Finding metaphor in grammar and usage: A methodological analysis of theory and research

Abstract

In much of his work, Gerard Steen has been concerned with the problem of how metaphors can be identified and how their conceptual structure can be analyzed in an intersubjectively reliable way. This endeavor appears to be a relatively modest goal in a field that has witnessed a large amount of grand (and sometimes speculative) theorizing about the relation between language and thought. However, Steen's research agenda is much more ambitious than it seems at first sight. His aim is the development of a multidimensional procedural approach to metaphor that, at each step of analysis, is guided by transparent methodological principles. It is only in this way, Steen claims, that an empirically sound theory of metaphor can be constructed. Steen proposes that researchers should proceed along five ''stages of research'': conceptualization, operationalization, data collection, data analysis, and data interpretation. Each of these stages ''feeds into the next one, but in practice also retrospectively affects the previous one'', and the stages are organized according to ''their own aims and norms'' (p. 382). Conceptualization concerns the construction of testable theories of metaphor. Operationalization translates conceptualization into a procedure of identification of metaphors (in contrast to other figurative meanings or literal language). Data collection deals with the well-known problem of sampling ''good'' and representative data. Three methods of data collection are discussed in some detail: introspection, observation (e.g., the use of corpus data), and manipulation (e.g., psycholinguistic experiments). Not too surprisingly, Steen is in general quite skeptical, or at least hesitant, about using introspection as a method of data collection. Contrary to the widespread use of introspective methods in data collection and interpretation in Chomskyan linguistics, and also partly in cognitive linguistics (see, e.g., Lakoff and Johnson, 1980; Langacker, 2008), Steen argues compellingly in favor of observational and experimental methods (manipulation). Data analysis and data interpretation are closely connected. The former is concerned with the development of qualitative and quantitative methods of metaphor identification and categorization; the latter relates analyzed data ''to an interpretive framework that makes them significant'' (p. 389). The author distinguishes a number of areas of research for metaphor (pp. 13–25) that are often not clearly kept apart in the literature, but that, in his view, are essential for a serious assessment of the plausibility of competing theories of metaphor. These research domains are defined in terms of contrasts between (i) grammar versus usage, (ii) symbolic structure versus behavior, and (iii) language versus thought (p. 14 et passim). The author considers four metaphoric models, whose empirical adequacy is painstakingly investigated within each of the research areas mentioned above. These metaphoric models are (1) the 'two-domain' approach (e.g., Lakoff and Johnson, 1980), (2) the 'many-space' approach (e.g., Fauconnier and Turner, 2002), (3) the 'class-inclusion' approach (e.g., Glucksberg, 2001), and (4) the 'career of metaphor' approach (e.g., Bowdle and Gentner, 2005). The four models share the view that metaphor is not merely a linguistic but also a conceptual phenomenon, a figure of thought (p. 49). In passing, Steen also mentions pragmatic accounts of metaphor, such as Searle (1979) and Sperber and Wilson (1995). The 'metaphor as thought' thesis is not shared by these pragmatic approaches, and that is probably the reason why Steen does not deal with pragmatic theories of metaphor in any detail in his book. However, some of his methodological reflections are certainly relevant to pragmatic models of metaphor as well, e.g., the development of intersubjectively reliable procedures for metaphor identification, and methods of data collection, analysis, and interpretation, to mention just a few. The focus of the book is on the two-domain approach to metaphor known as conceptual metaphor theory. It was first presented to a wider readership by George Lakoff and Mark Johnson in their influential work Metaphors We Live By