How not to learn from history

Abstract This article attempts to answer two questions. First, what are the relevant ‘how not tos’ when it comes to learning from history? I argue that from existing accounts of how policy-makers (mis)learn the lessons of history, we can derive four ‘how not tos’: 1) do not settle or fixate on the first, or most ‘available’ or ‘representative’ analogy; 2) do not dismiss differences between your favoured analogy and the case in question; 3) do not neglect alternative analogies; and 4) do not shirk from ‘testing’ the observable implications of your preferred analogy. Second, do policy-makers show awareness of these ‘how not tos’ as they use historical analogies? An examination of how they are using the Cold War analogy to interpret the nature and trajectory of contemporary US–China relations suggest that they seem to have avoided the worst pitfalls of analogical reasoning in foreign affairs. The most prominent users of these historical analogies show awareness of the first three of the above proscriptions; all, however, shy away from the fourth proscription—testing the prognostications of their favoured analogy. Although this is far from perfect, the signs point to the Cold War analogy being used in ways that avoid the general pattern of superficial and poor use documented in existing analyses of analogical reasoning in foreign affairs.