Simulation versus the thing itself: commentary on Markman and Tetlock.

Abstract

My comments take for granted the integrity and seriousness of the authors whose work is being reviewed, and nothing in the critical tone of my remarks has anything to do with their professionalism or good faith. Far from it; no reader can mistake the painstaking effort of Markman and Tetlock's (2000) work, and it is a four-square example of the social psychological laboratory research tradition that stretches back at least to the 1960s and E. E. Jones and Gerard's influential, experiment-endorsing Foundations of social psychology (1967). I say that because, inevitably, a commentary such as this one will contain disagreements with what was done. But among ourselves, in the collegial pages of the BJSP such complaints are meant with respect. Two sorts of comments are in order. One is to remind us all of the conflicts of the 1970s — the ‘crisis in social psychology’ — and the many arguments it generated about doing any social psychological research. That is a story well-known to most readers of the journal, and so needs only a light dustingdown. More interesting, I think, will be to swoop down from the general to the particular, and ask ourselves how one could do this sort of this research - research into people's use of counterfactual explanations - in a way that uses the debates of the 1970s to move away from the laboratory and its model of causes and effects.