Research in Music Series: A Talk by Nathan Martin
May 1, 2019
Fine Arts Center
UMass Amherst Campus
Corpus Studies and 'Close Listening'
In a recent polemical intervention, Markus Neuwirth and Martin Rohrmeier present machine-assisted corpus-based research as an essential departure from music theory’s more customary routines. Drawing on Franco Moretti's work in particular, they juxtapose machine-enabled “distant listening” with the discipline’s native tradition of close analytical attention to particular musical works: the latter is relegated at best, in their vision, to the status of a pre-scientific heuristic—a means of generating hypotheses that then stand or fall according to the rigors of empirical testing against an appropriately selected corpus.
This picture, I want to argue, results from a false antithesis, one that confuses tools with methods. The increasing availability of machine-readable musical corpora and the gradual refinement of their attendant strategies and methods both offer enormous possibilities to music theorists. Not all of that potential, however, is methodologically novel, and what I want to emphasize—over and against the excitement generated by “big data” in the “digital humanities”—is what is not new about corpus-based research.
The argument comes in three parts. First, I sketch the practices that corpus-based music theory allegedly replaces by reading David Lewin's celebrated analysis of Schubert's "Auf dem Flusse" against Cleanth Brooks on Donne's "A Canonization" (in the opening chapter of The Well Wrought Urn). Next, I turn to Vasili Byros’ analysis of Beethoven’s Eroica Symphony as an example of a corpus-based interpretation. Finally, I present a reframed version of another of Lewin's analyses, this one of Schubert's "Morgengruss," with the aim of arguing for a broad continuity between aspects of Lewin's and Byros' approaches. The moral of the story is that employing digital methods does not necessarily make music theory scientific and vice versa.