhaps easiest to observe in children as they continue through preschool and school. At the same time, the proficiency level of non-native children is difficult to assess once they have reached a certain level due to the fact that even the Swedish of their native peers continues to develop and change throughout the school years. Longitudinal studies of L2 production have often been focused around the intial acquisition of linguistic elements and the emergence of specific forms in the speech of the learners. Even if this is an important step towards full proficiency, many structures continue to develop and this development can sometimes only be 1