著者

一ノ瀬 正樹

出版者

日本哲学会

雑誌

哲学 (ISSN:03873358)

巻号頁・発行日

vol.2005, no.56, pp.42-62,3, 2005-04-01 (Released:2009-07-23)

参考文献数

26

---

The past is nowhere, but it is implied in our understanding of the world. We have to take into account these two fundamental characteristics of the past, i.e. absence and im-plicitness, in order to clarify historical knowledge, as history must be a part of the past highlighted by linguistic description. In this article I investigate the nature of historical knowledge by taking implicitness of the past to be causation between the past and our present state, and interpreting absence of the past as bringing about a probabilistic character of historical knowledge.After briefly examining the controversy about the reality of the past, I scrutinize the covering law model of historical causal explanation through considering its probabilistic form. In particular I mention the problem of how to apply Bayesian Conditionalisation to historical causal explanations. However, Bayesian theory is involved in serious difficulties like the problem of old evidence. This suggests that our choice of context must be ques-tioned, which calls the narrative theory of history into discussion. I argue that narrative theory will introduce backward causation from narration to past events, and that the theory will still imply a probabilistic contingent status of the past. Finally I take up the question of why we narrate our history, and I assert that a sense.of crisis causally motivates us to narrate it. As the sense of crisis itself is a sort of historical knowledge as well as the resultant narration, the probabilistic and contingent character extends throughout our historical knowledge.