Towards Song

Abstract

This paper combines the separate but overlapping interests of its authors – one a composer and sound-artist, who often works with voice and word; the other a poet with an interest in how page-based poetry is actually read – in an enquiry into the performance of lyric poetry and its relations to song. The enquiry owes at least as much to the studio, to sound-editing software, to close listening and close reading as it does to existing literature on connected topics. The main precedent is Douglas Oliver’s 1989 publication, Poetry and Narrative in Performance, in which the author adopts the word performance to refer to any act of reading, whether aloud or silent. Like Oliver, the authors rely on the visual representations of the sound-shapes of specific sounded readings and also rely on access to sound files of some of the readings and their manipulations. In addition to the analytic value of having access to recordings and wave-forms and spectrograms, the authors experimented with using software (Audacity and Melodyne) to make subtle changes in pitch, pulling particular vowel sounds from their original positions to the nearest notes in a tempered scale. These experiments test, perhaps, a border-zone between lyric poetry and sung lyric. They also suggest a possible value for poets in using such software as support for their technologically unaided ears in testing out their own compositions.

Notes

1 An early version of this essay was given as a paper at The Shape of Song: A conference on lyric poetry, Faculty of English, the University of Cambridge, Saturday 7 July 2012.

2 Gerard Genette demonstrates in The Architext: An introduction (1992) that the supposed source for the triadic typology, Aristotle’s Poetics, does not, in fact, include the lyric, nor does it refer to the mode in any detail.

3 In music processing, ‘quantization’ usually refers to the process of correcting performance discrepancies in the time domain, by moving recorded notes to their nearest bar, beat or fraction of a beat. ‘Pitch quantization’ refers to the equivalent process in the spectral domain, where recorded notes are re-tuned, usually to their nearest semitone.

4 For a discussion on the role of recording in the audio-poem, see McCaffery (1997: 149–68).

5 This three-tier description of vocal utterance is explored in detail by Mladen Dolar (2006: 23–32).

6 Hannah Silva raised this question in ‘On the “minimal performance”: Interview with John Hall’ (2013).

7 ‘The second problem arising in connection with privative opposition is that of the unmarked term. It is called the zero degree of the opposition. The zero degree is therefore not a total absence (this is a common mistake), it is a significant absence’ and ‘in rhetoric, where, carried on to the connotative plane, the absence of rhetorical signifiers constitutes in its turn a stylistic signifier’ (Barthes 1967: 77–8).

8 Audacity is a free digital audio editor and recording computer software application (https://sourceforge.net/ projects/audacity/).