Expressing evaluation without grammatical stance: informational persuasion on the web. Issue 1 (April 2018)

• Record Type: Journal Article
• Authors: Biber, Douglas; Zhang, Meixiu
• Is Part Of: Corpora. Volume 13:Issue 1(2018)
• Page Start: 97
• Page End: 123
• Publication Date: 2018-04
• Subjects: keyword analysis -- lexico-grammar -- opinion -- persuasion -- stance evaluation -- web registers; Corpora (Linguistics) -- Periodicals; 410.188
• DOI: 10.3366/cor.2018.0137
• Languages: English
• ISSNs: 1749-5032

Abstract

Abstract : Although 'stance' and 'evaluation' are closely related theoretical constructs, they have typically been analysed in fundamentally different ways. Stance is normally described in terms of explicit lexico-grammatical features. As a result, stance has been studied through corpus-based methods that result in generalisable descriptions of different registers. In contrast, evaluative language has been regarded as implicit and context-dependent. As a result, analyses of evaluative language have focussed on the connotations of particular words and phrases, or on detailed descriptions of particular texts (rather than on generalisable descriptions of a register). Against this background, this study is motivated by an apparent contradiction that emerged from a recent study of register variation on the web (Biber and Egbert, 2017, forthcoming): end-users identified two categories of opinionated/persuasive documents on the web, one labelled 'Opinion' (OP) and the other labeled 'Informational Persuasion' (IP). However, a multi-dimensional analysis of register variation indicated that only the OP documents were marked for the use of grammatical stance features, while the IP documents were marked for the absence of those features. Our primary goals in this paper are two-fold. First, we undertake detailed linguistic analyses of lexico-grammatical stance features in OP and IP registers to confirm the general results of the earlier analysis. And, second, we use keyword analysis to explore the possibility that evaluation might be expressed lexically – rather than grammatically – in IP. In our conclusion, we discuss the broader theoretical implications of the study, including the possibility of using keyword analysis to explore a particular discourse function (i.e., evaluation); the possibility of analysing evaluative language in ways that can be generalised to an entire register; and the possibility of two complementary discourse systems (i.e., grammatical stance versus lexical evaluation) that are used as alternatives for the expression of personal attitudes and assessments.