How do we know what we believe? Gilbert Ryle is often credited with the view that we know what we believe in much the same way that we know what others believe: namely, by inference from observation of behavior. This Rylean view cannot explain all of our self-knowledge, however, since we can know what we believe even when our beliefs make no relevant causal impact on our behavior.1 In opposition to this Rylean view, I will assume that there is an epistemic asymmetry to be drawn between first-personal and third-personal ways of knowing what we believe. Each of us has some way of knowing what we ourselves believe that is peculiar in the sense that it is different from any of our ways of knowing what others believe. What is more, I will argue, this first-personal way of knowing what we believe is privileged in the sense that it is immune from the rational uncertainty and error that affects our ways of knowing what others believe.2 What is this first-personal way of knowing what we believe? A preliminary answer is that we know what we believe by introspection. However, the term ‘introspection’ is nothing more than a placeholder for an account of how we know what we believe. We certainly cannot assume that our knowledge of what we believe has its source in anything like inner perception. So the task remains for a theory of introspection to fill in this placeholder by giving an informative account of our first-personal way of knowing what we believe. The aim of this paper is to argue that what I call the simple theory of introspection can be extended to account for our introspective knowledge of what

中文翻译：

---

信念与自知：摩尔悖论的教训

我们怎么知道我们相信什么？吉尔伯特·赖尔 (Gilbert Ryle) 经常被认为是这样一种观点，即我们知道我们相信什么，就像我们知道其他人相信什么一样：即通过对行为的观察进行推断。然而，这种瑞尔观点并不能解释我们所有的自我知识，因为即使我们的信念对我们的行为没有相关的因果影响，我们也可以知道我们相信什么。 1 与这种瑞莱观点相反，我将假设存在一个认识论第一人称和第三人称的方式来了解我们所相信的内容之间存在不对称性。我们每个人都有某种方式了解我们自己相信什么，这种方式很奇特，因为它不同于我们了解其他人相信什么的任何方式。更重要的是，我会争辩说，这种以第一人称方式了解我们所相信的事物是特权的，因为它不受理性不确定性和错误的影响，这种不确定性和错误会影响我们了解他人所相信的方式。2 这种了解我们所相信事物的第一人称方式是什么？一个初步的答案是我们通过内省知道我们相信什么。然而，术语“内省”只不过是一个占位符，用于说明我们如何知道我们所相信的。我们当然不能假设我们对我们所相信的东西的知识来源于内在感知。因此，内省理论的任务仍然是通过对我们了解我们所相信的第一人称方式的信息性说明来填补这个占位符。