Historical irrealism: Paul A. Roth and the epistemic value of narrative explanation

In The Philosophical Structure of Historical Explanation, Paul Roth undertakes the task of answering a philosophical question of substantive importance: that of ‘why narrative form typifies and is essential to historical explanations’ (xiii). The author warns about the undeniable fact of the widespread use of narratives to explain in all kinds of history: social, political and cultural history, history of science and history of philosophy. Narratives tell us that something happened: ‘A narrative explanation presumably presents an account of the linkages among events as a process leading to the outcome one seeks to explain’ (22). The philosophical question raised is whether this practice is to be tolerated or condemned (22). Roth takes on the challenge of revisiting the question with respect to the analytical philosophy of history of the 1960s by, among others, Arthur Danto, Louis Mink, W. B. Gallie and Morton White – thinkers labelled ‘narrativists’ by Dray (for the first time in Ely et al 1969 and later in Dray 1971). Although interest in narrative structure in general and in historical narrative in particular has continued to grow since that time – and has attracted all kinds of disciplines to its conceptual elucidation, the question about the explanatory capacity of narrative has attracted little or no attention. The issue of the status of narrative explanation that Roth raises in the book’s first pages was approached, on the one hand, in a negative way by neo-positivist philosopher Carl Hempel, who excluded narrative as a kind of scientific explanation. Let us remember that Hempel reconstructs explanation as an argument – a logical structure of premises and conclusion. To explain scientifically is to show that an event (or, more precisely,