EN
Historians and natural scientists are adept at inferring knowledge of the past from present traces and evidence. The repository of available methods has been rapidly expanding, and historians of the human past have learned that using techniques developed in other fields that study the natural past might prove beneficial to their endeavours in some cases. The network of inferences involved in historical discourse is vast and diverse. Influential philosophers of history and historical sciences like Jouni-Matti Kuukkanen and Aviezer Tucker have argued that it is necessary to expand the scope of the philosophy of history and to take a deeper look at underlying reasoning and inferential structure of history. This paper answers this call by analysing a group of inferences in human historiography that could be described in terms of hypothesis and hypothesis testing. Hypothesis testing has received some attention in the philosophy of historical sciences, but it is mostly underexplored in the philosophy of history in the context of human history. The paper examines four case studies and analyses their inferential structures by using concepts from the philosophy of historical sciences, such as trace-based and analogous reasoning, type/token distinction, etc. It will be shown that hypothesis testing helps generate knowledge about the past, but to fully appreciate it and to differentiate its types, philosophers of history must engage with the infrastructure of historical research.