Optimizing the analysis of metaphor in discourse

Abstract

This article presents a software-based methodology for studying metaphor in discourse, mainly within the framework of Conceptual Metaphor Theory (CMT). Despite a welcome recent swing towards methodological reflexivity, a detailed explication of the pros and cons of different procedures is still in order as far as qualitative research (i.e. a context-sensitive manual coding of a text corpus) is concerned. Qualitatively oriented scholars have to make difficult decisions revolving around the general research design, the transfer of linguistic theory into method, good workflow management, and the aimed at scope of analysis. My first task is to pinpoint typical tasks and demonstrate how they are optimally dealt with by using qualitative annotation software like ATLAS.ti. Software not only streamlines metaphor tagging itself, it systematizes the interpretive work from grouping text items into systematic/conceptual metaphor sets, via data surveys and checks, to quantitative comparisons and a cohesion-based analysis. My second task is to illustrate how a good research design can provide a step-wise procedure, offer systematic validation checks, keep the code system slim and many analytic options open. When we aim at complex data searches and want to handle high metaphor diversity I recommend compositional coding, i.e. tagging source and target domains separately (instead of adopting a “one mapping-one code” strategy). Furthermore, by tagging metaphors for image-schematic and rich semantic source domains in parallel, i.e. two-tier coding, we get multiple options for grouping metaphors into systematic sets.