International Journal of Corpus Linguistics - Volume 11, Issue 3, 2006

Lexical cohesion not only contributes to the texture of a text, it can help to indicate the rhetorical development of the discourse. This article looks at this argument-structuring function of lexical cohesion first by considering single texts using the techniques of classical Discourse Analysis and then by using the methodology of corpus linguistics to examine several million words of text. First, the nature of cohesive links within single articles is examined. Next, the link between headlines and the articles that follow them is studied. Finally, various concessive mechanisms which structure arguments are examined in detail. It is argued that an awareness of the mechanisms outlined in this article will help students to understand better the kind of argumentation presented in texts. All the texts studied are from English newspapers.

This paper discusses some approaches to the categorisation of cohesive devices with reference to spoken academic discourse, multi-word units, and strings of frequently co-occurring words (lexical bundles). It goes on to investigate the cohesive role of lexical bundles in a corpus of 160 university lectures (120 from the BASE corpus and 40 from MICASE). Like the bundles from the T2K SWAL teaching subcorpus, investigated by Biber, Conrad and Cortes (2004), the bundles in the lecture corpus included both ‘oral’ and ‘literate’ elements. The majority of frequently occurring bundles were found to be used to signal discourse relations, although their cohesive function was not necessarily obvious when listed out of context.

This paper describes lexical cohesion across participants in a discourse, and across discourse events, and the additional contribution made by speakers’ choices of prominence in the cohesive chains. The choice of prominence is made to communicate what is perceived to be situationally informative in the local context of interaction (Brazil 1985 and 1997). This study describes a connection between a speaker’s choice of a lexical or non-lexical word in a cohesive chain and the speaker’s choice of prominence and, in doing so, offers a partial explanation to the question Hoey (1991:17) asks of Winter (1974 and 1979): ‘under what circumstances do we use one (cohesive device) rather than the other?’. It also describes how the selection of prominence across speakers in a discourse can signal divergence and convergence between them. Lastly, the paper discusses the pedagogical implications of the study with regard to English language learning and teaching.

The study reported in this paper applies Sinclair’s (2004) descriptive model of lexical items, which consists of five categories of co-selection: two obligatory categories the core and semantic prosody, and three optional categories collocation, colligation and semantic preference. The study examines a selection of spoken discourse events collected in Hong Kong during and in the immediate aftermath of the SARS crisis in 2003. These discourse events form part of the Hong Kong Corpus of Spoken English (HKCSE). The findings show that once the overlapping patterns of co-selection of the most frequently occurring lexical words in the SARS corpus have been determined, it is possible to describe the cumulative effects of the habitual co-selection in the lexical items that contribute to textual meanings and coherence within and across the texts. It is argued that patterns of co-selection provide a fuller picture of textual and intertextual coherence than concentrating solely on lexical cohesion.

Signalling nouns are nouns which have cohesive properties across and within clauses. A signalling noun is potentially any abstract noun the full meaning of which can only be made specific by reference to its context. Examples of nouns which can function as signalling nouns are attitude, assistance, difficulty, endurance, process, reason, result etc. Signalling nouns in discourse are closely associated with nominalisation and are problematic for learners. Based on a corpus of argumentative essays written by Cantonese L1 learners of English, this paper presents a taxonomy of error types and frequency data of the different error types in the use of signalling nouns. The paper then compares the average number of signalling nouns used per essay with grades awarded to the essays, on the one hand, and the numbers of signalling noun errors according to grades, on the other. In both cases there is a significant correlation. The findings confirm the intuitive idea that the use of signalling nouns adds to the overall coherence of a text.

Cohesion is generally described with regard to two broad categories: ‘grammatical cohesion’ and ‘lexical cohesion’. These categories reflect a view on language that treats grammar and lexis along separate lines. Language teaching textbooks on cohesion often follow this division. In contrast, a corpus theoretical approach to the description of English prioritises lexis and does not assume that lexical and grammatical phenomena can be clearly distinguished. Consequently, cohesion can be seen in a new light: cohesion is created by interlocking lexico-grammatical patterns and overlapping lexical items. A corpus theoretical approach to cohesion has important implications for English language teaching. The article looks at difficulties of teaching cohesion, shows links between communicative approaches to ELT and corpus linguistics, and suggests practical applications of corpus theoretical concepts.