An examination of the validity of metaphor analysis studies: Problems with metaphor elicitation techniques

The last decade has seen a large number of studies employing metaphor elicitation techniques, especially using ‘X is (like) Y’ format to investigate language teachers’ and learners’ understandings of teaching and/or learning. Although a few recent studies have reported the proportion of unsuccessful answers to this type of task, and identified a number of issues connected with task difficulty, there appears to be little published work that has seriously addressed the validity of the method used. The aims of this paper are therefore to explore the discourse and contexts where failure/difficulty with the metaphor elicitation task occurs, to try and understand what causes the problems and to suggest approaches to resolving them. In so doing, this paper reports on two small-scale metaphor analysis studies that were primarily designed to reduce the incidence of difficulty with ‘X is (like) Y’ metaphor prompts, presenting the perceived reasons for the difficulties and discussing possible solutions by introducing training in the form of four ready-made metaphor related teaching sessions. Essentially, I argue that training both about metaphor and in using it are important, and that thought needs to be given to both the nature and the length of training. The hope is that the present paper can be a first step and will serve to shed light on the ways that can be employed by metaphor researchers to identify and then resolve their methodological problems.