Throughout recent theoretical and pedagogical literature on form, a commonly accepted part of the definition of Classical rondo, sonata-rondo, and expanded-binary rondo forms is the recurrence of the opening A section (the "refrain") in the tonic key between the contrasting episodes (B, C, etc.). However, as is well known, many pieces, most notably the finales of Haydn, play with these normative formal traits. Indeed, some apparent sonata-rondo finales seem to hover between two formal types, or change midway through the pieces to a "pure" sonata or "pure" rondo form because of formal peculiarities or omissions. The truncation of the second A section, or in some cases, its omission from a rondo-style movement, can significantly problematize a simple identification of a piece's form, and can also affect its harmonic structure, particularly when other formal complications accompany this truncation.

This paper, focusing on finale movements, will consider the second A section in terms of a continuum where one extreme is a full, unabbreviated second A section (non-problematic) and the other is no second A section. Between these two extremes lies a rich set of finales that defy the generic norms to varying degrees because of their truncated, dissolving, and/or re-composed second A sections. A representative sample of these finales ranging from Haydn and Mozart, through Schubert and Mendelssohn to Brahms, will be examined in order to show the techniques composers used at this point in the movement, and to explore the voice-leading structure of these movements within their historical-stylistic context.