The epistemic indispensability of understanding in historiography

In this paper I argue that understanding is an indispensable epistemic procedure when historians use texts as evidence. On my account understanding installs a norm that determines what kind of event or object a texts is evidence of. Historians can debate which norms should govern a body of texts, and if they reach consensus, they can use that body of texts as an empirical constraint over their historical claims. I argue that texts cannot perform this constraining function without understanding – it is thus indispensable. In order to argue for this position I first discuss two existing accounts of textual evidence in analytic philosophy of science by Kosso and Hurst. Both defend a coherentist position. I show that their coherentist position is flawed by applying it to the famous case of Lucien Febvre's argument that François Rabelais was not an atheist. I show that a coherence between texts leaves the debate concerning Rabelais' religious beliefs underdetermined, even though this should not be necessary. I argue that my account of understanding better captures Febvre's actual reasoning with texts. In the final section of the paper I show that the two most famous accounts of understanding in analytic philosophy by Hempel and Taylor ignore either the epistemic indispensability of understanding, or the actual success of evidentiary reasoning in the historical sciences that was enabled by understanding.In this paper I argue that understanding is an indispensable epistemic procedure when historians use texts as evidence. On my account understanding installs a norm that determines what kind of event or object a texts is evidence of. Historians can debate which norms should govern a body of texts, and if they reach consensus, they can use that body of texts as an empirical constraint over their historical claims. I argue that texts cannot perform this constraining function without understanding – it is thus indispensable. In order to argue for this position I first discuss two existing accounts of textual evidence in analytic philosophy of science by Kosso and Hurst. Both defend a coherentist position. I show that their coherentist position is flawed by applying it to the famous case of Lucien Febvre's argument that François Rabelais was not an atheist. I show that a coherence between texts leaves the debate concerning Rabelais' religious beliefs underdetermined, even though this should not be necessary. I argue that my account of understanding better captures Febvre's actual reasoning with texts. In the final section of the paper I show that the two most famous accounts of understanding in analytic philosophy by Hempel and Taylor ignore either the epistemic indispensability of understanding, or the actual success of evidentiary reasoning in the historical sciences that was enabled by understanding.B