Univerbation

Abstract

Abstract Univerbation is the syntagmatic condensation of a sequence of words recurrent in discourse into one word, as when the Spanish combination a tras (to back) becomes atrás ‘behind’. It affects both lexemes and grammatical formatives. Unlike processes of word formation, including conversion of a syntactic construction into a word, as in forget-me-not, and compounding, as in Spanish lavaplatos ‘dishwasher’, univerbation is a spontaneous process. There are two main types of univerbation: phrasal univerbation downgrades a phrase to a word, as when Latin terrae motus ‘earth’s movement’ becomes Spanish terremoto ‘earthquake’. Transgressive univerbation coalesces a string of words which do not form a syntagma into a word, as when French par ce que becomes parce que. A set of univerbations may share structural features and may therefore evolve into a pattern of compounding. Thus, blackbird originated by univerbation, but may now provide a pattern of compounding. As a consequence, univerbation and compounding are not always easily distinguishable. The discussion uses empirical evidence adduced in earlier work, mostly from Romance and Germanic languages. Its aim is not to present novel phenomena but to provide a theoretical background for the phenomenology and improve on available analyses.