a. A prehistory of poet-voice

The origins of poet-voice are elusive. Complaints about the speaking of poetry go back to classical sources, but the particular objections are often hard to pin down. Plato criticises the rhapsode who ‘drags out a long speech of lamentation’ (Plato, 2000: 327). Aristotle bemoans that ‘actors believe that the audience is incapable of understanding anything unless they emphasize it’, albeit the emphasis in question was gestural more than vocal (Aristotle, 2013: 54). In the early modern period, the discipline of pronuntiatio—taught as one of the five canons of rhetoric (Richards, 2019: 88–97)—gave rise to discussions of preaching style where poet-voice perhaps has a predecessor: Erasmus complained of the new manner adopted by priests of the 1520s who drew each Latin syllable out to the same length and stress (Erasmus, 1985: 427). In English, meanwhile, questions of delivery often hinged on the idea of ‘accent’, which by the end of the 16th century was moving away from technical matters of classical pronunciation and toward a broader sense of good reading style (Alexander, 2021). In Romeo and Juliet (1597), Mercutio inveighs against ‘such antic, lisping, affecting phantasimes, these new tuners of accent’ (Shakespeare, 1997: 1117). ‘Well spoken, with good accent and good discretion’, Polonius says to Hamlet (1599), congratulating him for not overdoing it when he recites a speech to the players; we later gain some sense of what this ‘good accent’ does not consist in when Hamlet tells them that ‘if you mouth it as many of our players do, I had as lief the town-crier spoke my lines’ (Shakespeare, 1997: 1206, 1209).

Detailed accounts of the norms of verse-speaking only start to appear with the elocution movements of the 18th century; from which point onward, remarkably little seems to have changed. The main principle, Thomas Sheridan told his audiences in the 1770s, should be simply to follow the cadence of ordinary speech. The problem is that readers tend to fall back on ‘certain tones and notes of the voice, which differ wholly from those employed in speaking; and which, being but few in number, and adapted to all sentences alike, destroy that endless variety of notes, with which Nature has furnished us, to express the endless variety of sentiments, and emotions of the human mind’ (Sheridan, 1775: 288–289). A similar ‘chanting’ style was often favoured by the Romantics (Perkins, 1991: 657–661). Rather than ‘poet-voice’, the 18th or early-19^th century term was ‘reading with a tone’ or ‘in a tone’: ‘a canting kind of speech, more resembling bad singing than good speaking’, as one encyclopaedist defined it (Rees, 1819). But Coleridge, in a lecture of 1811, mounts a defence of this style:

reading with a tone […] heightens the verse, and does not in any respect lower the sense. I defy any man, who has a true relish of the beauty of versification, to read a canto of The Fairy Queen, or a book of Paradise Lost, without some species of intonation. Every man who reads with true sensibility, especially poetry, must read with a tone, since it conveys, with additional effect, the harmony and rhythm of the verse, without in the slightest degree obscuring the meaning (Coleridge, 1987: 477).

Here we catch a first glimpse of the ambivalence poet-voice produces: ‘resembling bad singing’ to some, but to others, the badge of ‘true sensibility’.

Coleridge’s term ‘true sensibility’ is revealing because it associates his defence of poet-voice with the discourse of sensibility and sentiment that had evolved in the previous century. What had taken place, according to the standard accounts, was the rise of a newly autonomous sphere of feeling. The emerging social forms of the European bourgeoisie had opened up a new role for the expression and recognition of personal affect. William M. Reddy, in his influential work on the history of emotions, characterises the situation as a shift in ‘emotional regime’ from ‘strict’ to ‘loose’ (Reddy, 2001: 124–126), a regime in which literature, and poetry in particular, played a central role (McGann, 1996). So when Coleridge justified poet-voice as the badge of ‘true sensibility’, it was this regime of transmissible affect that he was invoking. The Romantics had inherited and intensified the role of the ‘man of feeling’ par excellence. The poet was the archetype, as Keats put it in 1819, of ‘those to whom the miseries of the world / Are misery, and will not let them rest’ (Keats, 1990: 294), and that tender-heartedness was assumed to be passed on to poetry’s readers and listeners. Coleridge, again, was particularly interested in

the excitement and temporary sympathy of feeling, which the recitation of the poem by an admirer […] calls forth in the audience. For this is really a species of Animal Magnetism, in which the enkindling Reciter, by perpetual comment of looks and tones, lends his own will and apprehensive faculty to his Auditors. They live for the time within the dilated sphere of his intellectual Being (Coleridge, 2014: 407–08).

The poet sensitises the reader, the reader sensitises the listeners, and each is pulled into a structure of ‘temporary sympathy’ in which sense dilates into shared sensibility, or so Coleridge pictures it. But why should ‘reading with a tone’ have this strange magnetism? What made the vocal pattern outlined by MacArthur—reduced pitch range, falling cadences, slow pace—so compelling?

b. A phonetics of poet-voice

Behavioural psychologists have long investigated the relation between vocal patterning and emotion. MacArthur’s set of traits corresponds to an emotional category often labelled simply as ‘sadness’ (Tatham and Morton, 2003: 232). But this finding seems strangely one-sided. If we theorise poet-voice in terms of communication of feeling, why should the feeling it communicates be mournful rather than cheerful? Why should it evoke ‘the miseries of the world’ rather than its joys? It might be better to think of the weepy impression of poet-voice as a way of reading with feeling per se. After all, weeping can be triggered by intensities of other feelings besides sadness, and its associated physiological symptoms—the lump in the throat, the catch in the voice—are often features of aesthetic experience in general and Romantic verse-speaking in particular (Perkins, 1991: 664–665). Silvan Tompkins associated the crying response simply with ‘a general level of density of stimulation’ (Tomkins, 2008: 290). The weepiness of poet-voice might therefore sound like a reasonable approximation simply of someone experiencing ‘the spontaneous overflow of powerful feelings’ (Wordsworth and Coleridge, 2005: 291), regardless of what those feelings are.

The Romantics themselves were fascinated by the flow of feeling from writing to reading, text to voice. Poetry, wrote Hazlitt, has the power of ‘producing, by sympathy, a certain modulation of the voice’ (Hazlitt, 1991: 308). But the flow was by no means seen as all one way, so that it is also the case that ‘as often as articulation passes naturally into intonation, there poetry begins’ (Hazlitt, 1991: 322). Language and voice form a feedback loop. The affective response expressed by poet-voice in the moment of recitation was seen by Hazlitt—and has often been seen by more recent critics—as already crucial in the moment of composition. As Mutlu Konuk Blasing argues, the distinctive tendency of poetic technique is to restore an original oral erogeneity to language use that since infancy is ordinarily excised (Blasing, 2007). Poets and readers of poetry do what Hamlet warns the players not to do with language: they ‘mouth it’, meaning that they let themselves activate a greater than usual bodily engagement. If poet-voice means reading with feeling, in this view, then it is the corollary of an entanglement between reading and feeling that is basic to poetry itself.

So rather than simply a symptom or a show of being affected, poet-voice has always been auto-affective: it embodies and amplifies a feedback loop between reading and feeling, language and the body, that is integral to poetic form. Admittedly, the same might be said of all speech. This contention was central to the early work of Jacques Derrida, whose application to the performance of poetry is analysed by Peter Howarth (2024: 125–130). But in poet-voice, over and above the speaking of poetry in general, the circuit of signification and sensation involved in hearing oneself speak takes on particular force and directiveness. This idea has considerable explanatory power when we turn back to the particular vocal patterning of poet-voice as we know it. What poetry does with the patterning of written language, poet-voice does with the patterns of speech. The poet, Blasing argues, takes their cue from the material resonance of language and embodied histories of feeling. In practice, that means foregrounding phonetic repetition and difference within the text, whether at the level of meter and rhyme or in the more fluid play of rhythm and assonance. ‘The musical in sound is the sustained and continuous; the musical in thought is the sustained and continuous also’, insists Hazlitt (Hazlitt, 1991: 321–322): on the one hand, this describes only one among many sound-worlds that poetry might inhabit, but on the other, the heightened sensitivity to a sustained non-semantic through-line is fundamental even to the most jagged manifestations of modernism. In poet-voice, we see that sensitivity voluptuously recapitulated.

Just as assonance creates a certain embodied response by gluing together the semantics of a text into phonetic sequence, so poet-voice, most notably in its adoption of a ‘sustained and continuous’ pitch range that carries through from one vowel to the next, amplifies the phonetic resonance—and hence the bodily feeling that the text affords—to a maximal extent. This is the logic of Coleridge’s view that to ‘read with a tone […] conveys, with additional effect, the harmony and rhythm of the verse’ (Coleridge, 1987: 477). But not only does it set off the prosody to new effect, it also imposes a parallel effect of its own, a one-size-fits-all euphony that could add reverb even to the flattest prose. Thomas Sheridan chastised poet-voice for smoothing out the phonetic variety of ordinary speech, but in doing so, it extends a smooth vocal surface pre-sensitised to the internal links and echoes of a poem’s musicality. The resemblance of that surface to a cry or a moan might express a bodily response to the music’s effect—but it also has to be understood as a condition of the reader’s responsiveness. By subordinating ordinary semantic prosody to its own drawn-out cadences, the voice makes itself felt and also makes itself feel.

Poet-voice thus has a paradoxical quality of willed passivity. The ‘sustained and continuous’ contour of the reader’s voice seems to work itself up to a maximum of feeling precisely via a minimum of vocal activity. In behavioural psychology, this manner of vocal passivity is sometimes termed ‘lax voice’. It has been interpreted as a response to a situation of reduced ‘control potential’ (Scherer, 1986: 154). Sadness, Scherer argues, is fundamentally linked to powerlessness, and a sad voice is the expressive symptom of a bodily mechanism that has let itself go.² The accompanying physiological explanation is complicated and largely speculative. But this reframing of the question in terms of agency is critical. To some hearers of poet-voice, it is the performer’s apparent abandonment of agency that is most problematic. The Romantic figure of the Aeolian harp, moved only by the air that moves through it, shades into a disappointing automaticity. It is telling that, with the exception of its falling cadences, the standard phonetic traits of poet-voice could imply boredom just as well as being overcome (Scherer, 2003: 233). The contradiction is captured by the poet David Bromige when he complains that ‘I go to these readings and that old singsong starts in—half rapt, half-assed’ (Bromige, 1989: 26).

We have come full circle from Coleridge’s defence of poet-voice back to its critique. To Coleridge, it was the badge of ‘true sensibility’ (Coleridge, 1987: 477) demonstrating and intensifying the reader’s capacity for sympathetic response. To Bromige, it is a pernicious disengagement through which the performer shirks their true responsibility; ‘poet-voice, that is, to lay a given mutual voice on top of the text’, Pester argues, ‘is a kind of opting out’ (Pester, 2019). But this tension is not new, for the discourse that Coleridge invokes has been shadowed by critique of this kind from the outset. 18th century writers acknowledged that ‘true sensibility’ risked shading into mere sentimentality, that is, into a narcissistic identification with the miseries of the world that leaves those miseries untouched (Boltanski, 1999: 96–113). A long archive attests to the historical slippage between being affected in the good sense of moved and being affected in the bad sense of just putting it on. More recently, however, theorists have often returned to the faultline between action and passivity that runs through affect as a conceptual category (e.g. Massumi, 1995). For Derrida, that faultline is in fact the place from which all ethical relation begins: ‘the general structure of auto-affection […] receives the other within the narrow gulf that separates doing from suffering [pâtir]’ (Derrida, 1997: 165). No sympathy, in other words, without the dangerous supplement of showy self-involvement, risking the exasperation that poet-voice so often elicits. But why should this performative pathos be such a touchy subject? Just as the practice of poet-voice has a history, so too does the discomfort it provokes, and the history of that discomfort brings the determination of poet-voice into focus.

c. A politics of poet-voice (or, bourgeois subjectivity makes nothing happen)

I have argued that poet-voice is a technique of the body that takes reading a poem aloud as the occasion to enact the body’s capacity for feeling, and that it has to be understood in the context of post-Enlightenment cultures of sentiment. It belonged to a range of practices through which the intensity of affective relations was cultivated and displayed. But many have agreed with T.S. Eliot that those practices depended on a ‘dissociation of sensibility’ that increasingly treated sentiment as its own domain, simultaneously fetishised and sidelined from more practical matters (Eliot, 2014: 380).³ The two spheres of pathos and logos, sympathy and logistics, the domestic and the economic, feelings and facts, were separate but not equal (no more equal than the separate spheres of gender that these other binaries were increasingly aligned with (Endnotes Collective, 2013)). As affect intensified, it was also sequestered off from industrial capitalism’s more direct modes of relation, most notably in the form of commerce, bureaucracy, and state violence. It might be said that the sentimental half of the dissociated dyad was in fact determined under conditions of diminished ‘control potential’, for sentimentalism’s intensity was locked in precise historical ratio to its impotence. The discourse of sentiment, in this account, was determined as a way of feeling out the impasse of the near-zero control potential of bourgeois subjectivity in a world structured by the objective forces of labour-power and capital.

However, sentimentality can also be seen as a way of feeling in control of this impasse. It offers an imaginary resolution to a real contradiction: in Lauren Berlant’s terms, ‘the turn to sentimental rhetoric at moments of social anxiety constitutes a generic wish for an unconflicted world, one wherein structural inequities, not emotions and intimacies, are epiphenomenal’ (Berlant, 2008: 21). Sentimentality, in other words, belongs to our collective repertoire of genres of wish fulfilment. It offers a consoling and self-aggrandising misrecognition: to accuse something of being sentimental is to say that it opts out of seeing the self-enclosure of the sphere that it operates in, or opts out of seeing that its affective relation to the modern world’s uncontrollable miseries in fact remains under its own control all along. It is to make the same accusation that Hamlet makes of the Player King, horrified and fascinated by the sentimental power that poet-voice could exercise: ‘Is it not monstrous that this player here, / But in a fiction, in a dream of passion, / Could force his soul so to his own conceit […]?’ (Shakespeare, 1997: 1207). As Freud put it, modern socialisation bans magical thinking of this kind, such that ‘control is then exercised by the higher psychical authorities, which have subjected themselves to the reality principle’ (Freud, 2002: 17). That reality principle reasserts itself in every injunction against poet-voice.

But between the desires that motivate poet-voice and the ‘psychical authorities’ that repress it lies a relationship that has to be historicised as well as psychologised. The periodisation invoked so far follows the rise of industrial modernity and its reaction formations in the guise of sentimentality and Romanticism, but the meaning of poet-voice cannot have remained constant over this period. Between then and now, the poetry of the 19th century is often seen as having reached an apogee of sentimentalism, developing what J.H. Prynne termed ‘an amazing degree of control over incantatory techniques, designed to preserve the cocoon of dream-like involvement’ (Prynne, 1963: 291). Poet-voice surely ranks among those techniques. But concurrently, commentators from Habermas onward have argued that the intimate ‘cocoon’ of the affective private sphere was increasingly subsumed into the logistics of the world outside it. The contrast between control in the inward realm of feelings versus uncontrol in the outward realm of facts—the classic structure of the sentimental subject—was becoming more and more blurred, theorised by Deleuze in terms of the new ‘societies of control’ that emerged from the 19th century onward (Deleuze, 1992). So what happens to poet-voice’s embodied performance of uncontrol in the era of the control society? What becomes of the display of disempowered feeling when every subjective tremor can be linked directly into the objective processes of ‘affective labour’ (Hardt, 1999)? To answer that question, we need to bring our explanatory framework into contact with the material conditions in question through two case studies which, between them, will bring our account of poet-voice from its theoretical and historical origins right up to the present moment.

a. Poet-voice versus the media age (1888–1931)

The telecommunications revolution of the late 19th century has often been described as the start of a ‘control revolution’ (Beniger, 1986); among other things, it was a revolution in the degree of control potentially exercised by a voice. Previously, transmission had largely relied on letters and messengers. This was the situation depicted, for example, in Robert Browning’s dramatic monologue titled ‘How They Brought the Good News from Ghent to Aix’ (1845), with its story of three heroic riders carrying a vital piece of military intelligence across 17th century Europe. But Browning’s poem was knowingly archaic: commercial telegraph lines had begun to operate some seven years earlier. The impact of this new technology in opening up new possibilities of distributed social coordination and control was to become widely apparent over the next few decades, notably in the Crimean War (Bektas, 2017). Tennyson’s ‘The Charge of the Light Brigade’ (1854) is a telling artefact of this transitional moment. Famously, the charge in question was the result of a control failure, a jumbled word-of-mouth message whose accuracy could not be checked because the telegraphic systems that relayed news of the disaster to the home front were still too cumbrous to deploy on the battlefield. When Tennyson wrote ‘their’s [sic] not to make reply’ (Tennyson, 2009: 302), in reality, this idealised stoic obedience of the British Empire’s cavalry was a residual feature of a media environment still in the early stages of its transformation.

If we except Edison’s rendition of ‘Mary Had a Little Lamb’, captured on tinfoil in 1878, then Browning’s recitation of a few lines from ‘How They Brought the Good News from Ghent to Aix’ in 1889 and Tennyson’s of ‘The Charge of the Light Brigade’ in 1890 are the two earliest instances of recorded poetry that have come down to us. These early recordings preserve a sample of poet-voice in its peak Victorian form. Both authors read with a consistency of pitch that approaches singing a note (Browning, 1889; Tennyson, 1890). This was a style with which Tennyson in particular was closely associated. Ford Madox Ford, who recollected the ‘ore rotundo’ style—literally, ‘with round mouth’—as a constant feature of the poetry readings of his childhood in the 1870s and ’80s, was under the impression that Tennyson had invented it (Hueffer, 1911: 236).

But the development of sound recording does not simply open a transparent window on poet-voice’s history. It belongs to a nexus of technological change that poetry and its performance were closely tangled up with. Jason Camlot has set out the context of new genres, debates, and possibilities that Tennyson’s reading arose from, including the technical limitations of Victorian recording equipment that required a reader almost to shout to be heard (Camlot, 2019: 100–136). A certain version of poet-voice, with its exaggerated volume and slowness, was produced literally as a recording technique. But the poetics of this moment were being reshaped by their conditions of mediation in more subtle forms as well. The new possibilities of telecommunication in the background of both Browning’s poem and Tennyson’s cast the poets’ own deployments of language in a new light. If poet-voice had previously dramatised the sentimental isolation of the speaking body, its self-contained separateness from the control circuits of the industrial society, then by the fin de siècle that separateness was increasingly coming into question.⁴

‘The Lake Isle of Innisfree’—written by W.B. Yeats in London just four months before Browning made his recording—belongs to a pastoral tradition in which this separateness is deliberately reinscribed. ‘I will arise and go now, and go to Innisfree, / And a small cabin build there, of clay and wattles made’ (Yeats, 1997: 35): this is a poem about the desire to go off-grid. Its central trope is the disjunction between the poet’s subjective inner life and his objective outward surroundings, mapped onto the polarity of the rural past versus the urban present, all coming to a head in the last stanza’s contrast between ‘the pavements grey’ that Yeats sees around him versus the sounds of the Irish countryside that he hears ‘in the deep heart’s core’. The invitation extended to its audience is to connect with the poet’s fantasy of disconnection, to identify with his disidentification from the modern city and what it represents.

When Yeats came to make his first radio broadcast in 1931 and chose to begin with ‘The Lake Isle of Innisfree’ (by then known as his greatest hit), the incantatory vocal style heard on the recording is a case in point of how poet-voice resonates with this urge to disconnect. Yeats’s poet-voice is extreme. ‘I am going to read my poems with great emphasis upon their rhythm, and that may seem strange if you are not used to it’, he warns the listener before beginning (Yeats, 1931a: 0:00)).⁵ But even more remarkable than his handling of rhythm is that of pitch. At first, he roughly carries a single note through each half line, before dropping approximately a semitone a syllable or a few syllables before the cesura or line break; then, when he reaches the second half of each quatrain’s second line or the shorter final line, he starts on a symmetrically lower note before rising to end on the same pitch (Yeats, 1931b). The effect discussed above—of a reading style that sounds profoundly affected, as if speaking in stylised fashion from ‘the deep heart’s core’—is very much in evidence. Ronald Schuchard has shown how for Yeats this style of delivery was bound up with a half-imagined tradition of bardic chanting and magical incantation (Schuchard, 2008). But as we have seen, the effect is a matter of not just mythology but physiology. The poem becomes a space where the voice goes off the grid of ordinary semantic prosody, insulated within the pathos-laden body just as Yeats imagines that body insulated, islanded, from the teeming logistics of the world outside.

Yeats’s 1931 broadcast therefore preserves a recognisably 19^th century configuration of poet-voice. Just like the ill-fated Light Brigade, it is not for Yeats the self-sufficient island dweller ‘to make reply’ to the city life he leaves behind. The lake isle is an ideal scene of reduced control potential, and poet-voice provides its house style. Even here, though, there is a faint indication that this style was starting to shift its coordinates. The tacit claim of poet-voice, I have argued, is to be called forth from the body by affect alone, over and above semantics; ‘producing, by sympathy, a certain modulation of the voice’ (Hazlitt, 1991: 308). What is new, however, is that Yeats found himself increasingly interested in the technical question of how such affect could be deliberately transmitted. A good performer of poetry, he wrote in 1902, is one who ‘understands how to assume that subtle monotony of voice which runs through the nerves like fire’ (Yeats, 2007: 15). But today, he worried, such performers were in short supply. Like the 18^th century elocutionists before him (Joshua Steele, notably, had had the same idea in the 1770s), Yeats and a small group of collaborators started to develop a system for notating the proper intonation of a spoken recital (Schuchard, 2008: 49–79). For Yeats, this was part of a deeply conservative cultural politics, in which a renewed art of recitation could help to renew organic Irish community. But it was also a reaction to a changing media network in which patterns of speech seemed more and more tied up with cheap printing and catchy music hall songs from London (Yeats, 2003: 95). It was no longer enough to rely on the poetic text ‘producing, by sympathy, a certain modulation of the voice’: by a certain modulation of the voice, Yeats hoped, it should be possible to produce sympathy. One commentator replied that Yeats’s system would turn speakers ‘into little more than human phonographs’ (Symons, 1902: 559).

When Yeats started work on ‘The Lake Isle of Innisfree’ in 1888, the project of technically transcribing poet-voice was still in the future. But even in the apparently simple technique of this early poem, we can see that project foreshadowed. It marks Yeats’s arrival at a new and looser prosody in which, in Richard Ellmann’s terms, ‘the accented vowels vary greatly in the amount of time that they force the tongue to linger over them’ (Ellmann, 1964: 124). This variable pacing (on which, see North, 1991: 22–27; Sullivan, 2018) is seen most clearly of all in the short lines that end each stanza, where eight or nine syllables are lengthened to fill the space of 13. This, at any rate, is implicit if the reader wants to preserve the isochronic regularity of the rhythm, with each line taking roughly the same time, and it is in this way that Yeats reads it on the recording. The writing thus actively notates a certain voicing. Yeats saw it as a triumph of self-fashioning: it was ‘my first lyric with anything in its rhythm of my own music’ (Yeats, 1999: 139). ‘The Lake Isle of Innisfree’ not only resonates with the traditions of poet-voice, but also consciously tries to put poet-voice on the page, like a sitcom that comes with its own laugh track. At the moment Yeats was writing, it would still be a few months before this vocal style was first put lastingly on wax, and 40 years before Yeats came to broadcast it himself. But it was already becoming clear that poet-voice was entering a new and contradictory stage in its history. Paradoxically, this way of reading, with its mournful, protesting performance of uncontrol, was beginning to link up with just those control circuits of modern communications that Yeats sought to push back against. The case of Yeats’s poem thus captures a key moment of transition. The sound of speech having passed the threshold of mediation, where could poet-voice’s island fantasies arise and go now?

b. Poet-voice in the control society (2011)

The poem carries its own executable vocal and affective protocol with it: we might say Yeats had learned to code. In the digital era we have come to think of the execution of programming language as something done by machines rather than people. But we have also got used to the possibility that the separate spheres of human control and machinery might blur into one another, even where the privileged category of affect is concerned. It is a mainstay of our theories of the contemporary that the networks of sympathy and the networks of logistics, pathos and logos, have become more enmeshed than ever. So is this the end of poet-voice, whose originating logic, as we have shown, has everything to do with the historical separation between those networks? On the contrary: we have now reached the point where we can see why, in the contemporary moment, poet-voice, though altered in its forms and affordances, has become more widespread than ever.

Consider the following 26-second passage from Kae Tempest’s spoken word poem ‘Renegade’:

you can recognize me cause I’m you, mate / it’s never too late to see deeper than the surface / trust me / there’s so much more to it / there’s a world beyond this one / that creeps in when your wits have gone soft and all the edges start shifting, I mean it / a world that is breathing / heaving its shoulders and weeping / bleeding through open wounds, that’s why I’m grieving / I’m down on my knees and I am feeling everything I’m feeling (Tempest, 2011: 0:32).⁶

The poem which this passage is taken was first published in written form in 2010, but surely reached its widest audience in a one-shot fixed closeup video of Tempest in front of a bookcase speaking the poem into a microphone, filmed in 2011 and posted in 2013, which at the time of writing has received somewhere over 300,000 views on YouTube (Tempest, 2010).

This digital performance can be situated broadly in the sentimental mode discussed so far. Grieving with the ‘world beyond this one’, Tempest belongs to the tradition of those to whom the miseries of the world are misery, whose performance of affect constitutes an invitation to join them in a separate sphere of shared feeling, disempowered but consoling. As with Yeats, Tempest’s delivery is a key part of the performance. As Hannah Silva has shown in a phonetic analysis of a recording of this poem, the narrow, falling pitch range associated with poet-voice is very much in evidence (Silva, 2021: 308–311), and interweaves with the subject matter. ‘Gone soft’, ‘weeping’, ‘down on my knees’, ‘feeling everything’: the cadences of ‘lax voice’, directly linked to a situation of reduced control potential, offer a stylised enactment of the poet’s enforced passivity in the face of others’ suffering.

But they do so in a new mode. The most obvious divergence from traditional poet-voice is simply the speed. While conforming to the template in terms of pitching, Tempest’s performance has none of the slow, halting rhythm that MacArthur found to be typical. The trope of hesitancy remains—the video of ‘Renegade’ begins with a long pause where Tempest fidgets and sways, several times seeming about to start but falling back—but once underway, the delivery is fluent, rhythmically intricate, and relatively rapid.⁷ In fact, in generic terms, its pacing and rhythm are closer to those of hip-hop, as too is the syncopated chain of assonance (‘breathing […] heaving […] weeping […] bleeding […] grieving […] knees and […] feeling […] feeling’) by which the rhythm is demarcated. It is perhaps in this propulsive assonantal patterning where it is most apparent that Tempest considered themself a rap artist before being a poet (Tempest, 2014). Their pacing of poet-voice arises at this intersection of genres and of the projects and priorities each represents.

We saw in the discussion of the phonetics of poet-voice how it bears a particular connection to the device of assonance. When a poem is written, it is through assonance that the auto-affective working of vocal resonance figures most clearly; when it is read aloud, it is assonance that most clearly prearranges the kind of phonetic throughline that poet-voice then picks up and amplifies. Poet-voice’s pitch repetition matches assonance’s phonic repetition. Doubtless, there have been many earlier poets for whom chains of line-internal assonance have become a central generative tactic (notably Dylan Thomas, the mid-century’s poet-voice par excellence). This has to be distinguished, however, from the pivotal role that the device has played in the textual construction of rap, a role often said to begin with Rakim’s lyrics in the late 1980s (Harrell and Fitzgerald, 2023). In rap, assonance works differently. For an acapella reader of a poem, assonance might comprise an invitation to linger, to sustain what Hazlitt called the text’s ‘sustained and continuous’ musicality just a little further through that slackening of vocal control that we have seen to be the crux of poet-voice (Hazlitt, 1991: 322). (Again, Thomas’s recordings are exemplary.) Against a beat, by contrast, the rapper’s voice cannot slacken in this way. Instead, the compression of as great a density of assonance (and other phonetic patterning) as possible into a given rhythmic space has become key to the genre’s characteristic performance of skill. When Rakim raps ‘listen / to the mix and think you’ll sink into the rhyme like quicksand holds and controls you / ’til I leave’ (Eric B and Rakim, 1987: 0:38), it is explicitly control, not uncontrol, that the assonantal patterning serves to project, and in place of the imagined community of sentiment we have the charismatic relation of star to audience. Rakim’s ability to ‘Move the Crowd’ (another song from the same album) refers to not emotion but motion, not passivity but activity, not affect but effect, all dependent on the spectacular pace that his own lyrics move at.⁸ Those lyrics were vital in forming the generic expectation of rap as the archetype of a high control situation. Its success as one of the world’s most popular genres reveals it, unlike poetry, as an artform able to work—albeit often subversively—in sync with the logistics of contemporary mass culture.

In Tempest’s style, the encounter between the archaic melodics of poet-voice and the contemporary rhythms of hip-hop has the character of an overlay between two contradictory historical possibilities. The former insists that Tempest is helplessly, auto-affectively moved by their own performance just as the viewer is. Form and content align in the message that ‘you can recognise me cause I’m you, mate’ (Tempest, 2011: 0:32): Yeats, with his programme of renewed bardic community, would have approved. But Tempest’s rhythmic control offers a different mode of spectatorship, closer to what Miriam Hansen describes as ‘the memory-spectacle rehearsed with each appearance of the overvalued erotic object, the star’ (Hansen, 1994: 281). Tempest is simultaneously a figure of sympathetic identification and idealisation. Under such circumstances, the sentimental workings of poet-voice are altered. But this is not to say that they are foreclosed. The success of ‘Renegade’ suggests how poet-voice in the era of mass mediation takes on a paradoxical new lease of life, with its promise of sentimental connection gaining new currency. Just as the closeup, for Hansen, was an essential part of the early cinema’s repertoire of ‘rituals of mutual recognition between star and fan’ (Hansen, 1994: 282), creating the impression of private intimacy within the new public medium, so in Tempest’s performance poet-voice comes to work in a similar way. Its stylistics offer a closeup of their own (paralleled once again by the video’s filming style). The ‘temporary sympathy of feeling’ that poet-voice makes available, drawn from the long archives of performative commiseration, is supplemented by the imagined thrill of being recognised by the face on the screen.

If today, poet-voice has gained a greater prominence than at any point in its long history, then Tempest’s example is key. At least in a British context, it would hardly be an exaggeration to say that Pester’s polemic in 2019 might just as accurately have been titled ‘Against Kae Tempest-Voice’. Another performance poet, Luke Wright, explains the conjuncture as follows:

In the sort of, like, I guess the late nineties, early, very early two-thousands, we’d go from town to town. […] And every town you went to, […] there was a voice. There was a Bristol voice. There was a Brighton voice. There was a Manchester voice. There was a Norwich style. […] And then what happened—later on, if you fast-forward to, so, about 2010, everyone sounds like Kae Tempest, because by then we’ve got YouTube, and these viral videos. […] I think YouTube, more than anything else, is what has enabled spoken word to flourish. Because you know, YouTube is—spoken word is ideally suited to YouTube. […] And it was amazing to see a young sixteen-year-old Geordie lass, like, you know, like, clutching her stomach and doing the Kae Tempest thing (Wright, 2023: 21:54).

In one sense, this viral quality has been part of poet-voice’s structure all along. The transference between author and spectator was already implicit when Coleridge’s described how ‘the enkindling Reciter […] lends his own will and apprehensive faculty to his Auditors’ (Coleridge, 2014: 407–08). But a final important transformation has to be noted: the contagious spread of poet-voice described by Wright has ceased to be simply a style of reading. Historically, poet-voice was a stylised performance of affective receptivity, a passive protest against the immiserating effects of the mode of production surrounding it. Today, the border between receptivity and productivity has blurred to the point that poet-voice has been rerouted into a mode of public self-fashioning. But then again, perhaps this has been the secret of poet-voice from the first. If ‘spoken word is ideally suited to YouTube’, perhaps this is because the desire of poetry’s readers to make of their reading a recognition scene for the pathos of distant others has never, in fact, been truly dissociable from the desire for recognition that social media has such a prevailing logic of contemporary life.

In Wright’s wistfulness for the lost diversity of styles in the UK poetry scene, we can once again make out a tactful reproach to poet-voice. We are also returned, however, to the fact of its unstoppable appeal, an appeal that today the proliferation of other, rival techniques for the social mediation of affect seems only to have enhanced. The separate sphere of private sentiment where three centuries ago poet-voice first took form may have lost its boundaries. But the wish to carve out a bounded, protected, imaginary space of reciprocal feeling in an unfeeling world is no less strong, or all the stronger for its anachronistic impossibility. Once again, the reading aloud of poetry, through its exemplary encounter between language and the body, mediation and affect, has come to provide the privileged cultural site where that wish goes to find fulfilment. Poetry continues to be popularly perceived as a space where pathos acquires a unique kind of social force; voiced as such, in turn, it continues to provoke the same reaction of stern logical inhibition as any process of wish-fulfilment that gives itself away too publicly. We can expect the profound ambivalence produced by poet-voice to continue for as long as we inhabit a system of collective life whose objective miseries seem to lie beyond our collective control.