2

Mélisande and the silence of music

Framing nothing

At the beginning of Act 3 of Debussy’s Pelléas et Mélisande, we encounter Mélisande alone for the first time. Seated at the window of her room in a tower of the castle, she sings to herself as she combs her hair for the night. Her song is unaccompanied, quite naked of any orchestral dress, but it is framed by a brief orchestral introduction, a musical outline of nothingness, the faintest of pencil lines on a white background. A single repeated pitch, flickering gently across three octaves in the harp, lightly ripples the scrim of viola and cello harmonics, a mere breath of sound, ghosting the single sustained note of the flute. To this is added the thin, muted line of the violins, floated in like the distant movement of clouds in the night sky, allowing a sudden moment of moonlight, marked by the punctuation of a B major chord, a momentary thickening of the texture, and the repeated three-note figure given by two flutes – a gentle premonition of the stirring of the doves who will take flight later in the scene (see Example 2.1). Opera composers have often found ways to conjure the presence of the heroine before her actual appearance on the stage. With these seventeen bars Debussy does something similar yet evokes an empty space. The sonic ripple that runs through the orchestra frames almost nothing. By means of its veil of sound, silence appears.

Example 2.1 Claude Debussy, Pelléas et Mélisande Act 3.i, bb. 1–17

Debussy’s only completed opera has been much discussed. It has in particular been much discussed for its symbols, as if the opera’s relation to Symbolism meant that it should be full of symbols, presenting a series of musical devices that stand for ideas of other (non-musical) things. I take a different view. My approach has nothing to do with ideas of expression or communication, the carrying of messages or the ‘freighting’ of meaning. Debussy’s Pelléas et Mélisande does not, I suggest, convey either ‘the interpretative message of the composer’ or ‘the psycho-dramatic meaning of the original play by Maeterlinck’.¹ Neither do I hear this music as the ‘encoder of the Unconscious’, whether that of the composer, the dramatist or the collective audience. Debussy’s opera does not translate or transpose some content previously expressed elsewhere: it writes and thus makes present. It foregrounds the primary activity of all aesthesis – the act of art’s appearing, of making itself present, of taking place.

Of all the aspects of Debussy’s opera, none has been more discussed than the nature of Mélisande herself. Jankélévitch sums up a prevalent understanding of her in his simple remark that ‘Mélisande is almost inexistent’. She is, he says, ‘a breath, a feather, a disappearing apparition’.² The association of Mélisande with silence, transparency, childlike innocence, luminosity, and weightlessness saturate accounts of Debussy’s opera. The quality of her voice and the listening it demands seems to lie at the heart of the opera’s concerns. Both Jann Pasler and Katharine Bergeron³ have underlined the definitive idea of a vocal delivery ‘presque sans voix’ by which naturalness becomes associated with a kind of artless truthfulness. Thus, Mélisande’s plain and unaccompanied singing, at the start of Act 3, Pasler suggests, allows us ‘to look for the truth not in clear expressive melodies, but in the emptiness and ambiguities of sound – Debussy’s “déclamation épurée” – as well as silence’.⁴ Carolyn Abbate suggests that, in the opera, it is only Pelléas that truly hears this quality of Mélisande’s voice for what it is, because he listens to her voice not the words: ‘Pelléas, not noticing what Mélisande says instead hears music: a grain that for him drowns out the literal meaning of her words.’ Abbate concludes that ‘what Pelléas hears in Mélisande’s voice is analogous to what Symbolist poets wished to hear in poetic language: sonority, not philosophy’.⁵ It is Pelléas alone who has this ‘capacity to hear sounds that are inaudible to everyone else’ but after his death, Abbate suggests, this ability passes to the audience who, in Act 5 of the opera, finally hear Mélisande’s voice merge with the ‘phantom’ voice of the orchestra.

It is salutary to remember that Debussy’s cultivation of a musical art ‘with almost nothing to say’⁶ was exactly contemporary with the high point of programme music in the tone poems of Richard Strauss and with the grand symphonic narratives of Gustav Mahler. The music of Debussy, by contrast, is often impersonal to the point of anonymity – brief, microscopic, non-discursive, focused on the elaboration of particularity, disinterested in narratives of the subject. As Schoenberg and his pupils explored a highly-wrought Expressionism, breaking musical language apart in the urgent impossibility of trying to say, Debussy’s quiet revolution took a different path. This arresting contrast marks the overlap, and short-lived simultaneity, of two quite different ideas of music. Strauss, Mahler, and Schoenberg stand at the end of a musical tradition whose paradigm was essentially that of nineteenth-century literature; Debussy stands at the beginning of one whose paradigm is closer to that of twentieth-century painting.

It was in relation to the composition of Pelléas et Mélisande that Debussy wrote, in a famous letter to Chausson (in September, 1893), about his fascination with silence as a means of expression, a discovery that helped him to distance himself definitively from Wagner. We might understand this, retrospectively, as part of a wider aesthetics of reticence in French music, a quality which, as Carlo Caballero has shown, is related to notions of both sincerity and interiority.⁷ In its refusal of messages, signs, and symbols, Debussy’s music finds common ground with both Fauré and Ravel; but also, in its distance from the idea of musical discourse, it shows a certain kinship with quite different kinds of later French music – with the neoclassicism of Les Six, with the organization of sound in Varèse, and with the constructivism of Boulez. In their different ways, all constitute a refusal of an idea of musical speech that had become vulgarized by Debussy’s time. It is interesting to hear the same aesthetic point of view expressed by a composer like Tristan Murail a century later. Talking about his orchestral piece La dynamique des fluides (1991), Murail reports that the work originated in ‘a desire for lightness, suppleness, polyvalency, a concern to avoid as much as possible grandiloquence and rhetorical effects’.⁸

Debussy’s comment about opera, after his second trip to Bayreuth in 1889, is well-known: ‘I am not tempted to imitate what I admire in Wagner: I visualize a quite different dramatic form. In it, music begins at the point where the word becomes powerless as an expressive force: music is made for the inexpressible.’⁹ He went on to outline what he looked for in a librettist in ways that come remarkably close to the Maeterlinck play he was to encounter a few years later: ‘A poet who half speaks things. Two related dreams: that’s the ideal. No country, no date, [ . . . ] scenes with different locations and of different types; characters who do not discuss, submitting to life and destiny.’¹⁰ In the same year, Debussy attended the Exposition Universelle in Paris. His encounter there with non-western music, including a Javanese gamelan orchestra, is well-known, but he seems to have been equally struck by the theatrical elements of what he saw. He was amazed, for example, at the simplicity of means in Vietnamese theatre: ‘There are simply more gods and less scenery. A bad-tempered little clarinet directs the emotion. A tam-tam organises the terror . . . and that’s all. No special theatre, no hidden orchestra. Nothing but an instinctive need of art, ingeniously satisfied.’¹¹

But, of course, Debussy’s opera is not nothing: it is a five-act drama lasting two and half hours and containing some of the composer’s most engaging and intense music. And for all that he sought to capture Mélisande’s enigmatic silence, she appears in ten of the opera’s fifteen scenes and must sing and move and fall in love. So it is the paradox of saying nothing that I explore here – not the idea of nothing itself but a palpably audible and articulate musical saying, a sonorous voicing of nothing. That quality was not sufficiently evident for Richard Strauss who protested, after hearing Pelléas et Mélisande in 1907, ‘There is not enough music [ . . . ] this is nothing, nothing at all’.¹² Adorno would later remark similarly, about Debussy’s music in general, that ‘everything seems to be a prelude [ . . . ] the overture to musical fulfilment [ . . . ] which never arrives’.¹³ This quality, definitive of much of Debussy’s work, was often a target for contemporary critics. Edmond Stoullig wrote, after a concert in March 1913 which included some of the second book of Préludes, that the first piece (‘Bruyères’), ‘contains nothing at all; the second, “Feuilles mortes”, not much more [ . . . ]’.¹⁴ Debussy may not have welcomed the implication of such remarks, but we might understand them as the flipside of what he quite deliberately cultivated within his music, summed up in his comment about trying to portray Mélisande musically: ‘I have spent days in pursuit of the “nothing” she is made of’.¹⁵

Understanding this ‘nothing’ as something of high value in Debussy’s music, rather than in the pejorative way Edmond Stoullig doubtless intended, has shaped a certain reception of Debussy’s music ever since the premiere of Pelléas et Mélisande. It has often been noted that, in Maurice Maeterlinck’s play, Debussy found his perfect libretto in a theatre of non-event, one whose drama emerges from the silences between characters and from the gaps within its language. ‘It is idle to think that, by means of words, any real communication can ever pass from one man to another’, Maeterlinck wrote in ‘Le Silence’: ‘from the moment that we have something to say to each other, we are compelled to hold our peace’.¹⁶ This might seem like an odd assertion for a writer, but it is not a call to cease writing or speaking. On the contrary, it is part of Maeterlinck’s aesthetic that only through the framing activity of what is said, only by means of the words that are used to demarcate a space within language, does such a silence become audible; in his own words, ‘the words that we let fall have no meaning apart from the silence that wraps them round’.¹⁷ Characters speak in Maeterlinck’s plays, just as they sing in Debussy’s opera, but, as Patrick McGuinness puts it, ‘dialogue is valid predominantly in so far as it advertises the unspoken or unspeakable, alluding to the ever present, incoherent, unformulated, voiceless world beneath it’.¹⁸ Such a use of language, he continues, ‘implies both the absent (what language leaves out) and the omnipresent (what language is none the less always about) [ . . . ] The unspoken governs, shapes, and confers meaning on the language that tries to exclude it’.¹⁹ Or, in the words of Jacques Derrida (who had plenty to say), the proper object of the critic is ‘nothing’, even though ‘nothing is not an object’, but rather ‘the way in which this nothing itself is determined by disappearing’.²⁰

The bare simplicity of Maeterlinck’s prose draws attention to the edges of language. By fracturing the surface of language he shapes the silences between the words – quite literally in his use of various kinds of punctuation (such as the dash and the ellipsis), but also through non-sequitur, repetition, and questions that seem neither to expect nor receive answers. Just as a composer writes a rest (un silence in French) to define a gap between musical events, Maeterlinck often inserts a direction for un silence between lines of his scripts.²¹ The result, as McGuinness comments, is ‘to leave the reader/spectator with an overwhelming sense of the inadequacy and provisionality of language’.²²

For Maeterlinck, silence, like speech, has its own expressive register, just as it has its own graphic code: a dash may denote interruption, rupture, or shock, while three dots may signify the petering out of discursive energy, the phrase dwindling into nothingness, an unfinished sentence, or the point at which dialogue ebbs away. It may be a pregnant pause or semantic discontinuity, and, as it is used by Maeterlinck, it is far from the ‘opposite’ of speech, if by this, we mean to imply that speech expresses while silence does not.²³

Maeterlinck’s text is characterised by constant repetition – of both single words and whole phrases. This too has the effect of undermining the signifying function of language; the semantic and syntactical logic of clear statement is broken up by the (em)phatic rhythm of the body. The abstract adequacy of language is displaced by a supplement of bodily repetition. Debussy made relatively few alterations to Maeterlinck’s text, but he often increased this phatic quality by adding an extra repetition of a phrase – such as Mélisande’s ‘ne me touchez pas’ in Act 1.i, or the threefold repetition of ‘je me suis enfuie, enfuie, enfuie’. In the tower scene, Act 3.i, Debussy adds a third ‘donne’ to increase the sense of Pelléas’s urgent and erotic ‘donne, donne, donne’ – a device carried over to the love scene in Act 4.iv.

As McGuinesss comments,

Maeterlinck’s introduction of silence, ‘l’Inconnu’ and the ‘néant’ into theatre parallels Mallarmé’s discovery of the potential of the blank page and the spaces that precede, surround, and interrupt the printed poem. Like silence in the theatre, the white page for Mallarmé is both an abyss of nothingness (a ‘gouffre’, a ‘vide’, and ‘abîme’) and a source of plenitude and multiple suggestion.²⁴

The abyss underlies Maeterlinck’s play in the form of the dark underground vaults of the castle, above which the precarious drama of its fragile characters takes place. It has often been noted that the dramatic space is structured on a vertical axis that runs from the height of Mélisande’s tower to the unknown depths of the grottos, and the risk of falling pervades a number of its key scenes.²⁵ The final scene of Act 1 takes place on a cliff-top looking out towards the open sea, the precipitous margin between the closed world of Allemonde and the infinite space of the ocean. But, as in Mallarmé, the ‘gouffre’ is first and foremost the terrifying space beyond language. It is the space to which Golaud, a man of rational relations between words and things, brings Pelléas to terrify him by the prospect of falling into a world without language. The same sense of a gaping abyss, and the asemantic world of noise that lies beyond language, haunts Debussy’s later operatic project based on Edgar Allan Poe’s The Fall of the House of Usher.

Debussy often opens up similar gaps in the musical surface, moments of silence framed by the music either side of it, one half of musical language which we have barely begun to explore.²⁶ His opera emerges, however, not from any simple parallelism of literary and musical devices but precisely from the tension between the two; less important than literal silences are the ways in which Debussy uses musical tone to make linguistic silences speak. In Act 2.ii, Mélisande tells Golaud that she is not happy. Golaud presses her for the reason with increasing agitation – ‘has someone been unkind? Is it the King? Is it my mother? Is it Pelléas?’. Mélisande answers at first with a rapid denial (‘No, no, it’s not Pelléas, it’s no one, you wouldn’t understand . . .’), the expressive tension in her voice accentuated by a high solo viola. And then, more calmly, she continues on a declaimed monotone, ‘It is something stronger than me . . .’ (C’est quelque chose qui est plus fort que moi). The final word is timed with the re-entry of the orchestra. The ellipsis at the end of the sentence, marked as such in Maeterlinck’s text, is not followed by a bar of silence, but a bar of music sounded from the mute and invisible orchestra. The harmonic slippage across these two bars might be explained as a simple series of dominant-ninth chords under the repeated D♮ of Mélisande’s vocal line, but more than that it is a slippage of affective register, a slippage of voice, from addressing Golaud outwardly to a turn inwards (see Example 2.2). It is above all the tone of the strings’ ‘voice’, the upper line doubled in the flute (très doux et très expressif) that makes audible what was framed by the silence of Maeterlinck’s ellipsis.

Example 2.2 Claude Debussy, Pelléas et Mélisande, Act 2.ii, bb. 70–77

One of the most explicit and concentrated examples of Debussy’s foregrounding of this gap is heard in the closing bars of Act 1. Descending the steep path from the cliffs, Pelléas offers to take Mélisande’s hand. Her reply signals neither agreement nor refusal: ‘Look, you see my hands are full of flowers’ (Voyez, voyez j’ai les mains pleines de fleurs), so Pelléas takes hold of her arm instead before, a few steps later, declaring ‘I may be leaving tomorrow’ (Je pars peut-être demain). ‘Oh’, Mélisande replies, ‘why do you have to leave?’ (pourquoi partez vous?). And with the very last word, the simplicity of the question is utterly changed by means of a sidestep in the harmony and a new orchestral tone (see Example 2.3). As the words imply future absence the music whispers a fragile promise of presence. Rarely does music present such an exposed edge between two worlds; a single barline here marks the gap between the absence of saying and the presence of music. Mélisande’s line should end with her G♮ falling to the tonic F♯, but with its twist up a semitone to G♯ instead, her question opens to a wholly new space. Coupled with the added D♯ in the chord, the resulting pentatonic harmony allows the unanswered question, and all the open potential it implies, to resonate well beyond the final bar (the chord will not resolve nor does it need to). It is one of the clearest examples to be found of music’s capacity to make resonate something unsaid. The closing bars of the orchestral music are a study in fragile presence suspended in absence. The music speaks infinitely what the words only hint at, neither touching nor parting. As the final bars fade, the flute takes up Mélisande’s question and turns it, wordlessly, twice more in a manner that has no end nor answer. The precipitous drop of the cliffs into the sea is as nothing compared to the gap that has been opened up within language.

Example 2.3 Claude Debussy, Pelléas et Mélisande, Act 1.iii, bb. 88–96

In Act 3.i, the tower scene, Mélisande’s words outline an internal drama of anxiety and desire, of wanting to let go but fearing to do so, a psychological study achieved with great economy by Maeterlinck. Encouraged to lean out of the window by Pelléas, her long hair falls down towards him, a dramatic moment marked by a preliminary climax, complete with a falling figure in the orchestra. Pelléas begins his lyrical hymn to her hair, ‘I hold it in my hands, I hold it in my mouth’ (Je les tiens dans les mains, je les tiens dans la bouche), in a state of quiet rapture, while Mélisande attempts to resist:²⁷ ‘Let me go . . . you’ll make me fall’ (Laisse moi! . . . tu vas me faire tomber). In Debussy’s setting, the lassitude of her ‘laisse moi’ already suggests that the sense of the words is quite distant to the sense of the body. The musical phrase, and the delivery it invites, ensure we listen to the voice not the words, or rather, that we attend to the gap that opens between them. Once again, it is with the sideways move of the final syllable that the gap is opened to reveal an utterly different musical space. As the descending vocal line ends on a C♯ (instead of the expected C♮) over the beginning of a gently rocking figure in the lower strings, a fragile melodic line emerges, first in the clarinet and then the oboe (see Example 2.4). It is Pelléas who sings, but it is Mélisande we hear.

Example 2.4 Claude Debussy, Pelléas et Mélisande, Act 3.i, ‘Je les tiens dans les mains’, bb. 107–17

The tower scene presents a gap in order to make it resonate – the gap between two lovers, between the absences of language and the presences of music, the gap that sparks between differentiated identity and release into the other, the vertiginous difference of desire, of body and language, music and words. After Mélisande’s hair tumbles down to him, Pelléas becomes rhapsodic.²⁸ The deliberately plain language of statement, question, and answer that makes up most of Maeterlinck’s drama is released into a musicalized language characterized by an unpunctuated flow of rich imagery; saying, in terms of declarative statements, gives way to poetic rhapsody. With this, the music comes into its own, because rhapsody is music’s default mode of saying. Or, to put it another way, at this point Debussy temporarily closes the gap between the restrained declamation of Materlinck’s truncated lines and the rich flowering of the orchestral voice. ‘My lips can’t reach your hand!’ (Mes lèvres ne peuvent pas atteindre ta main!), complains Pelléas: the mouth that speaks is divided from the hand that touches. And in urging her to give him her hand (Donne, donne, donne!), Pelléas anticipates the climax of the later love scene in Act 4.iv when it is her mouth that he demands (Ta bouche! Ta bouche!). Her impassioned response in the later scene (Oui! . . . oui! . . . oui! . . .) makes an inverted echo of her reply in the earlier one (‘Non! Non! Non!’). With the only moment of unison singing in the entire opera, the reciprocity of giving and receiving is marked by his ‘donne! donne!’ and her ‘toute! toute!’ sounding together on the very brink of Pelléas’s death at the hands of Golaud. At the dramatic and musical climax of the work, Debussy thus frames the bodily urgency of language; as words collapse back to the gestural violence of repetition delivered at full voice, the bodily force of music breaks through as nowhere else in the opera.

This breaking out is, of course, all the more powerful for the earlier restraint. Throughout Act 4.iv Pelléas has a passionately released music that almost trips over itself in its desire to be heard. Memory is insufficient (‘like carrying water in a napkin of muslin’); only the immediacy of an embodied presence/present will suffice. Saying thus takes on a new function in the epiphanic moment – here, one must say everything rather than nothing (il faut que je lui dise tout ce que je n’ai pas dit). This impossible desire to tell everything generates the performative act of the declaration of love which, famously, is almost lost in the saying. The everything he wishes to say is, at the climactic moment of his speech, close to nothing, merged with the resonance of the orchestral rush that proceeds it, a mere echo in the silence that follows. But the gap is a hinge, a threshold, an axis between two worlds. Three bars of unaccompanied vocal exchange separate the long-range build up to this moment and the new space that follows, three bars in which not only do the lovers speak their love for the first time, almost in silence, but immediately reflect on hearing and listening: ‘Oh, what did you say Mélisande! . . . I hardly heard it! . . .’ (Oh, qu’as tu dit, Mélisande! . . . Je ne l’ai presque pas entendu! . . .). Debussy is quite explicit in the way he inverts the usual relation between silence and saying. At the climax of Pelléas’s rush of words, his declaration (Je t’aime) signals a breakthrough into a musical resonance on the other side of the words. Maeterlinck’s image at this point is the breaking of ice with red-hot irons. The magical single sonority heard here (see Example 2.5) – the icy mirror of the high violin harmonic ‘E’ above, with the gently rocking chords in lower strings below – is a production of presence, a statement of immanent appearance, bringing the listener to an edge in order to gesture to a new fullness beyond it.

Example 2.5 Claude Debussy, Pelléas et Mélisande, Act 4.iv, ‘Tu ne sais pas pourquoi’, bb. 86–99

This is a threshold moment of the entire opera: it is not Mélisande’s words, but the transformed tone of her voice that Pelléas finally hears. In crossing this threshold, he hears the presence conferred by the voice rather than the absence which is the content of the words: ‘Tu dis cela d’une voix qui vient du bout du monde!’. As the tone of the voice becomes audible over the words it carries, the effect upon him is phatic rather than semantic: the touch of her voice is ‘fresher and purer than water . . . like pure water on my lips . . . on my hands’. With its definitive shift to F♯ major the music achieves a new space and sense of self-completion. Pelléas continues in his rhapsodic mode: ‘One would say that your voice had passed over the sea in spring! I had never heard it until now’ (On dirait que ta voix a passé sur la mer au printemps! Je ne l’ai jamais entendue jusqu’ici).

Hearing her (musical) voice for the first time is also the moment he accesses a new (musical) world. His rhapsodic lyricism, now at one with the orchestra, delivers a presence only occasionally hinted at in the rest of the opera. By framing what it is he now hears, Pelléas enables us to hear it too. After Pelléas dies, as Carolyn Abbate puts it, ‘his auditory peculiarity – a capacity to hear sounds inaudible to everyone else – survives his death. It is transferred, through music, to the audience.’²⁹ Which is why, for Abbate, Act 5 has a quite different sound quality. Just as, at Mélisande’s death, Arkel comments that ‘she has gone without a word’, so the music of Act 5 ‘says’ less and less, gradually shedding all its harmonic and motivic rhetoric. Even the tone becomes progressively thinner as presence dissolves into silence. Debussy marks the strings ‘dans une sonorité douce et voilée’, as the solo woodwinds withdraw to a childlike simplicity in the luminous nothingness of C♯ major with which the opera closes.

Golaud, of course, cannot bear the idea of nothing, hence his anger in Act 4.ii. When Arkel says, of Mélisande’s eyes, that he can see nothing in them but a great innocence, Golaud replies: ‘Listen – I am so close to them that I can feel a little breeze when the eye-lashes blink, and yet I am nearer to the great secrets of the next world than I am to the smallest secrets of those eyes!’. He cannot tolerate that the materiality of the world does not speak, that it stands outside language, that it says nothing and tells him nothing. Even in Act 5, as Mélisande lies on her death bed, he is still pressing her to tell him the truth – he wants discourse, words, the assurances of the linguistic world; he cannot accept that the world, in its fullness, is not contained by linguistic statement, and becomes increasingly angry as language ties him up in impossible knots. The two worlds, the two ways of being, are caught in the contrast of tone: the urgent tragedy of Golaud’s questions and the flat tone of the orchestra for Mélisande’s empty answers. For Golaud, being mortal brings with it the urgency of needing to say, to tell, in the face of death; for Mélisande, the timelessness of death brings with it only innocent silence.

Debussy’s opera is not a silent opera, but a work that uses the gap between words and music to frame a silence and to make it resonate. We may not be surprised that such a different musical personality as Richard Strauss concluded in bewilderment that the opera was ‘nothing at all’, but the coincidence of such antithetical aesthetic positions underlines the density of this historical moment. Debussy might have taken comfort in the fact that, nearly a century earlier, the paintings of his beloved Turner elicited similar responses; William Hazlitt famously quipped of Turner’s Snow Storm (1812) that the artist’s works were ‘portraits of nothing, and very like’.³⁰ The idea of framing a silence or an empty space, which pervades Symbolist literature, is made palpable in Rodin’s La cathédrale, whose two hands are the means by which something is made present that is not there (in an everyday or material sense). Only in the most basic sense is Rodin’s sculpture a representation of two hands; what this work projects is a specific field of tension, a quality of relatedness, which has to do with the space between the hands as much as the hands themselves. The ‘nothing’ between the hands, the empty space, is of course everything here. It makes appear, brings to presence, a relationship, that is – if one insists on the term – the content of the work. Kant, and a whole tradition of linguistically-constrained philosophy of art following him, wants to insist that this content is an idea, by which he means a concept, a word – say, ‘love’, or ‘tenderness’, ‘intimacy’, or ‘care’ – but this is a closure of the material which the work neither invites nor requires. Standing in the presence of Rodin’s hands, our experience is first and foremost ‘sensible’; I understand the work because I know what it is to touch and to be touched, because I know the affective language of the hands. What begins as a visual experience (I see the sculpture) is also at once tactile and affective, it engages the emotional intelligence of the body. Music ‘after Debussy’ does something similar.

Orchestral voices

Debussy’s opera unfolds in a highly distinctive way. Each scene is followed by an orchestral interlude that simultaneously prepares the next, passages of wordless music whose expressive fullness of tone contrasts strongly with the veiled quality of the scenes themselves in which characters deliver their lines in simple, declamatory statements, closer to speech than conventional singing. The result is an operatic paradox that the moments of emotional overflow, of most articulate expression, are largely to be found in the orchestral passages in which the singers are absent. This binary division is built into Maeterlinck’s drama, expressed in Arkel’s line that ‘We only ever see the reverse side of other people’s destinies, even the reverse of our own’ (Act 1). The idea proposed in a whole series of textual metaphors between inner and outer, visible and invisible, spoken and silent, is thus embedded in the structure of Debussy’s opera, alternating between the singing of on-stage characters and the wordless operation of the orchestra. If Debussy inherits from Wagner the idea of the orchestra as a kind of musical Unconscious to the words and actions of the characters on stage, he nevertheless departs from Wagner in thematizing the idea that words and music can never fuse – that they always work to opposite ends, that music is necessarily a supplement, a compensation for the losses and exclusions of language. The alternation in Pelléas et Mélisande of incommunicative speech and articulate orchestral music foregrounds this definitive gap. Although the interludes are often discussed in terms of a commentary on what has just taken place or what is about to happen, as a development of the ideas given in the text, the opposite is the case.³¹ The interludes accentuate the gap between what is spoken and what is felt, and underline the incommensurability of human actions and words with the vast seascapes of human life that cannot be spoken. One washes against the other, but the two remain separate – hence the aching sadness at the heart of both drama and music.

It is an astonishing thought, given their musical weight, that significant parts of these orchestral interludes were late additions as the opera neared its first performance in 1902. About a month before the opening, it became clear that the scene changes would take longer than expected and Debussy had to agree, ‘very reluctantly’ to extend some of them.³² Whatever their genesis, the nature of these passages, in their final form, is that they voice wordlessly something unsaid in the vocal scenes that come before and after them. But for all the presence of recurrent leitmotifs or figures amenable to semiotic coding, these interludes are not to be translated or decoded; there is no external action that equates to their interior drama. The interlude that joins Act 1.i to Act 1.ii is built on a march rhythm that might suggest the courtly world of Allemonde or Golaud’s royal provenance, but the twist of the harmony to the dark realm of F♭ minor and the rising trumpet theme (grave et expressif), just a few bars before the curtain rises on scene ii, tells a far more ambivalent story. Its sudden and portentous intensity leaves a stain on the blank silence of the letter scene that follows just a few bars later. In turn, the lyric emptiness of the letter scene is followed by a brief but passionate orchestral interlude that joins Act 1.ii to Act 1.iii. The heightened intensity here seems to arise simply through the insistent repetition of its dotted note motif and the focused tone of the strings, not through any musical discourse or argument (see Example 2.6). Momentary and short-lived, a mere twenty-six bars, one hears no statement but only a straining and thickening of the voice under the burden of unspoken feelings.

Example 2.6 Claude Debussy, Pelléas et Mélisande Act 1.ii, bb. 147–60

If the characters speak but communicate very little, by contrast, the orchestra seems to say everything by saying nothing. Where the characters seem not to hear one another and fail to communicate, the orchestral interludes are marked by a fullness of presence heard vocally only in the two love scenes. The orchestra gives voice not just to what characters have not said but also to thoughts and desires of which they are themselves not aware. In this sense, they voice an excess, something in excess of language, and part of the force of this most gentle of operas arises precisely from this appearance of music that is so palpably in excess of its words. It is striking that Proust, having listened to a performance of the opera on the théatrephone in 1911, wrote to Reynaldo Hahn that the parts he had enjoyed most were those without words.³³

Part of Debussy’s technique, in the orchestral music, is to allow the emergence of inner voices – a presentation of interiority that the flat, blank surface of the vocal dialogue seems to deny. In the passage quoted in Example 2.6, for example, from the interlude between Act 1.ii and Act 1.iii, it is the emergence of the cello in the middle of the texture that breaks out most forcefully at the climax. The same idea recurs in the interlude between Act 4.ii and Act 4.iii, where the orchestra has to deal with the emotional excess left by the end of a scene in which Golaud has thrown Mélisande to her knees and dragged her across the floor by her hair. As Alban Berg later found in Wozzeck, where the mute suffering of the onstage characters exceeds any words they have, it falls to the orchestra to voice a sense of care, concern, and pity that is left out of the telling of events. Debussy’s interlude here seems to expand on Arkel’s closing comment, ‘If I were God I would have pity on the hearts of men.’ For all its relative brevity, the orchestral passage achieves a searing intensity before giving way to a return of Melisande’s gentle (doux) theme in the oboe. In an opera which largely avoids counterpoint, this orchestral expansion opens up a dense weave of inner voices.

In a similar way to the emergence of inner voices within the orchestral texture, the orchestra more generally creates a kind of interiority to the otherwise ‘flat’ surface of the vocal dialogues, inhabiting the scenes as a wordless inner voice – not in the romantic sense of interiority, disclosing what a character feels but keeps secret, but in the sense of something that lies beyond the individuality of the characters. Act 1.ii begins in a singularly undramatic manner, with the blankness of Geneviève reading Golaud’s letter. And yet, the scene is poignant and deeply expressive; Mallarmé never encapsulated better the tension between music and letters than Debussy does in this scene. It is poignant partly because of the bare declamatory voice of Geneviève, but also because of the gap that divides it from the rich tone of the brief moments of orchestral commentary, a gap accentuated by Debussy’s avoidance of continuous orchestral underlay. Note, for example, the flowering of tone and melody, given by divided cellos and basses (très expressif) under the line in which the letter expresses Golaud’s hope that Arkel will accept and welcome Mélisande. Or again, a few bars later, where Golaud expresses his hope that a light will be lit in the tower as a sign that his ship will be welcome, now with the same melodic motif given in unison by two flutes over viola, divided cellos and basses (see Example 2.7).

Example 2.7 Claude Debussy, Pelléas et Mélisande Act 1.ii, bb. 27–46

We hear the words and what they tell, but we listen for what emerges from the orchestra. Debussy plays constantly with the disorientation of juxtaposing these two modalities. This is explicit as Geneviève reaches the end of the letter and turns to Arkel to ask ‘What do you say about it?’. The answer is given in the fulsome response of the orchestra (bb. 42–44), in slower tempo (in the full score, très expressif; in the vocal score, avec une grande expression): a rising melodic line in the cellos climbs towards a falling one in the violins, but the intensity of the phrase has as much to do with the tone of the string ensemble, its warmth contrasting with the flat coolness of the letter reading. Only after this brief but intense swell in the orchestral underlay does Arkel ‘speak’ for the first time, and that to say ‘I have nothing to say’ (Je n’en dis rien).³⁴

The gap is here presented in naked form – the inward nonverbal answer of the orchestra juxtaposed with Arkel’s outward ‘saying’ that he has nothing to say. In fact, in what follows he goes on to say more, but in order precisely to enlarge on the nature of this gap, insisting that we only see one side of the fate of others, an image centred on the notion of just such a dual aspect. This sense of the meeting of two sides of things is what defines the opera, not simply because it is written into Maeterlinck’s play but because it is played out in the constant oscillation between the edges of the words and the music. The contrast is all the greater here because of Debussy’s famously spare, declamatory text setting: a kind of degree zero of melodic writing. Language thus rises through the heightened speech of simple declamation and meets music coming in the opposite direction, a music divested of the conventional function of expressing any text.

Arkel is a key figure in this respect as the only member of the family who seems aware of this tension between words and silence. His most extended scene is Act 4.ii in which, alone with Mélisande, he embarks on what begins as a lengthy monologue on the darkness of the castle and his concern for Mélisande since her arrival. The unusually drawn-out prose hardly promises a fulsome lyrical aria, nor does it get one, except that the orchestra is richly eloquent and radiant, presenting in an expansive tone a sense of what the words alone do not say. Arkel’s meditation on old age and death versus youth’s affirmation of life might be prosaic or even banal at the level of the words alone, but the orchestra makes present what the words lament – a life force long absent within the walls of Allemonde. Debussy’s restraint in the word-setting makes space for the articulate presence of the orchestra. At first these are isolated moments – the lyrical addition of the violins giving Melisande’s theme (b. 123) or the intensity of the dense chordal scoring of the strings (bb. 131–33), but these are hints of a desired presence which the orchestra gradually makes present. Arkel’s statement that ‘from now on’ things will change is marked by the single D♭-ninth chord (b. 137), the beginning of a drawn-out process of enlargement (animez peu à peu) that delivers the old man’s vision which climaxes with a sense of arrival (b. 155) with his declaration ‘and it’s you, now, who’s going to open the door to a new future that I forsee’ (Et c’est toi, maintenant, qui vas ouvrir la porte à l’ère nouvelle que j’entrevois). What for Arkel is only a vision, the music momentarily makes real – witness how the rich sonority of the horns seem to recall the life-giving power of Freia’s golden apples in Wagner’s Das Rheingold (see Example 2.8).

Example 2.8 Claude Debussy, Pelléas et Mélisande Act 4.ii, bb. 152–58

The orchestral interludes, and the moments of orchestral commentary between and beneath the declamatory delivery of the text, might seem to embody Debussy’s oft-quoted words that he imagined an opera in which ‘music begins at the point where the word becomes powerless as an expressive force’. In this, of course, he was not as distant from Wagner as he imagined. It is not so much Debussy’s use of the expressive power of the orchestra that sets his own opera apart from that of other post-Wagnerian composers, but rather the gap he creates between what the voices do and what the orchestra does. The orchestra, which is linguistically mute and can ‘say’ nothing, appears to speak on behalf of the characters who speak but nevertheless say very little. The wordless instrumental music compensates for the inarticulacy and incommunicativeness of speech foregrounded in Maeterlinck’s text and underlined by the declamatory vocal style of much of the opera. It is perhaps not insignificant that while both Fauré and Sibelius were drawn to the drama, both wrote incidental music for use in performances of the play, and Schoenberg set out, in writing a tone poem, to explore the core of the drama without any words at all. Maeterlinck’s play needed a musical ‘setting’ no more than Mallarmé’s poem L’après-midi d’un faune needed a musical ‘prelude’, but allied with Debussy’s music both become quite different things. In both literary works, language reflects on the inadequacy of its own expression and points beyond itself, whether through the elliptical use of silence between words or in what is said by the musicality of speech as rhythmic sound rather than as abstract signifier. In Debussy, however, this straining of language to exceed its own limits meets, coming from the opposite direction, a music that foregrounds its own linguistic muteness, its refusal of saying, in favour of quiet appearing.

Being mute

Rodin’s La cathédrale had a companion piece, titled Le secret. Once again, two hands are on the brink of touching, but here they are separated by an object – unknown and indistinct, but nevertheless a hard and tangible thing. The difference between the two sculptures might help elucidate the difference Vladimir Jankélévitch attempts to define between a secret and a mystery at the beginning of his study of Debussy’s music.³⁵ Music, he insists, does not hide or disguise a meaning that might be otherwise disclosed – it contains no secret, ‘like a hieroglyph, essentially decipherable’ – but, rather, has to do with something essentially unsayable (indicible). The distinction he makes provides a way of approaching music’s articulate mutism: highly refined, clear and precise in all its details, it nevertheless resists being rendered into other terms. Like the space created by the coming together of the hands in La cathédrale, it is transparent but unnameable, expressive but unsayable. ‘The inexpressible, in Debussy, is a mystery in broad daylight’, Jankélévitch insists in the face of charges that this music is merely vague. After all, he insists, ‘what could be less labyrinthine than the naked, white simplicity of the Étude pour les cing doigts, the Tierces alternées, or Doctor Gradus ad Parnassum?’.

Rodin’s La cathédrale and Le secret both make palpable the idea of framing an empty space and, in that sense, ‘saying nothing’. Moreover, in the tangible, three-dimensional and hard presence of stone, Rodin reminds us that such an aesthetic does not necessarily lead to a kind of art opposed to its own materiality. On the contrary, it makes its own medium more vivid and particular, precisely because it draws attention to its edges; as in the poetry of Mallarmé, the drama of Maeterlinck, or the music of Debussy, it is not, paradoxically, the solidity of things that Rodin’s sculptures reveal, but their capacity to reach beyond themselves, to quietly insist that the ‘content’ of art is not its ostensible content (the object of its representation) but the material and manner of its presentation. It is an idea that finds emblematic illustration in a remarkable series of late still-life paintings from 1880 by Édouard Manet. In arrestingly simple works, such as ‘A Bunch of Asparagus’, ‘Still Life with Two Apples’, or ‘The Lemon’ (see Figure 2.1), Manet eschews any narrative of events or any historical, political, or social content that was so often the function of nineteenth-century public painting, and concentrated instead on the silence of mute objects with ‘nothing to say’. These are bare, stripped-back paintings, even by the standards of the genre, which foreground the nudity of the object over any idea of composition as such. But rendering these simple objects in all their material particularity, finding an arrestingly vivid character in their very everydayness, paradoxically throws attention back to the painting itself. As James Rubin puts it, the closer one looks ‘the more one discovers instances of pure visuality, passages of visual presence that are undetermined as anything other than indexes of the materiality of painting’.³⁶

Figure 2.1 Édouard Manet, Le citron (1880)

The French musicologist, Mathieu Guillot, in a wonderfully rich study of the idea of silence in music, says something similar about the effect of snow upon a landscape; it works, he suggests, as a ‘builder of silence’ (bâtisseuse de silence), making silence audible in a particularly intense fashion. It might recall Mallarmé’s ‘Sainte’, a poem about the silence of music.³⁷ The poem pictures, in a stained-glass window, a medieval vision of the patron saint of music with a viol, her finger poised to touch the string but suspended above it, like the way the hands of Rodin’s sculpture do not quite touch, and thus framing a silence by framing a space. Mallarmé’s closing line ‘Musicienne du silence’ stands as a metaphor for poetry in general – a musicalized language that frames a plenitude it does not attempt, like the noisy art of music, to sound out.³⁸

Both Mallarmé and Guillot point to ways in which framing silence detaches listening from its usual distraction by sound. As with Rodin, Manet, Maeterlinck, and Debussy, the art lies in this act of framing, the articulation of a frame of experience that allows a mute content to appear. As Jankélévitch has it:

Music is the silence of words, just as poetry is the silence of prose. Music, as sonorous presence, fills the silence, but yet music is itself a kind of silence [ . . . ] and likewise it is necessary to make music in order to coax out [obtenir] the silence.³⁹

Mallarmé found something similar in the mute art of ballet, an art formed by the coming together of two kinds of muteness, mime and dance, ‘two attitudes each possessive of its own particular silence’.⁴⁰ We might add a third, of course – the muteness of music. For all the flimsy narrative on which ballet is traditionally hung, and setting aside the few moments of explicit communication through mime, we do not read the dancer’s movements in terms of symbol or meaning but are engaged by them on their own terms. When we do so, Mallarmé suggests, we have achieved the ‘full daydream’ proper to art: ‘like the illiterate dancer enclosed in her circles or transported by flight, as she gives herself up to the games of her profession’.⁴¹ One must approach dance with ‘poetic instinct’ alone in order to grasp it; only then, Mallarmé insists, will the dancer truly reveal herself, ‘silently writing your vision in the form of a Sign, the sign that she herself is’ (silencieusement ecrira ta vision à la façon d’un Signe, qu’elle est).⁴²

Mallarmé’s word play, lost in translation, is a particularly resonant one. The ballet dancer’s prototypical role as a swan (cygne) is a transparent sign (signe); it does not designate or represent another thing (nobody goes to the ballet to watch dancers imitate large water birds). It is not just that the swan embodies the illusion of weightlessness and the elegant gliding motion the ballet dancer creates, nor that its immaculate white plumage is evoked by the dancer’s costume, but that both are mute. Mallarmé is insistent that the dancer’s mute presentation embodies ‘the sole principle’ of the theatre, for all that the latter it is an art of speech, movement, noise, drama. ‘[J]ust as the chandelier is supremely lustrous in itself, suddenly illuminating, in all its facets, anything at all – our diamond-clear-gaze – so a dramatic work shows the succession of the exterior aspects of the act without any of them containing reality – so that there happens, finally, nothing at all.’⁴³ Put another way, it is not the ostensible content of art that is of concern in art, but the kind of looking or listening to the world which it occasions. The pure art of ballet, ‘set free from the need to deal with characters, their clothes, their costumes, and their famous words’, is thus a kind of sensual diverting of the mind in order for it to focus on the central act of apprehension itself.⁴⁴ Ballet is thus ‘the ideal dance of constellations’ which takes us straight towards the state of non-individuality which is both ‘the abyss of art’ and its goal.⁴⁵ Writing of Messager’s ballet, Les deux pigeons,⁴⁶ Mallarmé comments: ‘after the artless prelude nothing takes place (except for the perfection of the performers) that’s worth even the briefest backward glance, nothing!’.⁴⁷

That ‘nothing takes place’ is the highest accolade from Mallarmé, as he might have reminded Debussy had he lived long enough to see the premiere of Pelléas et Mélisande. Or perhaps not, since in his review of the 1893 Paris premiere of Maeterlinck’s play, Mallarmé suggested that musical additions would be superfluous.⁴⁸ But it is certainly why Debussy was drawn to Maeterlinck’s play – not because he saw music as reinforcing the action or underlining the sense of the words, but rather because his idea of music was to articulate the resonant silence to which the words point. What joins all three is the idea of muteness not as the physical inability to speak, but as an active and voluntary act of not saying – an act of holding oneself silent. The French reflexive verb se taire captures this sense better than English, as a deliberate action, something one does rather than something one does not do. It is the act of the chorus of serving women who, in Act 5 of Pelléas et Mélisande, file into Mélisande’s chamber, unannounced and unsummoned, and simply stand silently around the edges of the room. At the moment of her death, instinctively and without a word, they kneel. Their presence affirms and bears witness to a silence. Se taire derives from the Latin tacere, whence comes tacet, an instruction musicians know well, denoting that here one keeps silent in order to enable listening to another. John Cage famously foregrounded the positive-negative quality of the instruction in 4’33”, a piece in which, for each of its three movements, the performer’s part is marked ‘tacet’ in order to enable listening.

The proper object of our enquiry, therefore, is not a literal silence but the audible means by which it is presented, not ‘nothing’ as a blank absence but rather its framing – an articulation of the relation between what is presented and what is not (as in Rodin’s hands), between the visible and the invisible, the hard objects and the mutually constituted space between them. As Jankélévitch insists, ‘it is necessary to make music to coax out the silence’. Sanford Budick and Wolfgang Iser, trying to get to grips with the notion of the unsaid or unsayable within language, characterize such an articulation as a ‘play of negativity’ that arises from within words themselves:

Wordlessness comes in two kinds. The first kind exists beyond language proper [ . . . ] The second kind exists by virtue of language itself. It is evoked by words, it is explored and elicited by words, it is even made possible by words. It is, in short, a post-verbal wordlessness.⁴⁹

The ‘play of negativity’ here might be compared to the effect of a photographic negative, the inverse of what appears in the printed image, an imprint of what is not made visible created by what is, but also the means by which the image appears. ‘[N]egativity’, Budick and Iser insist, ‘can only be described in terms of its operations, and not by any means in terms of a graspable entity’; it points to ‘a void within what is being said’.⁵⁰

Common to all accounts of this idea is the notion of a dynamic movement across a threshold, a movement characterised variously as play, oscillation, or flickering: ‘only through play can difference as oscillation be manifested, because only play brings out the absent otherness that lies on the reverse side of all positions drawn into interaction’.⁵¹ If there is a long tradition of thought and language that constitutes this otherness in terms of a ‘negativity’ (from theology to deconstruction), as Budick and Iser insist, such negativity ‘constantly lures absence into presence’. It does so not to displace one binary term with the other, to turn a negative into a positive, but precisely to put their relationship into dynamic play: ‘While continually subverting that presence, negativity, in fact, changes it into a carrier of absence of which we would not otherwise know anything’.⁵²

What allows the unsayable to speak is the undoing of the spoken through negativity. Since the spoken is doubled by what remains silent, undoing the spoken gives voice to the inherent silence which itself helps stabilize what the spoken is meant to mean. This voicing of the unsayable is necessarily multilingual, for there is no one language by which sayings of things can be undone.⁵³

This sounds not only like Mallarmé and Maeterlinck, but equally Jacques Derrida. In ‘How to Avoid Speaking: Denials’, Derrida answers the criticism that his works adds up to no more than speaking merely for the sake of speaking:

to speak for the sake of speaking, to experience what happens to speech through speech itself, in the trace of a sort of quasi-tautology, is not entirely to speak in vain and to say nothing. It is perhaps to experience a possibility of speech which the objector himself must presuppose at the moment when he addresses his criticism. To speak for nothing is not: not to speak. Above all, it is not to speak to no one.⁵⁴

Heidegger expressed a similar idea, insisting that speech and silence are mutually constitutive:

Keeping silent authentically is possible only in genuine discoursing. To be able to keep silent, Dasein must have something to say – that is, it must have at its disposal an authentic and rich disclosedness of itself. In that case one’s reticence [Verschwiegenheit] makes something manifest, and does away with ‘idle talk’ [‘Gerede’]. As a mode of discoursing, reticence articulates the intelligibility of Dasein in so primordial a manner that it gives rise to a potentiality-for-hearing which is genuine, and to a Being-with-one-another which is transparent.⁵⁵

Gerede, the language of (everyday) communication, is a quite different thing. We understand its language without having knowledge of the things spoken about because we understand only linguistic relationships not the things to which they are meant to refer: ‘the primary relationship-of-Being towards the entity talked about is not “imparted” by communication’,⁵⁶ as Heidegger has it, an idea that Maeterlinck had already encapsulated in Le silence.

One form of this negativity, of using the said to delimit the unsaid, is the concept of erasure that Derrida takes up from Heidegger. Erasure in philosophy necessarily functions by means of an essentially negative dialectic (not this, not here). But art has its own version: it is perhaps not possible to understand atonality, for example, except in this way, as a presentation of the absence of tonality. Atonality is thoroughly dialectical; its meaning is founded in being a kind of tonality. But unlike language, when art presents a kind of erasure (not here, not yet) it has to do so in material form, through colour, stone, sound, movement, speech, image, gesture. It has to use what is materially present in order to set up a resonance with what it is not. This, surely, is the critical resonance of all art – embodied and grounded, but pointing beyond itself to what is not heard or seen. Our focus is thus neither the content of art understood in the everyday sense (of representation or expression) nor an absent content (the invisible or unsaid), but rather the threshold between the two, the manner in which artworks construct a double movement, appearing and at the same time drawing attention to their own erasure through disappearing, in order to make something else appear.

This shift in focus displaces the idea of art as presenting some content and replaces it with the an idea of art as articulating a threshold; it displaces the idea of art as a kind of saying and replaces it with an idea of art as a kind of appearing. Such a shift, in both the making and experiencing of art, leaves behind any residual and ridiculous notion of concealed content or ‘hidden meaning’ in artworks, let alone the attribute of vagueness or ambiguity of meaning, all of which presume some (linguistic) content deformed or masked by the materiality of the artwork. The muteness of art is, rather, a play of negation that opens a space between the edges of signification and what the world might be when it is released from the insistence on signifying. Music and poetry open such a space to make resonant the gap between the sensuous and the rational aspects of language, the son (sound) and the sens (sense). Such a gap allows the non-identical and the non-determined to resonate and thus allows us, momentarily, to be similarly non-identical and non-determined. If talking about art sometimes borders on a quasi-religious way of speaking – of artworks possessing or occasioning a ‘redemptive’ quality – it is, perhaps, for this reason: that they sometimes afford a kind of experience we register as a recovery, a receiving back of something lost by our entry into language.

As Jean-Luc Nancy puts it: ‘Sense is consequently not the “signified” or the “message”: it is that something like the transmission of a “message” should be possible. It is the relation as such, and nothing else.’⁵⁷ Indeed, he goes further, arguing that the sense of art is marked by the absence of signification in a strictly linguistic sense. The content of art is rather ‘the nudity of existing’; art underlines, he insists, ‘that existence is (the surprise of) sense, without any other signification’.⁵⁸ Far from making art meaningless, in the everyday sense of being without value or sense, it is this essential aspect of art that makes it of the highest value, since: ‘To say this almost nothing is the sole task of a writing’ and thus ‘the end of philosophy.’⁵⁹ The pun is definitive: it is the goal of philosophy but also marks its completion. One might go further and invert Hegel’s progression of Geist by which art must give way to religion and religion to philosophy, in an ever-increasing process of abstraction from materiality. The mute presence of art seems to say the opposite, suggesting that, while we are still bodies, the highest form of our being in the world is found not in words but in a spiritual flesh.

Which brings us back to Debussy’s opera. Hegel’s demand that the embodied particularity of art should give way to the conceptual abstractions of philosophy is equally Golaud’s frustrated demand, that the muteness of Mélisande, which is also that of music, and of the momentary but epiphanic glimpses of the sea and sky, should be spoken, explained, and contained within the dark castle of his language. If this uniquely unoperatic opera, whose central character is without drama and often without voice, is uniquely sad, this is the reason. Golaud is no villain, though he becomes one. Because he cannot listen, all he hears is speechlessness. He cannot hear the music in Melisande’s voice. Even on their very first meeting, in Act1.i, when he finds her weeping and asks her why, he doesn’t hear the answer to his question because it is given not in words but in tone and gesture (a solo viola, oboe, and high violins). Melisande is often weeping, because weeping articulates the gap between speech and speechlessness. She is an ‘infinitely gentle, infinitely suffering thing’⁶⁰ because she is too musical for the linguistic world around her.