Abstract

In an unpublished manuscript of 1914 titled ‘Logic in mathematics’, Gottlob Frege offered a rich account of the paradox of analysis. I argue that Frege there claims that the explicandum and explicans of a successful analysis express the same sense and that he furthermore appreciated that this requires that one cannot conclude that two sentences differ in sense simply because it is possible for a (minimally) competent speaker to accept one without accepting the other. I claim that this is shown by Frege’s suggestive remarks about a cloudy grasp of a sense. I then argue that this fact calls into question a key assumption behind Frege’s master argument for the sense/reference distinction.