The acquisition of nominal compounding in Swedish

Abstract: In Swedish, as in other North Germanic languages, compounds are very common and a majority of the novel words in Swedish are compounds. This thesis aims at a presentation of compounding in Swedish and the problems involved in describing compounding, but mainly at an empirically founded description of children's creation and use of novel compounds and developing comprehension of compounds, especially nominal compounds. A Swedish compound contains two ore more word roots that elsewhere may function as independent words. The rightmost word is the head, which means that it determines the compound's gender and is declined for number, definiteness and case, and that the whole word is an instance (typically a hyponym) of the head. The compound is written as one word, and is pronounced with a particular intonation contour, characterized by two peaks. A noun that functions as modifier in a compound sometimes appears in a particular "liaison form". The empirical findings concerning children's acquisition of compounding are based on the following sources: (a) a corpus of around 340 spontaneous novel compounds collected from two children; (b) analysis of 258 novel nominal compounds produced by 10 children (aged between 3;5 and 6;8), who were encouraged to label picture cards.; (c) the results of a pictureidentification experiment where 60 children (aged between 2;0 and 5;4) were asked to interpret 24 novel nominal compounds; and (d) the paraphrases of 18 novel compounds that 70 children were asked to give, 4 years in succession. Some observations and results are that the compound stress pattern is established before my period of investigation, i.e. before age 2; that proper identification of the head is established around age 3; that the deletion of a final -a in a modifying noun is the first liaison form mastered by children; and that many different semantic relationships are attributed by the children to hold between the head noun and the modifying noun.