In Defence of Empty Realism

Abstract

In my article (Demeter 2009) I argued that there is a meaningful distinction to be drawn between two kinds of realism. One of them consists in a realist (factualist) interpretation of a class of sentences and is committed to the claim that the given class has independent truthmakers. The other, which I label as an almost "empty" one, does not entail anything more than an ontological commitment to some entity. I explore the distinction thus drawn concerning folk psychology in the first respect, and mental entities in the second. J?nos T zs?r (2009) has challenged my distinction between the two kinds of mental realism, arguing mainly on the ground that it overlooks the problem of the mark of the mental. T zs?r's general point is that it is impossible to be realist about anything without some knowledge of the entity itself that we are realist about (377, 379). The particular point is that with respect to realism about mental entities, at least some of the knowledge stored in our folk psychology must be true about them, otherwise our realism would not be about mental entities (379). Accordingly, there are two main points of disagreement between us. The first is whether realism about an entity is possible without some knowl edge of it. T zs?r's answer is negative, and in this respect he is in good company, indeed. He seems to agree with Descartes's (1985: 78) dictum: "we must never ask (fit is, until we first understand what it is"?we need to know first what we are talking about and only then can we decide if it is real or not. But if I am right in claiming that realism about an entity is logically presupposed by any kind of empirical investigation into it, T zs?r's negative answer entails that we must have some a priori knowledge of what we are realist about. If realism is logically prior to empirical investigations and if realism entails knowledge about the entity, then this knowledge must be a priori. In my eyes, this is an unwelcome con sequence of T zs?r's position, as I do not think that a priori knowledge about how things are in the world is possible. My answer is affirmative: we can be realists about an entity without any kind of knowledge about it. I suggest that understanding what an entity is presupposes at least tentative ontological commitment?which, as such, can be withdrawn as a result of, e.g.,