Given Liszt's experience in composing sonata forms prior to 1853, we should, perhaps, not be too surprised that the Sonata manuscript shows the overall structure was clear in his mind at an early stage (assuming we do not find a bundle of hitherto unknown sketches in some dusty Weimar attic). We might, however, allow ourselves some astonishment at the degree of confidence shown in the structural organisation of the basic layer of the manuscript. Many of the symphonic poems went through several complete versions before publication, the revisions sometimes drastically altering the formal design. With the Sonata there seemed to have been relatively little doubt or hesitation over even the most complex element of its structure: the accommodation within a sonata form of a slow section and fugal ‘scherzo’. The fugue, in particular, which has been the source of so much disagreement among analysts, was evidently integral to the original conception. The slow section, an afterthought as far as the Grand Concert-solo was concerned, was again part of the initial inspiration. In its basic design the tonal and formal conservatism of some elements of the Sonata are offset by what might be called ‘structural puns’ that relate to the binary–ternary dichotomy of the traditional scheme. The most obvious of these is the function of the fugue, heard as a development, recapitulation or even ‘scherzo’ movement, depending on the interpretation of what has gone before. These separate views are not necessarily contradictions. The fugue can perform each of these functions in the structure, although our conception of the Sonata as a whole will differ accordingly.