Philosophers have often wanted to state a principled way of demarcating empirical from non-empirical thought. This was a major concern of the Vienna Circle. In my view, this is an important intellectual project. Although it is not so common now to address the issue directly, it hovers in the background of many discussions. Non-empirical thought comes in different kinds. Perhaps some is a priori. Common candidates are mathematical, logical, modal and moral thought. Some non-empirical thought might be non-cognitive. Common candidates (...) are moral, aesthetic and religious thought. Or some non-empirical thought might be pseudo-empirical. Common candidates are astrological, theistic, Marxist or psychoanalytic thought; such thought masquerades as empirical, but in fact is not. These are all ways that thought can fail to be empirical. Here I shall focus for the most part on the case of morality, where there has been a lively debate about the empirical status of moral thought. I shall probe the prospects for a demarcation principle that can be used to decide whether or not moral thought is empirical. Some philosophers have recently presented what seems like a compelling case for thinking that moral thought is empirical, and so engaging with their arguments will raise the issue of what the empirical/nonempirical distinction amounts to, and also the issue of how to tell which forms of thought are empirical and which are not. (shrink)

I exist. That is something I know. Most philosophers think that Descartes was right that each of us knows that we exist. Furthermore most philosophers agree with Descartes that there is something special about how we know it. Agreement ends there. There is little agreement about exactly what is special about this knowledge. I shall present an account that is in some respects Cartesian in spirit, although I shall not pursue interpretive questions very far. On this account, I know that (...) I exist a priori; and I shall advance an explanation of how this a priori knowledge is possible and actual. I then consider the question of whether the belief that I exist is justified and, if so, how. I argue that the situation is different in important ways from the knowledge that I exist. (shrink)

I argue that attempts to demarcation ethics from science are not jeopardized by the fact that conjunctions of moral claims may have empirically verifiable logical consequences.

In this paper epistemic pluralism concerning knowledge is taken to be the claim that very different facts may constitute knowledge. The paper argues for pluralism by arguing that very different facts can constitute the knowledge‐making links between beliefs and facts. If pluralism is right, we need not anxiously seek a unified account of the links between beliefs and facts that partly constitute knowledge in different cases of knowledge. The paper argues that no good reasons have been put forward in favour (...) of believing in a unified maker of knowledge. It then appeals to the role of knowledge in order to argue that we have positive reason to embrace pluralism. (shrink)

Empiricist theories of knowledge are attractive for they offer the prospect of a unitary theory of knowledge based on relatively well understood physiological and cognitive processes. Mathematical knowledge, however, has been a traditional stumbling block for such theories. There are three primary features of mathematical knowledge which have led epistemologists to the conclusion that it cannot be accommodated within an empiricist framework: 1) mathematical propositions appear to be immune from empirical disconfirmation; 2) mathematical propositions appear to be known with certainty; (...) and 3) mathematical propositions are necessary. Epistemologists who believe that some nonmathematical propositions, such as logical or ethical propositions, cannot be known a posteriori also typically appeal to the three factors cited above in defending their position. The primary purpose of this paper is to examine whether any of these alleged features of mathematical propositions establishes that knowledge of such propositions cannot be a posteriori. (shrink)