Searching for Universals in Compounding

Abstract: Despite the fact that compounding is the most widespread word-formation strategy in the world’s languages – and for some languages, the only one – compounds have received little, if any attention in linguistic typology and in particular in studies on linguistic universals. This paper presents a preliminary crosslinguistic overview of the basic features of compounding, their distribution and expected degree of variation on the basis of the Morbo/Comp database. A basic conceptual framework of four important points for the study of compounding from a typological point of view is described, comprising the definition and classification of compounds, the position of the head constituent and the definition of compound type. These issues necessarily have to be dealt with prior to the actual examination of the data, for they play a fundamental role in the design and implementation of the research methodology. We attempt to extrapolate universal tendencies in compounding with respect to the following list of characteristic features: output categories, input categories, the relation between the constituents, headedness and the combination of categories. A series of universal hierarchies describing the world’s languages preferences in the domain of compound formation is proposed. Some well-known putative universals in compounding are discussed and rejected on the basis of our data, namely whether compounds are only right-headed, the root-compounding parameter and the head as locus inflectionis.