Retour vers liste

Détail de la contribution

Auteur: Ingmarie MELLENIUS

Co-Auteur(s): Maria ROSENBERG, Dept of Language Studies, Umeå University, Sweden

The Semantics of Compounds in Swedish Child Language

Abstract/Résumé: Compounding, in particular NN compounding, is a highly productive word formation process in Swedish. From the age of two years, Swedish children begin to produce compounds. They often produce creative coinages, which, for the most part, are context dependent. It is thus possible to stipulate that children’s use of compounds in particular situations draws from the need of naming. The objective is to examine the semantics of compounds (e.g. interrelations between constituents) produced by Swedish children. Are particular semantic relations more salient or frequent as compared to others? What is the status of the morphological head in the compounds? A Swedish compound can be defined following four criteria (cf. Mellenius 1997:32-33): i A compound is a complex lexeme which contains two or more lexemes (cf. Matthews 1974, Fradin 2003, Booij 2005). ii A Swedish compound is written as one word, without hyphens and spaces. iii The rightmost lexeme of the compound is the head: the compound is an instance of the head; the rightmost lexeme determines the compound’s gender and is declined for number, definiteness and case (cf. Williams 1981). iv Swedish compounds are pronounced with a particular intonation contour, characterized by two peaks (cf. Chomsky & Halle 1968). Štekauer’s (1998) theory of word-formation seems to suit the purpose of our study and will be evaluated according to our data. We will however take Grzega’s (2005) proposals to slightly revise Štekauer’s (1998) theory into account. Jackendoff’s (2009) and Lieber’s (2004, 2009) theories will also be taken into consideration in our theoretical discussion. Štekauer (1998) defines word-formation (clusters) as being one hundred percent productive, as well as regular and predictable. This definition does not seem to run counter to the creative coinages in child language. All of them are fully interpretable in the context in which they are produced. The corpus consists of NN compounds from four monolingual children. The compounds have been noted down together with information of the context in which they were produced. An amount of approximately 400 NN compounds have been attested and analysed. All of the compounds in our corpus correspond to non-established compounds in contemporary Swedish. Mellenius (1997) shows that the acquisition follows a developmental sequence: phonological properties (compound stress) are mastered around two years of age; a flat structure seems to be used before the age of three years, after which the head position (right-hand) begins to set; the semantics is the last parameter to be mastered. Consequently, the present study intends to shed more light upon the semantics of Swedish NN compounding in acquisition.