Leksikalisering av gammelnorske sammensetninger

Abstract

In linguistic theory compounds are recognized as occupying an intermediate position between the syntax and the lexicon. They share features both with typical syntactic expressions and with typical lexical expressions. The process by which compounds gradually develop from largely syntactic items to typical lexical items is in this paper termed the lexicalization process, and a hypothetical model of this process is presented. This hypothesis is then tested at one specific point by the use of an Old Norse material of noun + noun compounds. If the lexicalization model is a true reftection of the way compounds actually behave, one can predict that compounds which are lexicalized to a large extent, probably will show a larger tendency toward being written as one graphic unit (i.e. without a space), than those which are lexicalized to a !esser degree. The Old Norse material confirms this prediction, and the lexicalization model is thereby strengthened as a hypothesis.