This article argues that the so-called Charge from Psychology does not refute the project of defining art. The charge entails that the project is misguided because it falsely presupposes that the concept of art is classically structured. The charge is challenged by distinguishing philosophers’ normative from psychologists’ descriptive aims. Unlike what many philosophers of art themselves believe, defining art is a normative project, since proposed definitions formulate conditions under which the concept of art should be applied, rather than is applied. (...) Therefore, it is not inherently problematic that definitions do not fit empirical data: a normative definition need not reflect how people categorize items as art. In the end, this article shows the significance of empirical data for definitions of art. (shrink)