That gap-fill melodies and changing-note schemata are perceived differently was confirmed by an experimental study Meyer undertook together with Burton S. Rosner (1986). More ambitiously, Meyer's latest book, Style and Music (1989) projected the gap-fill schema distinction onto an historical narrative. In brief, Meyer conceives of music history, the shift from the Classical style to Romanticism and then to the modern age, as a retreat from Culture towards Nature. If the changing-note schema epitomised the rule- governed, learned and conventional aspects of classicism, then the Romantics' "repudiation of convention" (164) was rejected in the increased importance "of secondary parameters in shaping musical process and structure" (208). Secondary, statistical parameters, such as texture, density, dynamics and timbre, "seem able to shape experience with minimal dependence on learned rules and conventions" (218). "For many listeners, the power of sheer sound - as music slowly swelled in waves of sonic intensity, culminating in a statistical climax or a plateau of apotheosis - in a very real sense shaped experience 'naturally'" (218). If changing-note melodies are characteristic of eighteenth- century music, than the values of 19th-century music were typified by "axial" schemata. Axial patterns are subject to the natural psychological satisfaction of return. They are, furthermore, "consonant with values of Romanticism such as openness, reliance on natural rather than learned means, and appeal to less sophisticated members of the elite egalitarian audience" (242).
Perhaps the most suggestive element of Meyer's late theory is the fashioning of a nature/convention continuum in terms of music theory. As we have seen, the appraisal of signs according to their level of utilitarian or biological motivation is a practice associated with the visual arts. The originality of Gombrich's version of mimesis is that it concerns not the relation of the sign to its referent, but rather, at a much higher level, the relation between the signification process and the invariant principles of perception and of learning. Meyer's theory would therefore seem to point to the possibility of a "painterly" music theory, an analytical discourse finally able to represent musical structure in relation to the world. Nevertheless, even by the standards of his most rigorous work, the earlier Explaining Music, Meyer's Style and Music is methodologically incomplete. For example, the "convention" side of his nature/convention dichotomy is analytically extremely limited. On the natural side, Meyer has had no trouble in graphing entire movements in terms of gap-fill (Gestalt) principles. By contrast, Meyer's classification of melodic schemata has only been applied to short phrases, and we still await our first movement-length schematic analysis. What might a "musical iconography" of nineteenth-century music look like? I will close with some suggestions.