3.3 Causation in history
A third important set of issues that received attention from analytic philosophers concerned the role of causal ascriptions in historical explanations. What is involved in saying that “The American Civil War was caused by economic conflict between the North and the South”? Does causal ascription require identifying an underlying causal regularity—for example, “periods of rapid inflation cause political instability”? Is causation established by discovering a set of necessary and sufficient conditions? Can we identify causal connections among historical events by tracing a series of causal mechanisms linking one to the next? This topic raises the related problem of determinism in history: are certain events inevitable in the circumstances? Was the fall of the Roman Empire inevitable, given the configuration of military and material circumstances prior to the crucial events?
Analytic philosophers of history most commonly approached these issues on the basis of a theory of causation drawn from positivist philosophy of science. This theory is ultimately grounded in Humean assumptions about causation: that causation is nothing but constant conjunction. So analytic philosophers were drawn to the covering-law model of explanation, because it appeared to provide a basis for asserting historical causation. As noted above, this approach to causal explanation is fatally flawed in the social sciences, because universal causal regularities among social phenomena are unavailable. So it is necessary either to arrive at other interpretations of causality or to abandon the language of causality. A second approach was to define causes in terms of a set of causally relevant conditions for the occurrence of the event—for example, necessary and/or sufficient conditions, or a set of conditions that enhance or reduce the likelihood of the event. This approach found support in “ordinary language” philosophy and in analysis of the use of causal language in such contexts as the courtroom (Hart and Honoré 1959). Counterfactual reasoning is an important element of discovery of a set of necessary and/or sufficient conditions; to say that C was necessary for the occurrence of E requires that we provide evidence that E would not have occurred if C were not present (Mackie 1965, 1974). And it is evident that there are causal circumstances in which no single factor is necessary for the occurrence of the effect; the outcome may be overdetermined by multiple independent factors.
The convergence of reasons and causes in historical processes is helpful in this context, because historical causes are frequently the effect of deliberate human action (Davidson 1963). So specifying the reason for the action is simultaneously identifying a part of the cause of the consequences of the action. It is often justifiable to identify a concrete action as the cause of a particular event (a circumstance that was sufficient in the existing circumstances to bring about the outcome), and it is feasible to provide a convincing interpretation of the reasons that led the actor to carry out the action.
What analytic philosophers of the 1960s did not come to, but what is crucial for current understanding of historical causality, is the feasibility of tracing causal mechanisms through a complex series of events (causal realism). Historical narratives often take the form of an account of a series of events, each of which was a causal condition or trigger for later events. Subsequent research in the philosophy of the social sciences has provided substantial support for historical explanations that depend on tracing a series of causal mechanisms (Hedström and Swedberg 1998).