Introduction
In the Hans Christian Andersen fairy tale The Little Mermaid, the young mermaid at the centre of the story allows the sea-witch to cut off her tongue in order that she may procure a pair of human legs and the opportunity to win a human soul by obtaining the love of a handsome young prince. In the version of the story made into the 1989 Disney film of the same name, the mermaid, now called Ariel, makes a rather less gruesome trade: signing over her voice to the sea-witch Ursula before expelling it via a soaring vocalise. Despite their significant aesthetic and narrative differences, each version hinges on the value of the voice as a medium of exchange: a voice for a pair of legs. Beyond this (meta)physical exchange, Andersen’s story plays on the multiple significance of the voice as a form of identity within Euro-western thought. In particular, caught between voice-as-sound and voice-as-site (or, as Brian Kane puts it, between echos and topos), the voice in The Little Mermaid has an unstable relationship to individual agency and subjectivity.Footnote 1 Is the voiceless human the same creature as the envoiced mermaid, or, as Andersen’s central character suggests, is it that to lose a voice is to undergo ontological change: ‘If you take away my voice, what is left for me?’
Writing in the 1830s, Andersen could not have predicted the advent of recording, but the development of recording and playback technology (and the resulting separation between vocal site and sound) shapes one of the key narrative differences between the original and its 1989 adaptation. In the Disney version, Ursula captures Ariel’s voice in a magical shell, before adopting it as her own in order to enchant the prince and claim dominance over both Ariel and the realms of the sea. In this form of impersonation, the physical sounds of Ariel’s voice (sung in both guises by Jodi Benson) are layered on top of a different body and mind and fundamentally alter Ariel’s capacity to act in the human world. In the years since, the questions around whether the voice can be traded or ceded (and what the implications might be for such a transaction) have shifted, but have not gone away. Fast-forward to 2023, and the artists Holly Herndon and Grimes have invited fans to use software (Holly+ and Elf.Tech, respectively) to create new songs by deepfaking the artists’ voices. The idea of the synthesized singer is not itself new: Nicholas Cook, Nick Prior, and others have written about the Japanese holographic artist (or, if you prefer, the Vocaloid voicebank) Hatsune Miku.Footnote 2 What is new is the idea of synthesizing replicant voices, of vocal cloning.Footnote 3 In an age of rapidly expanding AI-generated vocal material, this raises significant questions about the intersection of voice, creativity, and economics. Strikingly, Herndon frames her project as an ‘experiment in communal voice ownership’, while Grimes has suggested that she will split royalties 50:50 with anyone who creates a commercially successful song using her deepfake vocals.Footnote 4 The difference illuminated in the transformation from Ariel and Ursula to Herndon and Grimes revolves around how the voice is being formulated: from a deeply embodied manifestation of an individual to a distributed means of creativity with commercial implications.
Whether centred on mermaids and sea-witches or on AI-powered singers, the potential divisibility of the sonic attributes of the voice from an underlying material body and subjectivity is reflected in the way a peculiar division of labour structures modern Euro-western thinking about the voice as a sonic and material phenomenon. Under ordinary vocal circumstances, we might say that some metaphysical entity one might think of as ‘oneself’ triggers vocal folds to vibrate, lungs to expand, and muscles to tighten until self and body combine to produce sounds that become ‘the voice’. Even where the physical apparatus of the voice is disrupted (for example for the scientist Stephen Hawking, who began using a speech-generating device in the 1980s), the separation between the initial command and the musculature of vocal production is maintained.Footnote 5 The voice originates in the body, but it is not fully of the body. This is why the voice has been thought to grant a special kind of access to subjectivity, but its liminal existence means it is simultaneously a product and has the potential to become a commodity.
Furthermore, this is not limited to the sonic material of the voice: even when framed as a metaphor for the self-expression of either an individual or a collective, the voice is frequently thought of as something that must be produced, usually through a process of self-exploration and experimentation (e.g. ‘finding your voice’). Yet once fashioned, this metaphorical voice can become a thing to be possessed and used as well as being something that is vulnerable to being suppressed or taken away by processes that might leave the physical body untouched. From this perspective, one striking aspect of the turn to voice-as-material is the way that it enfolds the subjective and metaphysical aspects of the voice into its potential as a commodity. Herndon and Grimes are not participating in a sonic version of a stock photo album; they are marketing the distinctive sounds of their voices, which are significant because they are linked to their established commercial, artistic, and even physical identities.
These idea(l)s of the voice as simultaneously metaphor for individual autonomy and agency and as material phenomenon that connects inner and outer worlds arose in specific historical and social conditions. However, they have become deeply sedimented in Euro-western thought. The voice-as-individual-expression has become, in Lakoff and Johnson’s sense, a metaphor that we live by even as the implicit tensions in constructing the voice as an object to be possessed and as an irreducible sign of identity have come under pressure.Footnote 6 This pressure has come from multiple directions, among them the critique of Cartesianism evident in an increased interest in the embodied materiality of voices, a disruption of the equation between the voice and individual subjectivity in postmodernity, and increasing questions around agency and potentiality evident in a larger societal disconnect between rhetorical power and effectual change-making.Footnote 7 In this context, it is abundantly clear that the neoliberal ideal of the individual finding/fashioning, possessing, and using something considered to be their own (sole) voice is not the only way to conceive of how the voice might act in the world. In this short provocation, I consider whether reconceiving the economy of the voice might offer a new way of thinking about the entangled relationships between individual and collective, voices and their agencies.
Possessing the Voice
In a well-known US legal case from 1988, the singer Bette Midler sought a judgment against the Ford Motor Company for their having used another singer to imitate her voice for an advertisement. The decision ‘centr[ed] on the protectibility of the voice of a celebrated chanteuse from commercial exploitation without her consent’ and went on to claim that the voice ‘is as distinctive and personal as a face […] one of the most palpable ways identity is manifested’.Footnote 8 This is an unusual case in that it was based not on grounds of copyright but rather on the property right of publicity. Did Midler have a right to exclusive use of her voice as a commercial endeavour? The court eventually ruled that she did and established the principle that when the voice in question is that of an artist renowned for their voice, ‘to impersonate her voice is to pirate her identity’.Footnote 9
Compare this with the near-contemporary case of ‘Ride on Time’ (1989), in which the Italian group Black Box first sampled (sans credit) the disco singer Loleatta Holloway, then ‘fronted’ her sampled voice with a lip-synched performance by the model Katrin Quinol in the accompanying video and promotional material. Barbara Bradby suggests that in the popular discourse surrounding this event, Holloway was understood as doubly disenfranchised by the occlusion of both her voice and her person.Footnote 10 Although permission for the vocal sample was eventually obtained from the copyright holders, Holloway was never credited; instead, the track was re-released with an unknown singer re-recording the vocals.Footnote 11 These two examples demonstrate some of the challenges around extricating the voice from an economy based around proprietorial use. The legal and societal critiques in the Midler and Holloway cases are predicated on the assumption of a constituent relationship between voice, subjectivity, and body that treats a voice (and its particular expressions) as property.
These examples also reveal that the relationships between voice(s) and property can never be stable. In the context of this round table, the contingent control of voice and, conversely, of voicelessness resonates strongly through the ways in which individuals who have crossed the boundaries of nation states are routinely presented in the media and in scholarship as having lost their voice along with the protections of citizenship.Footnote 12 This, too, presumes that voice is both property and a property. As the contributions from Kyratsou and from Murphy and Chatzipanagiotidou demonstrate, listening closely to refugees reveals multiple voices and voicings existing within what Murphy and Chatzipanagiotidou term a ‘dynamic assembly’ or constellation of expression. In their Benjaminian guise, constellations allow us to trace (and trace differently) the relations between objects while preserving both autonomy and connection, making this a natural fit for thinking about the complexities of the agential economy of the voice.
In following this path of thinking about the articulation points of vocal subjectivity, I take as a point of departure the familiar phenomenon of sampling, or the re-use of pre-existing material in a new creative work. As Marianne Franklin suggests, sampling is a subset of the venerable practice of musical borrowing, but it is one that takes on a particular character in the age of the remix.Footnote 13 Much of the scholarship on sampling has focused on its dual identity as a compositional technique and as a central part of Black creative practices of intertextual communication (broadly dubbed ‘signifyin(g)’ by Henry Louis Gates Jr) that has spread widely across different genres.Footnote 14 Perhaps inevitably, musical sampling has been rapidly drawn into questions of agency and creativity due to its direct connections to questions of copyright. In the past thirty years, the troubled relationship between sampling as a practice and prevalent intellectual and sociopolitical structures of ownership and authorship that seek to restrict and/or monetize creative work has resulted in a sharp increase in legal action over purportedly sampled popular song. Despite this, Franklin claims that to reduce the process of sampling ‘to struggles over “fair use” or artistic license’ is to miss the key point: sampling is indicative of an artistic relationship that emphasizes an ethos ‘not based on music taking premised on aesthetic or cultural entitlement but, rather, on music making that understands itself as part of mutually reinforcing networks of visible and audible signification’.Footnote 15 While this may not apply equally to every use of sampled material, this conviction offers an important route to evaluating cultural products that are in conversation with the past.
Within these networks of signification, the status of the borrowed voice is particularly salient because of the way it plays on the separation between the body and the sounded voice. Working outward from the embodied tradition that links voices and bodies, Lauren Redhead argues for rethinking the ‘material affects of the voice in the world’.Footnote 16 Voice in this telling is vibrant and vibrating, acting as it moves outward into the world and leaving behind both the body and the mind that brought the voice forth. Hence the capacity for vocal action does not cease, but rather changes at the moment of inception. To illustrate this, Redhead turns to the myth of Echo, who loses her voice — at least in the sense that she loses her ability to form new sentences and thereby express her subjectivity. Yet the result of Echo’s metaphysical diminishment is not a wholesale loss of agency, for by retaining a physical voice that responds to the words of others, ‘she is able to bend and exceed the intended meaning of the original utterances when she reproduces them’.Footnote 17 When linked to the ways in which sampling and other forms of borrowing re-situate material in new contexts — bending and exceeding its original meaning — the voice’s echoes become far more than the mere reflections of its sound.
To this I add the relational ontology developed by Adriana Cavarero, which begins by acknowledging that while the voice is ‘incarnate singularity’, there is never one voice; rather, every voice (including the multiple voices which may be emitted from a single body) exists in a complex web of entanglement with other voices.Footnote 18 In her 2005 book For More than One Voice, Cavarero frames this as a turn away from logos and towards phône, but unlike Kane’s ‘model voice’, this voice/phône is explicitly an intersubjective phenomenon.Footnote 19 In other words, the subject is always speaking from and to an audience, thus upending the traditional liberal autonomous subject. Moreover, rather than seeing these relations as a network (into and out of which any voice might be plugged or unplugged), Cavarero understands this intersubjective entanglement as a fundamental condition of the voice’s existence. We neither have a voice nor are we given a voice; voice and the condition of vocality come into being in and through interaction rather than transaction.
From these starting points of intersubjective relationship, Echo’s material and affective excess, and sampling as amplifying signification, vocal borrowing in music emerges as a kind of gift which ricochets and reverberates through a relational web formed through musical and other connections. In what follows, I trace these reverberations through examples in which the borrowed voice expands and resounds far beyond its original context. Such borrowings inevitably result in the re-situating of material, yet it is clear that this is not a displacement but rather a multiplication of potential. While this does not sidestep questions of power or obviate the potential dangers of diminishment, it does point towards a hopeful prospect for thinking through what having and using a voice in and through relationship might reveal.
Performing Vocality
I open this discussion of borrowed voices in an unexpected place: so-called western art music. Just as in quotidian contexts the production of the voice may be divided between the embodied and the metaphysical, a similar division of labour surfaces in our colloquial understanding of music composition and performance. In particular, composers may be exhorted to find their own musical and metaphorical ‘voice’ as a sign of artistic maturity and marketability. The mark of success is found in performance, wherein the composer’s metaphysical voice is given its material sounding through the efforts of other bodies and selves. At times this Cartesian separation intensifies into a model of composer as ventriloquist, which in its most caricatured form requires the performer to give over their body so that they may speak with the composer’s voice. Put another way, the composer must extricate themselves from a ventriloquizing relationship with tradition, only to promptly initiate their own performing ventriloquism with the performer of their work.Footnote 20
As traditionally conceived, this metaphor of the composer’s voice strips the performer of meaningful agency. It is a view strongly associated with both Romantic and, especially, modernist thinking, but it has come under sustained challenge. One provocative reframing is that of Cathy Berberian, who proclaimed the importance of what she called a New Vocality in the 1960s. She wrote in 1966:
What is the New Vocality that is so threatening to the old guard? It is the voice which has an endless range of vocal styles at its disposal, embracing the history of music as well as aspects of sound itself; marginal perhaps compared to the music, but fundamental to human beings. Unlike the instrument, which can be locked up and put away after use, the voice is something more than an instrument, precisely because it is inseparable from its interpreter.Footnote 21
Berberian acknowledges the hermeneutic density of the voice as a collection of sounds which is both ‘fundamental’ to and ‘inseparable’ from its human originator, but she rejects the idea that music is the sole expression of the composer. Rather, the vocalist is not only capable of acting upon the composer’s work, but responsible for doing so by altering and amplifying the work in original ways through their performative expression.Footnote 22 In Berberian’s career, this is clearest in her early collaborations with John Cage and Luciano Berio; take, for example, her cooperation with Berio on Visage (1961), described by David Osmond-Smith as one in which Berio ‘cajoled, suggested, explored new directions, and Berberian “invented vocal situations” in response’.Footnote 23 Although Berio then extracted the material he wanted (and with it, the lion’s share of both recognition and economic benefits), this suggests a more collaborative and dialogic idea of the voice that renders the intersection between the composer for the voice and the performer’s active envoicing a rich area for thinking through the multiple agencies at work.
Sampling the Past
In contrast, sampling seems to reverse the performative vocality-as-agency suggested by Berberian by allowing compositional creators to re-use particular performances to inflect their works. Take, for example, the track ‘Blood on the Leaves’ from Kanye West’s 2013 album Yeezus. The track opens with a pitch-altered fragment of one of the best-known anthems of the US civil rights era: Nina Simone’s 1965 recording of ‘Strange Fruit’. As we hear Simone sing the phrase ‘blood on the leaves’, West’s first verse drops in a nervous, tentative register. Lyrically, West’s verses are a furious portrayal of an isolated man and a relationship turned sour by fame and fortune, and their seduction of the unnamed woman he addresses. As the song proceeds through a masterfully produced collision of autotuned vocals, towering brass fanfares, and drum riffs, Simone’s voice keeps appearing and disappearing, but it is she who has the last word, a repeated fragment of ‘breeze’ altered to sound almost exactly like ‘please’. This raises the spectre that the track is not — or not only — the result of one artist borrowing the voice of another, but rather a mutual reverberation that sketches possible relationships between and beyond the two voices.
The track was divisive when it was released, with some critics hard-pressed to explicate the connections between the Simone sample and the narrative content of ‘Blood on the Leaves’ or reacting negatively to the choice to sample Simone’s voice.Footnote 24 Others sought to claim that the iconic status and subject matter of the sample were irrelevant: so long as it fitted the beat (and it does), who is to say what is inappropriate? Craig Werner even raises the possibility that this is West’s version of what Toni Morrison calls the work of ‘re-memory’: by involving Simone (and by implication, Billie Holiday and others), West is ‘keeping the voices of ancestors and the awareness of the history alive […] It’s not about the layers, it’s about understanding that this is his call, formed in response to the history.’Footnote 25 It is true that the structures of violence exposed in ‘Strange Fruit’ are amply evident in this song and in the rest of Yeezus, yet it is the seeming dialogue between the strained sound of Simone’s altered voice and West’s near-robotic autotuned vocals that is most immediately arresting. If, following Jason Stanyek and Benjamin Piekut, we conceive of this interplay of voices as an outgrowth of late capitalism’s investment in ‘ever-replenishable value, ever-resurrectable labour, ever-revertible production processes’ within the music industry, it also creates an echoing space in which new relationships and new constellations of meaning emerge.Footnote 26
Conclusion
A recent colloquy in the Journal of the American Musicological Society leads with the question: why voice now? As the colloquy convenor Martha Feldman notes, ‘whatever congeries of things we may find voice to be, it remains various and refractory to explanation’.Footnote 27 This richness certainly plays a role in its contemporary significance, but the felt urgency of the voice also stems from its centrality in a neoliberal framework that foregrounds choice and agency even while insisting that those are grounded in a system of individual rights. I have argued here that the voice in Euro-western thought has been caught up in discourses that presume the existence of a unified individual subject who voices (or fails to voice) according to the division of labour between creator and producer. This invests a voice with non-transferable agency and turns the play with voice into a zero-sum game. In re-focusing on vocal multiplicity and abundance within a distinctively intersubjective economy, we can more productively view the voice within a web of entangled agential relations.
One implication of this constellational reframing would be to focus our attention on the different configurations of what Feldman calls the ‘interstitial’ voice, or what I might describe as the way any voice weaves relations among people, objects, and meanings.Footnote 28 Another would be to shift the locus of action and creativity within the world of the sampled/borrowed music. After all, if we were to transpose the myth of Echo onto music sampling, Echo is not the sampled but the sampler. Within this sonic reflection that is simultaneously a transcendence, neither sample nor sampler is passive; rather, each resounds into and alongside the other in a network of collaborative, intra-active music-making. It might also mean a new recognition that the sounds we call our voices might emanate from within, but they turn strange and alien at the moment of their birth.
* * *
A moment of inattention and you’ll miss it, but approximately forty seconds into Childish Gambino’s ‘II. Worldstar’ from Because the Internet (2013), the distorted voice of a young woman interjects ‘Daddy, I love him!’ Gambino’s response is incredulous, noting that this mermaid love is just a ‘first date’, thereby tipping off the listener to the fact that this is a sample of Benson’s Ariel. It is a strange juxtaposition and one that perhaps would only happen in a digital/remix age, but Gambino’s inclusion of Benson’s lovestruck teenager reveals how sampling the voice can amplify, distort, and resignify as it echoes. Listen to this stuff; isn’t it neat?
Open access
