American television and off-screen registers: a corpus-based comparison

Abstract

In this paper, we present an analysis of a corpus of American television programmes, comprising 930 texts from 191 different TV programmes (excluding commercial breaks and infomercials), totalling 5,320,159 tokens. This analysis compares different programme types (live politics, movies, news desk, reality shows, etc.) against one another and against non-televisual registers (face-to-face conversations, telephone conversations, prepared speeches, press reviews, etc.), using the multidimensional (MD) approach to register variation ( Biber, 1988 ). The goal of this analysis is to determine how homogeneous/heterogeneous the language of television is as well as to how it compares to the registers of English ( Biber, 1988 ). We sought to detect both the similarities and differences among the TV registers and with respect to the five major dimensions of register variation previously identified by Biber (1988) . Linguistic studies of television language to date (e.g., Al-Surmi, 2012 ; Bednarek, 2010; Quaglio, 2009 ; and Rey, 2001 ) have generally focussed on few or individual TV registers. This study provides a much more comprehensive view of American television discourse by relying on a large multi-register corpus optimised for representativeness through the application of Biber's (1993) method for sampling adequacy. The results show distinct differences among the registers' dimensions, suggesting that, on the whole, present-day American TV language is varied and patterned. In addition, the findings indicate that several TV registers have close counterparts in off-screen communication, such as conversations, speeches and interviews. All things considered, this study shows in detail the contact and separation points across different TV registers, as well as across TV and off-screen registers.