Philosophers often invite their readers to categorize imaginary objects. These objects are not only hypothetical: many of them cannot exist because of physical or technological reasons. They are unprecedented or unheard-of objects. By categorizing imaginary objects, philosophers expect to gain knowledge about our concepts. In this paper, I challenge this general assumption: not every conceivable object can be described in terms of our existing categories. Although prominent philosophers held similar views in the past, they made no effort to provide a (...) satisfactory explanation for such impossibility. The argument that I will develop in the following pages rests mainly on three philosophical assumptions: a) that concepts whose content philosophers attempt to describe denote “monothetic classes”; b) that concept formation depends largely on “ecological salience” (a notion developed in the field of ethnobiology); c) that in any monothetic class, we can draw a proper distinction between definitional and correlated properties. (shrink)