Factors in the selection of compound-types in spoken and written Hebrew

Abstract

This study investigates the factors that constrain and promote the selection of noun compound types in spoken and written Hebrew: Construct-state (e.g. anfey∧ha-ets ‘branches∧tree’, where ∧ stands for the bound genitive relation); analytic (e.g. ha-anafim šel ha-ets ‘the-branches of the-tree’, with the free genitive particle šel); and double anaf-av šel ha-ets ‘branches-its of the-tree’ with two genitive markers: a bound suffix and šel. Three types of data were examined, one spoken and two written (journalistic and school textbooks). Lexical, semantic, syntactic and pragmatic analyses revealed that construct-state compounds are the default form for expressing classifying relations, while analytic compounds typically denote possessive and partitive relations. Syntactic complexity promotes the selection of analytic compounds in written language owing to their relative transparency. Double compounds are the most marked of the three, and are very restricted owing to the close relationship between a bounded adjunct and its head. Implications of these findings for language acquisition and the development of literacy are discussed.