Uncovering the extent of the phraseological tendency: Towards a systematic analysis of concgrams

This paper offers an analytical procedure for identifying phraseological variation within 'concgrams' (Cheng et al. 2006), which are sets of words that co-occur regardless of constituency variation (e.g. AB and A (*) B), positional variation (e.g. AB and BA), or both. It argues that examining concgrams takes us closer to more fully appreciating and understanding the idiom principle (Sinclair 1987) which underpins the claim that phraseology is at the heart of all language use. Central to a description of phraseology is the identification of 'meaning-shift units' (MSU) (Sinclair 2007a, 2007b) and the analytical procedure for concgrams described in this paper can lead to their identification. In this paper, the concordance lines of a two-word concgram, PLAY/ROLE, are analysed to identify all of the possible concgram configurations and their frequencies of occurrence. Based on frequency, the canonical form is identified and its meaning described. The canonical form then serves as the benchmark against which all of the other concgram configurations are compared, resulting in a ranking of the concgram configurations relative to their adherence to the canonical form. At the end of the process, a meaning-shift unit is identified and described with all of its potential variations; in other words, a paraphrasable family with a canonical form and patterns of co-selection. Lastly, this paper proposes initial theoretical statements to account for key phraseological patterns so far observed, and explores the implications of a shift in emphasis towards descriptions of phraseological variations for the field of applied linguistics.