In this dissertation, I argue that listeners’ ability to apprehend the expressive power and stylistic signature of late eighteenth- and early ninteenth-century music lies as much in its sonority (timbre and chord spacing) as in the traditional syntactic parameters (of harmony, rhythm, and form). This dissertation's chapters divide into theoretical (2 & 3), historical (4, 6, & 7), and analytical (5 & 8) ones. In the theoretical ones, I argue that music theory’s emphasis on textual analysis causes us to privilege what is most visible in the score over what is most audible in musical sound. To remedy this, I propose using affordance theory to ground musical analysis in the perceptual capacities of average (rather than idealized) listeners, whose ability to process music is often dependent on sonority. In the historical chapters, I examine a broad range of repertoire across decades and countries to challenge prevailing narratives, including: deterministic narratives of how changes in keyboard technology drove compositional practices, "Great Man"-centric narratives that treat figures like Mozart and Haydn as either innovators or unproblematic stand-ins for their eras' style, symphony-centric narratives that ignore operas and concertos when discussing the development of the orchestra, and Vienna-centric accounts that do not compare practices across Europe. Finally, in the analytical chapters, I build a series of tools to allow us to discuss sonority without resorting to vague metaphors, and I closely examine contrasting recordings of select pieces to show how sonority is exploited by performers to renegotiate supposedly fixed elements of the music (such as which notes belong to which line, or what the form is).