M. RUFFINO EXTENSIONS AS REPRESENTATIVE OBJECTS IN FREGE’S LOGIC 1 ABSTRACT. Matthias Schirn has argued on a number of occasions against the interpret- ation of Frege’s “objects of a quite special kind” (i.e., the objects referred to by names like ‘the concept F ’) as extensions of concepts. According to Schirn, not only are these objects not extensions, but also the idea that ‘the concept F ’ refers to objects leads to some conclusions that are counter-intuitive and incompatible with Frege’s thought. In this paper, I challenge Schirn’s conclusion: I want to try and argue that the assumption that ‘the concept F ’ refers to the extension of F is entirely consistent with Frege’s broader views on logic and language. I shall examine each of Schirn’s main arguments and show that they do not support his claim. In his essay “Über Begriff und Gegenstand” (from 1892), Frege famously addresses an objection raised by Beno Kerry against his sharp ontological distinction between concept and object. Kerry claims, against Frege, that the ontological status of something as a concept or as object is not absolute, that is to say, something that seems like a concept may “behave” like an object in certain contexts, and vice-versa. Kerry uses the famous example of the expression ‘the concept horse’ to illustrate his point. The thing des- ignated by this expression seems to be a concept in contexts like ‘Silver falls under the concept horse’. But in contexts like ‘The concept horse is easy to grasp’, ‘the concept horse’ turns out to designate an object due to its position as grammatical subject. Since there is no reason to suppose that ‘the concept horse’ refers to different things in each one of these contexts, Kerry concludes that the status as concept or as object of the thing designated is relative and not absolute, contrary to Frege’s view. An essential part of Frege’s reply to Kerry’s criticism is his well known claim that, due to the presence of the definite article in ‘the concept horse’ the expression has to refer (if it has a reference at all) to an object, and not to a concept. And this is so in all contexts. There is here, as Frege claims, a systematically misleading effect of ordinary language: it makes us refer to an object, while in fact we intend to refer to a concept. The expression ‘the concept horse’ refers, according to Frege, to an object that “represents” the intended concept in logical investigations (KS 170). 1 Hence Frege’s apparently paradoxical dictum that “the concept horse is not a concept” (KS 170). Erkenntnis 52: 239–252, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.