Register variation and the multi-word item

Abstract

Past research in language has posited numerous reasons for language variation. The findings of such research seem to have focused more on how language varies in different contexts, where contexts cover factors of geography, ethnicity, social class, age, sex and culture on the one hand, and situational variables such as mode, domain and tenor on the other. But beyond these, a new paradigm has emerged with the advent of computer-assisted language research, which has primed investigations into similarities and dissimilarities between register types based on various linguistic phenomena. One of the central issues in such research is lexical behaviour, which is the main interest of this research, in particular, the use of multi-word items which are postulated to characterize individual registers. This study focuses on this relatively new area of linguistic enquiry, to account for how a text type can be distinguished based on the word sequences which compose it. The question at issue is - are there genre-specific multi-word items? To answer this question, the paper investigates the use of multi-word items in five different texts: fictional writing, academic writing, religious speech, political speech, and memorial speech. The results show some indication that the multi-word item can serve as an index of text variability, but precisely which multi-word items are expected to occur in a register, or not occur, remains inconclusive. We therefore recommend further and continual research into the use of this item until a complete picture emerges. © 2012 ACADEMY PUBLISHER Manufactured in Finland.