A Voice of Her Own? Feminine Voice Effects in George Du Maurier, George Bernard Shaw and Isak Dinesen


The focus of Pygmalion, George Bernard Shaw’s 1913 stage classic, is on the linguistic and vocal makeover of the flower girl Eliza Doolittle. In the first act of the play, Professor Henry Higgins happens to hear the young woman at the Covent Garden vegetable market and immediately begins to phonetically transcribe her cockney accent typical of the working class of the London East End. In spite of his unbearably snobbish and condescending manner, she later decides to ask Higgins to teach her to speak ‘proper English’ as she would like to improve her social position. The professor on the other hand bets his friend Colonel Pickering that he will be able to get rid of Eliza’s accent and to introduce her as an alleged duchess to High Society. With this project of a modern Pygmalion, Shaw presents a provocative refiguration of the ancient myth as we know it from Ovid’s Metamorphoses (Ovid 2005, p. 350–352)._2 In it, the bachelor and sculptor Pygmalion creates a supposedly ideal woman in the form of a statue and promptly falls in love with his own creation. He views, kisses, touches and showers gifts upon the ivory sculpture, and upon his beseeching Venus brings his work of art to life. Galatea’s figure – which is both eroticized as well as animated – represents a palpable materialization of his sexual and creative desire. The woman appears literally as the pure projection of the male artist protagonist. The question of her subjectivity is thus entirely elided in favour of his artistic self‑expression.

As in the case of Pygmalion in the ancient myth, we can also see Higgins in terms of a self‑projection. Instead of recognizing the essential share Eliza has in her makeover, he regards his bet as an opportunity to demonstrate his powerful knowledge, abilities and skills. In contrast to the sculptor, the phonetician does not model the bodily shape, but the pronunciation of the woman. “[Y]ou have no idea,” he once explains to another figure, “how frightfully interesting it is to take a human being and change her into a quite different human being by creating a new speech for her” (Shaw 2003, p. 65). Higgins is clearly an experienced ‘reader’ of minute phonetic differences and can place individuals precisely in the social geography of the British class society on the mere basis of their pronunciation. Yet in contrast to the other characters in the play, he is unable to recognize Eliza’s ‘voice’ which Shaw uses as a cipher for her subjectivity and the fact that she is a person in her own right.

After the experiment has come to a successful conclusion – Eliza is taken to be a princess at an ambassador’s reception – Higgins is celebrating himself as the winner of the bet. Unlike Pygmalion’s sculptural Galatea, who remains silent even after being animated, Eliza objects. She confronts Higgins with the fact that his supposed success is essentially her own achievement. She also insists that she has been instrumental in winning his bet, but does not seem to matter as a person. After trying to make her point in this confrontation, she leaves Higgins. At the end of the play, he admits that he misses her voice and appearance. He has indeed for months shaped the pronunciation of his Galatea, but only when left to himself does he perceive the ‘voice’ Eliza has as a subject with her own fantasies and desires. She however insists: “I notice that you dont [sic] notice me” (Shaw 2003, p. 101). Indeed the question of whether Higgins realizes that his project was essentially bent on effacing her voice remains open to the very end.

It is not by chance that Shaw wrote Pygmalion during an important phase in the history of the struggle for the women’s suffrage in Great Britain, a political goal the author strongly supported. In a manner similar to the suffragettes, who at times vociferously appeared in public to express their demand for a political voice, Eliza also raises her concrete voice so that her ‘voice’ might find recognition. Shaw’s play and Eliza Doolittle, his vocally assertive character, underline that a discussion of femininity and voice hinges not only upon the concrete audible sound (and much less on its reduction to phonetically distinctive features) but always also upon the voice in a metaphorical sense. Over and above the concrete sound of voices, which can actually be heard, the voice also carries a wide range of metaphorical associations, including the notion of one’s ‘inner voice’ or the political question of who is given a voice. Importantly enough, to have a voice is not synonymous with having a voice of one’s own. Especially in a gender context, voice is always also a political category. It is decisive whether the woman in question can speak for herself or whether someone else speaks for her. Does her singular voice find a hearing, or is it absorbed by another voice and used in its interest?_3

These questions also inform my readings of George Du Maurier’s novel Trilby (1894) and Isak Dinesen’s story The Dreamers (1934). As my contribution will show, the late Victorian novelist and the modern storyteller construe, and construct, the female voice in markedly distinct ways. At the same time, they both deal with the voice of a female singer and, in so doing, refer to the same cultural tradition. While women in nineteenth‑century bourgeois culture hardly had a voice in the political public arena, we can observe an undisguised fascination with public female voices appearing in the theatre from this very period. On the opera stage, female singers begin to sing with great passion and devotion. In romantic opera for example the soprano rates as the musically privileged voice par excellence. A similar vogue can also be observed in literature. Nineteenth‑century texts feature strikingly numerous female singers and actresses who perform, speak and/or sing in public._4 This raises the question as to what degree the literal voice of these performers correlates with a voice in a figurative sense. Does the vocal presence of a female singer signify that she has a voice of her own? Is her figure subsumed by other characters so that similar to the Galatea sculpture, which represents Pygmalion’s desire, her voice turns into an acoustic mirror?_5 Or does the voice possibly bring into play an element than is inalienable, and if so, how is this indicated in the literary narrative text?

The Special Effects of Voice

As has often been observed, the voice represents a curiously elusive phenomenon, which cannot be pinned down to a single category. It oscillates, for instance, between spirit and body, between presence and evanescence._6 Traditionally the voice has often been understood as a privileged marker of individual selfhood. After all the timbre of any and every individual voice is distinctive and unique. At the same time, voices are never isolated. In fact, it is important to note that the voice mediates between the individual and the collective, between the self and the other. Every voice constitutes itself through dialogue: it is only by addressing others and by being heard that an individual voice comes into existence. Moreover, the voice marks the embedding of the subject in a cultural community. Voices never just resonate with themselves alone but always also echo with other cultural voices, texts and traces. Indeed subjects are defined among other things by the very fact that they are spoken and/or sung by language, culture and history.

In his book A Voice and Nothing More (2006), Mladen Dolar further suggests that the voice forms a surplus or excess. There is a dimension of voice which exceeds verbal language. Voices do not only and not always transport meaning, but they also harbour a surplus that is potentially disturbing. Dolar’s point of departure is Jacques Derrida’s well‑known critique of Western phonocentrism. Ever since Plato, Derrida (1997) argues, the metaphysics of presence has founded itself on the voice. The individual who hears him‑ or herself speak serves as a trope for transparent meaning and self‑presence, while writing is relegated to a mere supplement, secondary to the fullness of speech. However, as Dolar points out, there is an aspect of voice which disrupts, rather than supports self‑presence. Although the voice can inspire both speakers and listeners with a reassuring sense of self‑identity, there is a vocal dimension which moves us beyond symbolic and imaginary codes. As a result, the voice is at once uniquely familiar and radically alien.

But how can the voice be conceptualized for a discussion of literary texts? Clearly a theory of the voice such as Dolar’s is not a text theory. Nor are there any concrete voices in narrative fiction that could actually be heard._7 The voice, in other words, is alien to the type of literary language under discussion. Nevertheless, I want to ask whether literary texts can be seen to evoke virtually what escapes them medially. How do narrative texts produce voice as an aesthetic special effect? Similar to concrete voices, which are shot through with various colours, affects, moods and intonations, textual voices also mark a surplus – namely in the form of a multilayered complexity they introduce into a text. Important concepts here are Mikhail Bakhtin’s notions of dialogue, polyphony and heteroglossia, which he uses to describe the discourse of the novel as an arena of different, and potentially differing, voices. According to Bakhtin (1981), narrative fiction occupies a special status as a literary genre and aesthetic medium which juxtaposes various textual voices and hence confronts different social accents, positions and perspectives. It is through the dialogue or even the ideological discord of various character voices and the positions of the narrator and the implied author that narrative fiction puts the multiplicity of textual voices centre stage.

Moreover, I suggest that with respect to narrative fiction we can speak of a ‘voice effect’ in analogy to what Shoshana Felman (Felman 2007, p. 15–22) calls a ‘reading effect’. As Felman emphasizes, the effect of a text lies not in its thematic subject alone, but also in the ways in which it speaks to its readers (ibid., p. 18). Its effect resides in our relation to the text as well as in the impact the text has on us. As readers and critics we not only interpret texts but actually reproduce them (ibid., p. 21). This is an idea that can be fruitfully transferred and applied to issues of voice not merely because any voice in narrative fiction is entirely spectral and acousmatic since it has to be imagined and evoked in the solitary act of reading. The ‘voice effect’, as I define it, also allows us to bring into play the tone or ‘voice’ of a literary text as we pay attention to, and indeed foreground, the dialogue and/or dissonance between various textual voices.

Literary examples such as Du Maurier’s or Dinesen’s narrative texts about female singers link several vocal levels. The singer’s voice described on the thematic level raises the question as to what degree the concrete voice in her singing stands for a ‘voice’ in the sense of a self‑determined position and agency. The dialogue between the different narrative and character voices in turn refers us to the question of the ‘voice’ of the text: What is the tone that emerges as a result of the way in which the text in question orchestrates the various textual voices?


Trilby, the immensely popular late Victorian novel by George Du Maurier, the French‑born British author and illustrator, resembles Shaw’s later Pygmalion in that it also revolves around the vocal makeover of a female character. Du Maurier’s fin‑de‑siècle text, which is set in Paris in the mid‑nineteenth century, introduces its protagonist through her unusual voice. When Trilby O’Ferrall, who was raised in France as the daughter of an Irishman and a Scotswoman and who works as a model in different Paris art studios, makes her first appearance in the text, her voice is described before anything else: “a portentous voice of great volume […] that might almost have belonged to any sex” (Du Maurier 1998, p. 12). Her sexually undefined voice envelops an immense volume, yet Trilby lacks any sense of musicality. When at an early point in the novel she sets out to sing a song, what is immediately evident is the remarkably uncontrollable character of her boundless voice. “It was as though she could never once have deviated into tune, never once have hit upon a true note, even by a fluke – in fact, as though she were absolutely tone‑deaf, and without ear […]” (ibid., p. 19). In a radical reversal of the grotesque song which Trilby produces at the outset of the novel and which anticipates Higgins’ characterization of Eliza’s cockney accent as monstrous, we later get to hear a voice that has undergone a miraculous change. After a prolonged absence, Trilby has turned into an internationally acclaimed singer called ‘La Svengali’. As before her deep contralto voice envelops an immense volume with the crucial difference that she now intones her song with immaculate purity.

As the novel emphasizes in an extensive description of one of her concerts, La Svengali transports her audience to ecstatic rapture. However, while the sublime singing of the stage star may be irresistible, Trilby’s ‘voice’ remains inaudible. As the readers gradually learn, Trilby in her persona as a spectacular singer finds herself under the hypnosis of the Jewish musician Svengali. Since Svengali himself has no singing voice that would permit him to express his musical genius, he hypnotizes Trilby, appropriates her voice and, as it were, ‘sings’ through her._8 Du Maurier’s musical genius thus turns out to be a male artist in Pygmalion’s tradition. In a manner reminiscent of the sculptor who in the ancient myth fashions the ideal woman in the form of a statue, Svengali creates the perfect stage star, and as later Shaw’s Professor Higgins, Svengali concentrates on the voice. But in contrast to Eliza Doolittle, Trilby cannot resist this appropriation. As the hypnotized medium and most literal instrument of another voice, she is aware neither of her vocal transformation nor of her stellar career.

However, Svengali is not alone in his instrumentalization of Trilby. The three British amateur artists Little Billee, Taffy and the Laird, who have devoted themselves to the study of art in Paris, have all fallen in love with the young woman and claim her for their respective interests. In fact, before she falls under Svengali’s spell, Trilby serves them as a model for their art. First the three Britons appear fascinated by the bizarre appearance of the tall, almost gigantic Trilby, who at the opening of the novel wears the overcoat of a military uniform over her short underskirt and who in addition to her standard English speaks a slangy French. Gradually however, under the influence of the Britons, the initially self‑assured Trilby is feminized and domesticated. Outwardly her transformation is reflected by her increasingly refined looks, which according to the text anticipate the ideal of beauty later developed by the English Pre‑Raphaelites (Du Maurier 1998, p. 90). At the same time, the culturally hybrid Trilby drops her French slang and reads the British novels lent by the three artists. As the narrator affirmatively states and comments: “She grew more English every day; and that was a good thing” (ibid., p. 64).

Given that the three British artists as well as the Jewish musician all claim Trilby for themselves, we may well speak of an antagonism between these different character voices. The tension between the divergent positions becomes particularly manifest in a text passage towards the end of the novel in which a violinist explains the phenomenon of Trilby’s two disparate voices to one of the British artists and clearly sides with the British artist trio.

There were two Trilbys. There was the Trilby you knew, who could not sing one single note in tune. […] She could no more sing than a fiddle can play itself! […] Well, that was Trilby, your Trilby! that was my Trilby too […].
But all at once – pr‑r‑r‑out! presto! augenblick! … with one wave of his hand over her – with one look of his eye – with a word – Svengali could turn her into the other Trilby, his Trilby – and make her do whatever he liked […].
He had but to say “ Dors! ” and she suddenly became an unconscious Trilby of marble, who could produce wonderful sounds – just the sounds he wanted, and nothing else – and think his thoughts and wish his wishes – and love him at his bidding with a strange, unreal, factitious love … just his own love for himself turned inside out – á [sic] l’envers – and reflected back on him, as from a mirror … un écho, un simulacre, quoi! pas autre chose! […]
Well, that was the Trilby he taught how to sing […]. That Trilby was just a singing‑machine – an organ to play upon – an instrument of music – a Stradivarius – a flexible flageolet of flesh and blood – a voice, and nothing more – just the unconscious voice that Svengali sang with […]. […] When Svengali’s Trilby was being taught to sing … when Svengali’s Trilby was singing – or seemed to you as if she were singing – our Trilby had ceased to exist … our Trilby was fast asleep … in fact, our Trilby was dead … (Du Maurier 1998, p. 298–299).

As the narrative perspective of this passage suggests, the novel is by no means neutral in its orchestration of the various character voices. The implied author Du Maurier directs sympathies unequivocally towards the three British artists. Svengali, the eastern European Jew and ingenious musician, is in turn demonized in the text in stereotypical fashion, namely as an anti‑Semitic vampire figure, who, like Count Dracula in Bram Stoker’s novel published three years later, haunts the British characters as a menacingly alien threat coming from the East.

Yet not only Svengali, but also the three British artists privileged by the text contribute to the silencing of the female voice. Their feminization of Trilby implicitly anticipates her passive state under hypnosis, while their enjoyment of La Svengali’s singing reinforces the negation and ultimately the extinction of her voice. When towards the end of Trilby, the eponymous heroine lies dying, literally exhausted by having been absorbed as the medium of another voice, Svengali’s photographic portrait arrives, sent according to his will after his death. As it were from the grave he hypnotizes Trilby one last time, whereupon she intones her swan song. For one the deathbed scene illustrates the fatal effect of Svengali; his vocal appropriation literally extinguishes Trilby. For another the passage shows that the three British artists owe their highest enjoyment to a both aestheticized and usurped voice. As already in an earlier concert scene, they enjoy the sublime as well as perfect singing of the singer La Svengali, from which Trilby’s monstrous voice has been utterly obliterated. What is even more, they are ecstatically moved to tears while Trilby under Svengali’s hypnosis is literally sung to death.

She hardly seemed to breathe as the notes came pouring out, without words – mere vocalising. It was as if breath were unnecessary for so little voice as she was using, though there was enough of it to fill the room – to fill the house – to drown her small audience in holy, heavenly sweetness.
The usual effect was produced. Tears were streaming down the cheeks of […] Little Billee. Tears were in the Laird’s eyes, a tear on one of Taffy’s whiskers – tears of sheer delight (Du Maurier 1998, p. 283).

Of course Du Maurier’s starkly coloured novel can be read against the grain. We can for example ask ourselves whether the violent vehemence with which the feminine voice in this novel is contained and controlled by the male characters as well as the implied author possibly testifies to its subversive power. Still the tone or the ‘voice’ of the text remains misogynistic. Du Maurier’s novel puts a feminine voice centre stage – only to strip it of any agency and independence. In the course of the novel the female protagonist turns into an internationally celebrated voice, but only thanks to a male artistic genius figure. Instead of singing with her own artistic voice, Trilby fulfils the function of an acoustic mirror: La Svengali’s sublime singing mirrors the desire and the enjoyment of the three British artists in the same manner as it assures Svengali in his artistic self‑expression. The question of a female subjectivity is thus elided. What the woman really wants – so the novel suggests with the figure of the singer who unconsciously sings under hypnosis – is not just to sing, but to turn into the mechanical instrument or the selfless medium of another voice.

The Dreamers

As a text that draws on both romantic and Gothic traditions, George Du Maurier’s Trilby forms part of the literary archive of texts which Isak Dinesen, the Danish author, often rewrites in her stories. Like Du Maurier’s novel, Dinesen’s text The Dreamers from her first story collection Seven Gothic Tales revolves around an extraordinary female singer in the mid‑nineteenth century. Yet in contrast to Du Maurier, whose singer appears as the creature of a male Pygmalion figure, Dinesen is interested in the self‑fashioning of the female artist. When the author Baroness Karen Blixen invented herself as a professional writer in the mid‑1930s and published her English texts under the masculine pen name Isak Dinesen, she created a highly artificial public persona, which she ostentatiously staged as a flamboyant mask and masquerade._9 In a photographic portrait for example, the writer can be seen to wear a pierrot costume she already possessed in her youth, while her strong make‑up underlines the mask‑like quality of her face._10 Dinesen, who identified with role of the storyteller, also fictionalized herself by using a score of other names in addition to her nom de plume._11 She would assert that she was 3000 years old and had already dined with Socrates, or again that she had promised her soul to the devil so as to be able to turn her life into tales._12

Isak Dinesen’s extravagant self‑staging was crucial to her artistic self‑articulation, and a similar self‑fashioning of the feminine artist also underpins her story The Dreamers. Significantly enough, Pellegrina Leoni, the literary singer figure in her text, is not ‘sung’ by another voice but presents herself as a consummate artist in her own right. As Marcus Cocoza, her Jewish impresario relates, the spectacular stage star is motivated by two great passions.

She had in her life two great, devouring passions, which meant everything to her proud heart.
The first was her passion for the great soprano, Pellegrina Leoni. […] In her relation to this idol she had no forbearance and no rest. […] She worked in the service of Pellegrina Leoni like a slave under the whip, weeping, dying at times, when it was demanded of her.
She was a devil to the other women of the opera, for she needs must have all the parts for Pellegrina. She was indignant because it was impossible for her to perform two rôles [sic] within the same opera. […]
And the other great passion […] of this great heart was her love for her audience. And that was not for the great people […] but for her galleries. Those poor people of the back streets and market places […] she loved them beyond everything in the world. […]
And she was adored by the people (Dinesen 1969, p. 402–405).

Driven by her boundless love for her audience and her unconditional commitment to her star persona, Dinesen’s singer is ready to undergo any sacrifice in order to turn herself into an idol which is revered and worshipped by her audience as well as by herself. On the one hand, her voice is composite, multiple and plural because she performs various operatic roles. When necessary, that is “when it was demanded of her” by the operatic scripts and scores, she will weep and die – only to constantly resurrect herself on stage. On the other hand, her voice is monologic because she privileges one single role. As mentioned by Marcus Cocoza, “she needs must have all the parts for Pellegrina”. Instead of actually transforming herself into various operatic figures and lending them her voice, she lays claim to all parts so as to use them as her own vehicle and subsume them under her one superlative role, “the great soprano Pellegrina Leoni”.

When a fire breaks out during a performance of Mozart’s Don Giovanni at the opera house in Milan, the singer is hit by a falling beam. She makes a narrow escape from death, but as a result of the shock, she loses her voice as a singer. How is this near‑fatal accident to be read? With the loss of the singer’s voice, Dinesen’s tale can be seen to refer to such narratives of the nineteenth century in which female performer figures routinely fall silent at the end._13 Seemingly in this tradition, The Dreamers reverberates with a tragic sense of loss. Indeed the singer leaves the stage forever, never to be heard again. However, by using what would be a classic ending in a nineteenth‑century text as her point of departure, Dinesen reshifts the narrative argument. Rather than stressing the tragic loss, the accident can in fact be seen to point to a problem and even a deadlock in Pellegrina’s persona as a star singer. Investing everything in one single role, the incident suggests, is fatal.

Dinesen’s artist loses her professional voice and her social role as a revered idol, but she gains access to a different form of self‑articulation in exchange. After the star soprano Pellegrina Leoni has been buried in a public fake funeral, she explains to Marcus Cocoza, her former impresario: “There are many that I can be. […] I will not be one person again, Marcus, I will be always many persons from now. Never again will I have my heart and my whole life bound up with one woman, to suffer so much” (Dinesen 1969, p. 417–418). What Pellegrina opts for is a protean performance of the self. Having realized the lethal effects of a single self‑construction, she fragments her former persona into a myriad of masks and masquerades: “I will not be one person again”, and “[t]here are many that I can be”. Following her wish never again to be trapped and caught up in one single role, Pellegrina turns into a traveller, and for every one of the lovers she encounters on her journey, she plays a different part. To the Englishman Lincoln Forsner in Rome, she presents herself as the courtesan Olalla, to Friedrich Hohenemser in Lucerne as the milliner and revolutionary Madame Lola and to Baron Guildenstern in Saumur as Saint Rosalba. Whenever a part becomes too constraining, her former Jewish impresario Marcus turns up and helps her to disappear and to assume a new mask. While Svengali appropriates Trilby’s voice for his own expression, Marcus supports the former singer Pellegrina in her protean self‑staging. The actual theatre may have gone up in smoke and flames, but because of the peregrinations of Pellegrina, the wanderer, all the world becomes a stage.

Pellegrina serves as a figure who reflects Dinesen’s poetics. She refers us to a voice which disappears in order to resound as pure text. Her voice can no longer be attached to a person but instead dissolves into writing. Importantly enough, Dinesen’s narrative technique recounts the story of Pellegrina’s voice through the reports of other character voices. Her stellar career, the loss of her voice and her embodiment of multiple roles are the chronological events of the story which we can only reconstruct after having read the entire text. The actual structure of the text, however, consists of several narratives that frame Pellegrina, who is almost the only figure not to tell a story. The first narrative level of the authorial narrator frames a second‑level narrative, that is, the story that Lincoln Forsner tells to the famous but weary storyteller Mira Jama. According to Lincoln, he was searching for Olalla, the prostitute who had made him a dreamer, when he met Hohenemser and Guildenstern, his two friends who, in turn, told him their stories about Madame Lola and Rosalba.

This complex narrative structure refers us to the voices in the text or, as Mikhail Bakhtin would put it, its narrative polyphony. As mentioned before, Bakhtin regards narrative texts as fundamentally dialogic. They stage the dissonances between various narrator and character voices together with their various social positions, interest and accents. The Dreamers enacts a heightened form of polyphony. For in contrast to Trilby, Dinesen’s story revolves not just around the antagonism between the voices of the various male character‑narrators, namely Lincoln, Hohenemser and Guildenstern. The fiercest conflict can be observed between their narrative desire and Pellegrina’s metamorphoses. While Pellegrina continuously reinvents herself so as to avoid being read and appropriated, each of the three male characters seeks to reduce her to the particular role which she plays in his respective story. Fixations and transgressions are thus mutually implicated.

The conflict between Pellegrina and the three men reaches its climax in a stormy winter night. Lincoln, Hohenemser and Guildenstern have just told each other their stories about Olalla, Madame Lola and Rosalba in a hotel in the Swiss Alps when, all of a sudden, they see a veiled woman. All three believe to recognize their respective lover and chase the woman as she is running towards a mountain pass in the blizzard. Because she rejects and escapes any one defining role, the question Lincoln asks her when he finally catches up with her is inevitable.

[…] Who are you?’
She did not turn, or look at me. But the next moment she did what I had always feared that she might do: she spread out her wings and flew away. […] she threw herself from the earth clear into the abyss, and disappeared from our sight.
[…] I thought then of how it had been my question to her which had driven her into this great white full‑moon death, in the end (Dinesen 1969, p. 395–397).

Pellegrina refuses to answer and tries to escape from this scene of interpellation altogether. The attempt to limit her to one single identity turns out to be lethal in the last instance; as the voice did before, her bodily figure disappears as she leaps off the precipice and into the void and thus withdraws from the visual field by literally vanishing off the face of the earth.

What then is the ‘voice’ of Dinesen’s text? By asking this question, I am not trying to make out one single voice that would unite or even reconcile the various textual voices. Instead the question is one of tone, namely the tone that emerges as a result of the ways in which the implied author orchestrates the various voices in the text. In contrast to Trilby the narrative mode of The Dreamers is feminine, or even feminist, in the way in which it shows, and actually performs, the violence used by the male figures in their narrative framing of Pellegrina. Significantly enough, it is only as the woman lies dying that Marcus tells the name and story of the great soprano Pellegrina Leoni. In a deft gesture, Dinesen has his belated commentary on the singer’s rigid star persona coincide with the actual death of the woman. Or put differently, the dramaturgy of the text suggests that her death is brought about by a narrative desire that seeks to reduce her to one single role.

Yet the feminine mode of Dinesen’s text goes further than that: shortly before Pellegrina passes away, her narrative containment is disrupted by a strange voice effect.

Her whole body vibrated under her passion like the string of an instrument. ‘Oh,’ she cried, ‘look, look here! It is Pellegrina Leoni – it is she, it is she herself again – she is back. Pellegrina, the greatest singer, poor Pellegrina, she is on the stage again. To the honour of God, as before. Oh, she is here, it is she – Pellegrina, Pellegrina herself!’
It was unbelievable that, half dead as she was, she could house this storm of woe and triumph. It was, of course, her swan song.
The old Jew was in a terrible state of pain and strain. […]
Of a sudden he took up his little walking stick and struck three short strokes on the side of the stretcher.
‘Donna Pellegrina Leoni,’ he cried in a clear voice. ‘En scène pour le deux [sic].’
Like a soldier to the call, or a war horse to the blast of the trumpet, she collected herself at his words. Within the next minute she became quiet in a gallant and deadly calm. She gave him a glance from her enormous dark eyes. In one mighty movement, like that of a billow rising and sinking, she lifted the middle of her body. A strange sound, like the distant roar of a great animal, came from her breast. Slowly the flames in her face sank, and an ashen grey covered it instead. Her body fell back, stretched itself out and lay quite still, and she was dead (Dinesen 1969, p. 426–427).

Initially, Pellegrina’s famous singer persona seems to be reconstituted by the cue given by her impresario. She slips into her former role as the star soprano Pellegrina Leoni, ready to resume her part as Donna Anna in Mozart’s Don Giovanni at precisely the point at which she was interrupted by the near‑fatal accident. The accumulation of the words “she”, “herself” and “Pellegrina Leoni” in her speech suggests unreserved identification with the role of the singer: “It is Pellegrina Leoni – it is she herself again […]. Oh, she is here, it is she – Pellegrina, Pellegrina herself!” Yet ironically the singer Leoni, a figure to whom she refers almost exclusively in the third person, is just as much a mask as all her other roles. Rather than making a theatrical comeback, she has never actually left the stage. And indeed, while the other character voices seek to unmask her, this last performance shows not just the fatality but also the inherent impossibility of their attempt to lay bare and thus expose her identity.

What is even more disruptive, however, is the culmination of Pellegrina’s swan song in a monstrous utterance towards the end of the passage – “a strange sound, like the distant roar of a great animal”. With verbal means, Dinesen’s text evokes the effect of a voice beyond all meaning by comparing her articulation to a non‑human utterance. The sheer sound of Pellegrina’s voice subverts not just the narrative desire of the three men. It also undercuts all symbolic and imaginary codes – and hence all social roles. The sublime singing Pellegrina used to produce as a singer allowed her and her audience to mirror themselves in each other. Her swan song articulates the very reverse, that is, a voice of radical alterity. It is at this point that we can invoke Mladen Dolar’s notion of the voice as a surplus or excess. The passage marks a moment where the voice articulates a radical subjectivity beyond cultural interpellation. What becomes audible is an aspect of the subject that is absolutely singular and intimate.

Among the elements that underline the reification of the singer in Du Maurier’s novel Trilby are the fan pictures that display the star in ancient dress and in the pose of the Venus of Milo._14 The photographic portrait of La Svengali, which is reproduced and offered for sale, refers us not only to the commodification of the female performer in the context of modern celebrity culture, but also to her instrumentalization as art material. Trilby may be able, under hypnosis, to transform well‑known melodies into an endless series of variations._15 However, at the same time, her superlative song arrests her in a quasi‑sculptural pose. As “an unconscious Trilby of marble, who could produce wonderful sounds – just the sounds he wanted, and nothing else –”, she is, similar to the statue of Galatea, nothing more and nothing less than the projection of a male artist. If Du Maurier’s late Victorian novel revolves around a misogynist scenario of artistic mediumship and ventriloquism, Dinesen’s modern tale proposes an altogether different trajectory. In The Dreamers, the ambivalent interplay of audiovisual ‘masks’ and the complex orchestration of the different voices can be read as a literary critique not only of the cult surrounding the modern star, but also of a concept of creativity in which feminine life and/or subjectivity are sacrificed in favour of masculine expression and art production._16

Dinesen’s protagonist has a ‘voice’ although, or perhaps rather because she has lost her concrete voice as a star singer. When her accident breaks open her rigid star persona, she embarks on a self‑staging that is both protean and evanescent. By suspending the boundary between life and art in a typically modernist gesture and by transposing her theatrical scenarios into everyday life, the former singer recreates herself, both as an artist as well as her own work of art. Isak Dinesen’s polyphonic text formulates an idea of modern feminine subjectivity which articulates itself in a two‑fold manner: as a radical self‑fashioning of the feminine artist who conceives of herself as her own work of art, and also as a radical alterity which eludes any appropriation. Dinesen’s fictional alter ego may not ultimately sustain her fleeting, changeable self‑fashioning in view of the narrative desire of her three lovers to fix her in a certain role. What is important, however, is the vocal effect of the text. In my reading I have been exploring the various voices that are at stake in The Dreamers. It is by tracing their relation that we can fill a text with our own critical tone. The voice as an aesthetic category in the sense of Bakhtin’s polyphony is closely connected to a political dimension. Or put differently, the voice effect of narrative fiction involves us as readers and critics and refers us to an ethical dimension of literature. It is by discerning various voices that we lend out ears to the moral imagination of literature. Pellegrina may die and thus fall silent. Yet Dinesen’s complex orchestration of voices refers us to the ‘voice’ her fictional artist has as a subject and, in so doing, positions her as the most resonant voice.


The original German version of this text will be published in Sabina Brandt – Maren Butte (eds.): Bild und Stimme. Fink, Munich (forthcoming)
On the cultural afterlife of the Pygmalion myth see the essay collection edited by Mathias Mayer – Gerhard Neumann (1997) as well as The Pygmalion Effect: From Ovid to Hitchcock by Victor Stoichita (2008).
My discussion of the difference between having a voice and having a voice of one’s own is indebted to Stanley Cavell’s discussion of George Cukor’s film Gaslight (1944). Cukor’s melodrama revolves around the niece of a murdered singer and, as Cavell argues in his compelling reading (A Pitch of Philosophy, 1994, p. 134–136, Contesting Tears, 1996, p. 47–78), around the question of whether Paula (Ingrid Bergman), who follows her aunt’s legacy by training as a singer, has a voice of her own. Her husband Gregory (Charles Boyer), who has murdered the great singer, is after a rather more tangible legacy, namely her jewels, which he assumes are still in the house inherited by his young wife. He systematically drives Paula out of her mind and into silence so as to deprive her both of her property and her voice. Yet the heroine finds her voice when in the climactic denouement she confronts her husband with a knife. Her speech or ‘aria’, as Cavell calls it, picks up her aunt’s signature role as Lucia di Lammermoor. Rather than suffering the fate of the operatic character Lucia (her unhappy marriage, her madness, the stabbing of her husband), Paula rescripts the libretto in a happier key. She takes on the legacy of her aunt and simultaneously speaks for herself having found her ‘voice’. My reflections on the feminine voice in this text also pick up on Elisabeth Bronfen’s discussion of Stanley Cavell in her text Silencing Voices (2009), notably her suggestion that the transformation of the woman into a narcissistic mirror image is tantamount to an obliteration of her ‘voice’ or even of her existence.
For discussions of the figure of the female singer in nineteenth‑century literature (and beyond) see Susan J. Leonardi and Rebecca A. Pope (1996), Felicia Miller Frank (1995), Jörg Theilacker (1989) and Corina Caduff (2003). In addition to romantic texts that feature female singers such as Don Juan and Rat Krespel by E. T. A. Hoffmann, there is a host of nineteenth‑century novels about female performer figures – singers, speakers and actresses – who speak and/or sing in public, including for instance Germaine de Staël’s Corinne ou l’Italie (1807), George Sand’s Consuelo (1842), George Eliot’s Daniel Deronda (1876), Henry James’s The Bostonians (1886) and The Tragic Muse (1890) as well as Jules Verne’s Le Château des Carpathes (1892).
In her seminal study The Acoustic Mirror (1988), Kaja Silverman discusses the function of the female voice as an acoustic mirror from a psychoanalytic perspective.
Sybille Krämer, for instance, describes the voice as a threshold phenomenon and writes that with respect to binary organizations and classifications of concepts, the voice is “always both: body and spirit, nature and culture, affect and intellect” (Krämer 2006, p. 290; my translation).
In narrative fiction, the concrete voice only comes into play when texts are read out loud and thus brought into a performative setting.
A different substitution of voices in a musical context occurs in Rat Krespel by E. T. A. Hoffmann. Krespel (who happens to have a disagreeable voice) forbids his daughter Antonie to sing but plays the violin so that she can ‘sing’ through his play.
Blixen’s pen name ‘Isak Dinesen’ is, strictly speaking, a half‑pseudonym. Dinesen was her maiden name, while the assumed Hebrew name Isak means ‘the one who laughs’.
A reproduction of this photograph can be found, for example, in the richly illustrated biography The Life and Destiny of Isak Dinesen by Frans Lasson and Clara Svendsen (1976, p. 181).
Blixen had several nicknames such as Tanne and Tania, and she used a number of other pseudonyms, including Osceola and Lord Byron. See the standard biography Isak Dinesen: The Life of a Storyteller by Judith Thurman (1995, p. 5).
See Judith Thurman (1995, p. 140, 258, 331, 337) and Susan Brantly (2002, p. 66). Blixen felt that she had been cut off from life after realizing that she had contracted a severe form of syphilis from her husband and after she lost both her lover Denys Finch Hatton and her coffee farm in Africa. It was against this backdrop that she turned to a second life in her writing and, as part of her artistic practice, fashioned herself as a living artefact.
Female performer figures who fall silent in nineteenth‑century narrative fiction include, apart from Trilby, the genius improviser Corinne in Germaine de Staël’s Corinne ou l’Italie (1807), the speaker Zenobia in Nathaniel Hawthorne’s Blithedale Romance (1852) and the speaker Verena Tarrant in Henry James’s The Bostonians (1886). Corinne dies a heroic death, Zenobia commits suicide, while Verena Tarrant is forced from the public platform into marriage.
Like the photograph mentioned in the text, Du Maurier’s original illustrations for the novel also depict the figure and pose of the singer in a manner that is strongly reminiscent of the iconography of ancient sculptures. On the important role played by the Pygmalion myth and the sculptural metaphor on the Victorian stage in general see Gail Marshall (1998). The commercialization of La Svengali in the text was reduplicated by the reception of Du Maurier’s novel at the time. The Trilby mania which was triggered by the publication of the novel and its adaptation for the stage was accompanied by an early instance of merchandising, namely the sale of various consumer goods, including soap, ice cream and the so‑called Trilby hat, which was used in the first London production of the play.
According to Nina Auerbach, the transformative power of Trilby’s voice is to be read together with her characterization as a figure: “Her ability under hypnosis to ring endless variations upon familiar tunes is the power of her character to transform itself endlessly […]” (Auerbach 1981, p. 286).
A detailed analysis of the ways in which art feeds on the cultural conjunction of death and femininity is offered by Elisabeth Bronfen in Over Her Dead Body: Death, Femininity and the Aesthetic (1992). For related discussions also see Sigrid Weigel (1996) and Corina Caduff (2006).


Auerbach, Nina: Magi and Maidens: The Romance of the Victorian Freud. Critical Inquiry 8.2, 1981, p. 281–300.
Bakhtin, Mikhail: Discourse in the Novel. In: idem: The Dialogic Imagination: Four Essays. U of Texas P, Austin 1981, p. 259–422.
Brantly, Susan: Isak Dinesen: The Danish Scheherazade. Scandinavian Review 90.2, 2002, p. 58–66.
Bronfen, Elisabeth: Over Her Dead Body: Death, Femininity and the Aesthetic. Manchester UP, Manchester 1992.
Bronfen , Elisabeth: Silencing Voices. In: Beate Neumeier (ed.): Dichotonies: Gender and Music. Universitätsverlag Winter, Heidelberg 2009, p. 23‑28.
Caduff, Corina: Die Literarisierung von Musik und bildender Kunst um 1800. Fink, München 2003.
Cavell, Stanley: A Pitch of Philosophy: Autobiographical Exercises. Harvard UP, Cambridge MA – London 1994.
Cavell, Stanley: Contesting Tears: The Hollywood Melodrama of the Unknown Woman. U of Chicago P, Chicago – London 1996.
Derrida, Jacques: Of Grammatology. The Johns Hopkins UP, Baltimore – London 1997.
Dinesen, Isak: The Dreamers. In: idem: Seven Gothic Tales. Putnam: London, 1969, p. 327–430.
Dolar, Mladen: A Voice and Nothing More. MIT Press, Cambridge MA – London 2006.
Du Maurier, George: Trilby. Oxford UP, Oxford 1998.
Felman, Shoshana: Writing and Madness – From ‘Henry James: Madness and the Risks of Practice (Turning the Screw of Interpretation)’. In: Emily Sun – Eyal Peretz – Ulrich Baer (eds.): The Claims of Literature: A Shoshana Felman Reader. Fordham UP, New York 2007, p. 15–50.
Kramer, Sybille: Die ‘Rehabilitierung der Stimme’: Über die Oralität hinaus. In: Doris Kolesch – Sybille Krämer (eds): Stimme: Annaherung an ein Phanomen. Suhrkamp, Frankfurt am Main 2006, p. 269–295.
Lasson, Frans – Svendsen, Clara: The Life and Destiny of Isak Dinesen. U of Chicago P, Chicago – London 1976.
Leonardi, Susan J., – Pope, Rebecca A.: The Diva’s Mouth: Body, Voice, Prima Donna Politics. Rutgers UP, New Brunswick NJ 1996.
Marshall, Gail: Actresses on the Victorian Stage: Feminine Performance and the Galatea Myth. Cambridge UP, Cambridge 1998.
Mayer, Mathias – Neumann, Gerhard (eds.): Pygmalion: Die Geschichte des Mythos in der abendlandischen Kultur. Rombach, Freiburg im Breisgau 1997.
Miller Frank, Felicia: The Mechanical Song: Women, Voice and the Artificial in Nineteenth‑Century French Narrative. Stanford UP, Stanford 1995.
Ovid: Metamorphoses. Norton, New York – London 2005.
Shaw, Bernard: Pygmalion. Penguin, London 2003.
Silverman, Kaja: The Acoustic Mirror: The Female Voice in Psychoanalysis and Cinema. Indiana UP, Bloomington – Indianapolis 1988.
Stoichita, Victor I.: The Pygmalion Effect: From Ovid to Hitchcock. U of Chicago P, Chicago – London 2008.
Theilacker, Jörg: Vorwort. In: Jörg Theilacker (ed.): Der hohe Ton der Sängerin: Musik‑Erzählungen des 19. Jahrhunderts. Luchterhand, Frankfurt am Main 1989, p. 7–22.
Thurman, Judith: Isak Dinesen: The Life of a Storyteller. Picador, New York 1995.
Weigel, Sigrid: Musen und Junggesellenmaschinen – Mythen vom Geschlecht der Künste: Vorwort. In: Corina Caduff – Sigrid Weigel (eds.): Das Geschlecht der Kunste. Böhlau, Köln – Weimar – Wien 1996, p. vii– xv.


What does it mean to have voice of one’s own? And how does narrative fiction, as a medium that works without any concrete sound, evoke “voice” as an aesthetic special effect? Taking my cue from George Bernard Shaw and his play Pygmalion (1913), which refers us to the notion that in a gender context, having a voice is not coterminous with having a voice of one’s own, I compare and juxtapose the two female singer figures in George Du Maurier’s novel Trilby (1894) and Isak Dinesen’s story The Dreamers (1934). With her radical departure from Du Maurier’s construction of Trilby’s voice, which is ventriloquized by Svengali, a male artist who sings through her, Dinesen foregrounds the subjectivity of the modern woman artist. In a tragic accident, Pellegrina Leoni loses her voice as a professional singer but instead gains a “voice” of her own as she recreates herself, both as an artist and as her own work of art. By picking up on Mikhail Bakhtin, Shoshana Felman and others, I show in my reading of these two texts about singers how the “voice effect” of narrative fiction invites us as readers and critics to become aware of the tone or “voice” of a text as we trace the dialogue and/or the dissonance between various textual voices.