This chapter proposes a new theory of the sonata-form exposition from the viewpoint of contrast rather than from the more common perspectives of unity, symmetry or balance. Expositions end differently to how they begin, and so do sonata forms. Why, then, does music analysis, with an eye to tonal return, celebrate sonata form chiefly as a paradigm of conflict resolution, as epitomized in the dynamics of recapitulation? An equally pertinent issue is surely the formal contrast within an exposition, and moreover, the kind of contrast which survives into the recapitulation, when the modulation to the second group is diverted into the tonic. Thus a level of contrast more fundamental than the tonal rhythm of modulation and resolution which has occupied sonata theory since Tovey and Rosen. At this level, I will argue that an exposition divides into two groups because it needs to say the same thing twice – it must present its material in two ways, the first time conventionally, the second time distinctively. I call this the ‘two-time rule’, a rule which attends to the dialectics of Enlightenment semiotics, whereby convention is equally the vehicle for communication and its straitjacket. Given that communication involves schemata, it entails what psychologists identify as a dilemma between gist and detail in schematic perception. Much of this is common sense. On the one hand, something will not be perceived or remembered if it cannot be fitted into a stereotypical pattern.