skip to primary navigationskip to content

Shadows of meaning: Webern’s Piano Variations on record

Shadows of meaning: Webern’s Piano Variations on record

Funded by a AHRC Fellowship, the aims of this project were to examine, through a case study of the first movement of Webern's Piano Variations Op. 27, the processes through which performance traditions are established; to link such processes to the cultural and discursive contexts that condition them; to draw conclusions for the relationship between analysis and performance; and to explore new approaches for working with recordings that use computational technology to extend the scope of close listening. The project pages include the following:

Key findings

Webern's Piano Variations Op. 27 were first performed two years before the outbreak of Second World War, and it was only after the war—and in very different aesthetic and ideological circumstances--that a performance tradition developed. Through a combination of computer-assisted close reading and contextual analysis, the project traced the competing stylistic and ideological directions of early performances—one deriving from the pre-war tradition of the Schoenberg circle, the other from the European avant-garde associated with Darmstadt and die Reihe—and the processes of negotiation through which a more or less consensus interpretation developed by the 1970s.

I set this narrative into a variety of broader cultural and ideological contexts. One is the fraught relationship between the pre- and post-war avant-garde: largely because of Peter Stadlen, who premiered the work, the Piano Variations were constantly in the front line of controversy. I show that the links between pre-war and post-war traditions were much stronger than Stadlen was prepared to admit, for example through the influence of Arnold Schoenberg on Leonard Stein, and of René Leibowitz on Jacques-Louis Monod. Another is the ideological context of the post-war avant-garde, with its aim to make a new beginning and its links to existentialism and structuralism. While for a brief period Webern was seen as emblematic of this new beginning, however, the focus of the Darmstadt avant-garde was on his scores rather than his music as performed or heard.

Key modernist performers of the Piano Variations from Monod and Stein to Paul Jacobs and Yvonne Loriod drew on historical performance practices to a far greater extent than has been generally recognised, and recordings by the so called Darmstadt hardliners—none of whom were German—were also much more distinct and idiosyncratic than the literature suggests. There was a disconnect between Darmstadt-style analysis and performance practice: the evidence of Stein's 1965 DMA thesis is that he, at least, was thinking in pre-war terms even as he made perhaps the most uncompromisingly literalistic of all the recordings of Op. 27. Only in Jean-Rodolphe Kars's recording from 1969, and perhaps those of the Takahashi siblings during the 1970s, is there a suggestion of what an interpretation might have been like that proceeded from the basic tenets of post-war serial aesthetics.

Intertwined with this narrative is that of the development of North American music theory and what I call the ‘page-to-stage’ approach that long dominated academic thinking about performance. While music-theoretical discussion of serial music has focussed overwhelmingly on structure, performers of Op. 27 have always drawn to a greater or lesser degree on the stylistic assumptions and performance practices of tonal music: indeed Adorno claimed that it was necessary to rely on old-school performance practices if Webern's late music was to be given even 'the shadow of meaning'. Adorno’s implication was that, in itself, Webern’s serialism is meaningless. A more positive interpretaton is that performers have played a far greater and more creative role in generating musical meaning, in Webern’s Piano Variations and elsewhere, than most music theorists have been prepared to admit. They have, in short, invented tradition.

The research was first presented in July 2014 at the Performance Studies Network conference at Cambridge, with other presentations taking place during the year in London, Lisbon, Tallahassee, and Beijing. Two project outcomes are forthcoming. One ('Inventing Tradition: Webern's Piano Variations in Early Recordings') is a substantial article presenting the material described above: it appeared in Music Anaysis 36 (2017), pp 163–215. The other (‘Seeing Sound, Hearing the Body: Glenn Gould Plays Webern’s Piano Variations’) is a study of physical gesture in performance, structured around two recordings that Gould made for television, in 1964 and 1974; as well as tracing the interaction of the audible and visible dimensions of performance, I set these television performances into the larger context of stylistic development traced by Gould's five audio recordings of the Piano Variations. This was published in The Oxford Handbook of Music, Sound, and Image in the Fine Arts, ed. Yael Kaduri (New York: Oxford University Press, 2016), pp 124–37, and is forthcoming in German translation. Abstracts of both articles may be found here.

The project website hosts those audio and video recordings cited in the two outputs that are in the public domain, together with Sonic Visualiser session files that enable readers to integrate bar numbers, tempo and dynamic graphs, and other information into their playback environment. In addition the very substantial corpus of performance data for all three movements that was generated by the project has been placed on the web for use by researchers. Unfortunately copyright restrictions make it impossible to share other project materials, including the remaining recordings and analytical software, on an open basis.

Performance data

The project was based on a corpus believed to include all commercially released recordings of Op 27 from 1948 to 2012. Although only the first movement was analysed in depth for the project outputs, data were captured for all three movements and would support extensive follow-up research. Data capture involved extracting onset times and associated global dynamic values for all events in each recording, using tools developed by the CHARM Mazurka Project. These data may be accessed below in the form of Excel spreadsheets containing both timing and dynamic data for each movement:

Note events as listed in these spreadsheets and identified in the Sonic Visualiser session files may be identified by reference to the following annotated scores:

Further details of the recordings are provided in the discography. In a few cases recordings of movements listed in the discography will not be found in the performance data spreadsheets, owing to technical problems with the recording in question (for example omitted or incorrectly repeated segments).

The project additionally made use of Sonic Visualiser session files for each recording, together with custom designed web-based software. This ranked recordings in terms of timing and/or dynamic correlation between a specified segment of a given recording and the same segment in other recordings; it also retrieved and labelled that segment from each recording, delivered in the form of an mp3 file and associated Sonic Visualiser session file. These tools cannot be made freely available owing to copyright in the recordings. Interested researchers may contact Nicholas Cook or Craig Sapp.


The software used in this project was created by Craig Sapp (Center for Computer-Assisted Research in the Humanities, Stanford University). Data capture was carried out by Miriam Quick and Georgia Volioti. The collection of recordings was carried out in collaboration with Miriam Quick, and the discography is based on hers. This research could not have been carried out without the support of the Arts and Humanities Research Council.