¹ La Musique et I'ineffable (Paris, 1983), 172, ‘S'il est vrai que la “loquela” … est le bruit humain par excellence, le mutisme qui supprime ce bruit sera un silence privilégié. La musique est le silence des paroles; tout comme la poésie est le silence de la prose. La musique, présence sonore, remplit le silence, et pourtant la musique est elle-même une manière de silence … Et de même il faut faire de la musique pour obtenir le silence’.Google Scholar

² Ibid., 93, ‘Et I'ineffable … est inexprimable parce qu'il y a sur lui infiniment, interminablement à dire’; 95, ‘La musique est done inexpressive non pas parce qu'elle n'exprime rien, mais parce qu'elle n'exprime pas tel ou tel paysage privilégié, tel ou tel décor à l'exclusion de tous les autres; la musique est inexpressive en ceci qu'elle implique d'innombrables possibilités d'interpretation, entre lesquelles elle nous laisse choisir.’

³ Stefan Jarocinski cites a maxim of Goethe: ‘the symbol transforms the phenomenon into an idea, and the idea into an image, and does this in such a way that the idea in the image has infinite repercussions, and remains intangible; even when expressed in every language it will always remain unexpressed’, as well as a similar sentiment dating a century and a half later, from Maurice Blanchot, ‘[the symbol] merely makes present - by bringing us into its presence - a reality which cannot be grasped in any other way, seeming to emerge suddenly, prodigiously far away, like some strange apparition’. See Debussy: Impressionism and Symbolism, trans. Myers, Rollo (London, 1976), 23, 24. Without rehearsing the immense bibliography in linguistics, psychology or philosophy on the nature of the symbol, one can none the less point out that twin notions of inexhaustible resonance and distance (entailing an interpretative quest towards an infinitely receding object) recur in most nineteenth- and twentieth-century definitions.Google Scholar

⁴ La Musique et I'ineffable, 129–30, especially his discussion of Andersen's fairy tale The Bell. Jankélévitch sees fictions about unlocalisable cities and legendary hidden sites as allegories about music itself, ‘Et de même que l'âme récuse les localisations cérébrales et Dieu les localisations terrestres, ainsi la céleste Kitiège, la Kitiège absent et omniprésente, lointaine et prochaine, qui est la pure musique en elle-même ne figure sur aucune carte … échappe à toute topographic’ One is reminded inevitably of Hölderlin's most famous aphorism, the opening line of ‘Patmos’, ‘Nah ist / und schwer zu fassen der Gott’.Google Scholar

⁵ I am grateful to Caryl Emerson for this translation from Chekhov's Russian text.Google Scholar

⁶ Operatic instances include Schreker's Der ferne Klang, as well as Ravel's unfinished operatic project La Cloche engloutie, which was based on Gerhard Hauptmann's Die versunkene Glocke.Google Scholar

⁷ Schaeffer, Pierre, Traité des objets musicaux, rev. edn (Paris, 1967), 90–9;Google Scholar on disembodiment as mark of omniscience or divinity see also Wolterstorff, Nicholas, Divine Discourse: Philosophical Reflections on the Way God Speaks (Cambridge, 1995), 56–8.CrossRefGoogle Scholar

⁸ Simon Morrison, writing on the Russian symbolist reception of Wagner, offers an elaborate meditation on the dilemma of realising music so freighted with poetic ambition that silence must result; see ‘Skryabin and the Impossible’, Journal of the American Musicological Society, (forthcoming, 1998).Google Scholar

⁹ This paradox is discussed in Abbate, ‘Ventriloquism’, in Meaning in the Visual Arts: Views from Outside, ed. Lavin, Irving (Princeton, 1995), 305–12.Google Scholar

¹⁰ In this respect the sound engineering for such apparitions in operatic recordings is telling: those from the classic era of LP stereo sound (such as Colin Davis's performance of Don Giovanni on Phillips) tend to distort such voices with acoustic devices that amplify without literally increasing volume, such as echo effects. This habit of enhancing decorporealised speech with synthesised resonance is ingrained in Hollywood cinema. The pervasiveness of the gesture is one reason why the disembodied voice of omniscient Hal in 2001 is so disorientating. Hal's voice is never booming or resonant, but wholly closed in and intimate, a whisperer's lips planted perilously close to our skulls.'Google Scholar

¹¹ Wayne Koestenbaum discusses this phenomenon of ‘singing with the singer’, as one based not primarily on narrative identification with character but on an illusion of bodily identity with the woman performing; see The Queen's Throat Opera, Homosexuality and the Mystery of Desire (New York, 1993), 42–3.Google Scholar

¹² Dieter Borchmeyer refers to this identification (of which, he notes, Wagner himself was well aware) by Holda, calling ‘das germanische Inkognito der Frau Venus’, see Das Theater Richard Wagners (Stuttgart, 1982), 203–6.Google Scholar

¹³ See , Abbate, ‘Mythische Stimmen, Sterbliche Korper’, in Richard Wagner: Ansichten des Mythos, ed. Bermbach, Udo and Borchmeyer, Dieter (Stuttgart, 1995), 75–86; and Orpheus and the Trumpet: Essays on Opera (forthcoming), Chapter 4.Google Scholar

¹⁴ Chion, , La Voix au cinema (Paris, 1972), 25–33.Google Scholar

¹⁵ Ibid. I, 26–7.

¹⁶ La Voix au cinema, 27–8, ‘on est done loin de la voix de coulisse théâtrale, perçue réellement en retrait par rapport à une scéne qui, elle, contrairement au champ cinématographique, ne vous fait pas sauter d'un point de we à un autre … l'acousmêtre du cinéma, lui, est à la fois “hors-champ”, done, pour le spectateur, en-dehors de l'image, mais en même temps il est dans cette image de derrière laquelle il provient, réellement (cinéma classique) ou imaginairement.’Google Scholar

¹⁷ Audio Vision: Sound on Screen, ed. and trans. Gorbman, Claudia (New York, 1994), 129–30. Audio- Vision is a brief English compendium of previous and newly revised writings by Chion; in it he refreshes various passages on the acousmêtre from La Voix au cinéma; see ‘The Acousmatic’, 71–3; ‘The Invisible Man’, 126–8, and ‘The Acousmetre’, 129–31.Google Scholar

¹⁸ See for instance Zizek's, Slavoj chapter ‘The Ideological Sinthome’ in Looking Awry: An Introduction to Jacques Lacan through Popular Culture (Cambridge, Mass., 1991), 124–8, esp. 126, in which Zizek misreads Chion's acousmatique as stricdy limited to an uncanny typus and excluding commentary voice-over, ‘the voice without bearer, which cannot be attributed to any subject and thus hovers in some indefinite interspace. This voice is implacable precisely because it cannot be properly placed, being part neither of the diegetic “reality” nor of the sound accompaniment (commentary, musical score), but belonging, rather, to that mysterious domain designated by Lacan as “between two deaths” … as was demonstrated by Chion in his brilliant analysis, the central problem of Psycho is to be located on a formal level; it concerns the relation of a certain voice (the “mother's voice”) to the body for which it searches.’Google Scholar

¹⁹ Looking Awry, 127–8. On the mediumistic, spiritualist and morbid origins of the telephone as invention (many of which seem to have emerged in telephone movies with disembodied voices), see the associations documented in Ronell, Avital, The Telephone Book (Lincoln, Nebr., 1989), 192–6.Google Scholar

²⁰ Carol Clover has argued that males with disturbed sexual identities who are sexually ambiguous and (occasionally) possess androgynous voices regularly turn up as villain slashers in horror films, where they are pitted against heroic and masculinised female heroines; see Clover, , Men, Women, and Chain Saws (Princeton, 1992), 48–60.Google Scholar

²¹ See Silverman, Kaja, The Acoustic Mirror: The Female Voice is Psychoanalysis and Cinema (Bloomington, 1988), 48, ‘to the degree that the voice-over preserves its integrity, it also becomes an exclusively male voice’; and 50, ‘dominant cinema also holds the female subject more fully than the male subject to the unity of sound and image’. Silverman also takes up Chion's discussion of disembodied voice, agreeing (49) with the truism that ‘voice-over is privileged to the degree that it transcends the body. Conversely, it loses power and authority with every corporeal encroachment, from a regional accent or idiosyncratic “grain” to definitive localization in the image. Synchronization marks the final moment in any such localization, the point of full and complete embodiment’ (Silverman's ‘synchronization’ corresponds to Chion's ‘deacousmatization’).Google Scholar

²² Telephone answering machines proliferate in films where villainy is associated with disembodied voices; an early instance is Kiss Me Deadly (1955), from an era in which the answering machine, hardly yet a common household object, still had an aura of t he technologically feral and dangerous. Chion, Audio- Vision, 131, cites the villain's voice in Kiss Me Deadly as an example of a ‘typical form [of the acousmatic voice] in detective and mystery films, when the “bigboss” who pulls all the strings … is finally revealed’ and notes that while we constantly see a part of this villain's body (his shoes), the deacousmaticisation occurs only when we see his face. Thomas Y. Levin notes that fantasies attached to sending one's voice via telephone, and then inscribing it via machine, view capture of the floating telephonic voice by the answering machine (or voice-mail programme) as a kind of rematerialisation of one's body in another form, as physical traces on the recording medium.Google Scholar See his ‘Before the Beep: A Short History of Voice Mail’, in Essays in Sound 2, ed. Jonson, Annemarie (Darlinghurst, Australia, 1995), 59–67.Google Scholar

²³ For rehearsals of classic literature on this image of maternal speech, as well as critiques of the assumptions that feed it, see Stanton, Domna, ‘Difference on Trial: A Critique of the Maternal Metaphor in Cixous, Irigaray and Kristeva’, in Miller, Nancy, ed., The Poetics of Gender (New York, 1986), 142–59,Google Scholar and Kahane, Claire, ‘Rethinking the Maternal Voice’, Genders, 3 (Fall 1988), 82–91.Google Scholar

²⁴ Critiques that challenge the assumption of operatic music's epistemological certainty, as well as the ideological component in correspondence between music and text or action, are rare indeed; Adorno's Versuch über Wagner is one model; and it is significant that this critical scepticism tends to attach itself to Wagner's operas, perhaps because Wagner himself was so suspiciously compensatory in coining words like Gesamtkunstwerk. For instance, Zizek's, ‘“There is no Sexual Relationship”: Wagner as Lacanian’, New German Critique, 69 (Fall 1996), 7–36,Google Scholar reprinted and expanded in Gaze and Voice as Love Objects, ed. Salecl, Renata and Zizek, Slavoj (Durham and London, 1996), 208–49, demythicises Wagnerian and film-music conceits in terms familiar from Adorno.Google Scholar

²⁵ As, for example, in McClary, Susan, Georges Bizet: ‘Carmen’ (Cambridge, 1992),CrossRefGoogle Scholar or Kramer, Lawrence, ‘Culture and Musical Hermeneutics: The Salome Complex’, this journal, 2/3 (1990), 269–94, esp. 284–94.Google Scholar For one counter-argument, see my ‘Opera, or the Envoicing of Women’, in Musicology and Difference, ed. Solie, Ruth (Berkeley and Los Angeles, 1993), 225–58.Google Scholar

²⁶ Such assumptions underlie notions such as Mary Ann Doane's ‘fantasmatic body’, the illusion of unity created by certain elements in film (such as correspondence between sound and image, or continuous music) that work counter to film's fracturing forces (anything from uncertain or group authorship of the ‘text’ to ragged visual cuts), speculating that the viewer/listener, similarly wishing to understand him- or herself as a ‘unified subject’, consumes the illusion and identifies with this ‘body’; see Doane, , ‘The Voice in the Cinema: The Articulation of Body and Space’, Yale French Studies, 55/1 (1980), 33–50. Adorno and Eisler also, famously, deflated any presumption of continuity as provided by film music that is both fairly continuous and ‘redundant’ in merely reinforcing the emotional or narrative message conveyed by dialogue or action within the visual field;CrossRefGoogle Scholar see Composing for the Films (London, 1986), 23–4. I am grateful to Scott Paulin for this point.Google Scholar

²⁷ Hence the title of Gorbman's, ClaudiaUnheard Melodies: Narrative Film Music (Bloomington, 1987).Google Scholar

²⁸ This music is a classic Leitmotif in the Wagnerian sense; it has been associated from the first with menace, and is eventually pinned exclusively to the Terminator, accompanying claustrophobic shots of Schwarzenegger and his raptor eyes as he cruises Los Angeles looking for his victim.Google Scholar

²⁹ In La Voix au cinéma, 122, Chion writes that the effect of the floating maternal voice in Psycho would have been destroyed by the availability of Dolby sound, which would probably have inspired creation of a spectral physical locus (perhaps somewhere within the theatre) for Mrs Bates's voice.

³⁰ Novalis, , Heinrich von Ofterdingen (Stuttgart, 1976), 45–6.Google Scholar

³¹ See Orledge, Robert, Debussy and the Theatre (Cambridge, 1982), 102–27, for a summary of the project's history.Google Scholar

³² See Lockspeiser, Edward, Debussy et Edgar Poe (Monaco, 1961), 95, Rod: Tenez, j'ai retrouvé cet antique et curieux bouquin de savoir oublié. II y ait parlé des anciens satyrs africains et des Aegipans … Pendent des heures, j'ai rêvé sur la musique qui devait accompagner leurs étranges cérémonies … Lisez … ici … ne croirais t'on pas entendre comme une danse funêbre et passionné (Pendent qu'ils lisent on entend - vaguement - la musique qui imagine R.U.).' Lockspeiser published here in transcription the latest of the various libretto drafts made by Debussy.Google Scholar

³³ Orledge, Debussy and the Theatre, 112–14, argues that an unlabelled sketch page in the Bibliothèque nationale represents this imagined African satyr music from La Chute de la maison Usher, certain orientalisms in the sketch (tam-tam and cymbal beats) are suggestive.Google Scholar

³⁴ Transcriptions in Orledge, Debussy and the Theatre, 115–16, are more accurate than the reconstruction of this scene in the edition of the opera by Juan Allende-Blin (Paris, 1979); the autograph sources are listed in Debussy and the Theatre, 354 n. 24.Google Scholar

³⁵ So strong was the association of A with the pastoral mode that minor talents were inspired to exploit the code; in 1922 Gottfried Bohnenblust, an amateur German poet of small literary gifts, published a collection of pastoral poems - including classical forms such as eclogues as well as verses entitled ‘Siciliane’ - under the title A Dur.Google Scholar

³⁶ La Musique et l'ineffable, 129, ‘La réalité musicale est toujours ailleurs, comme les paysages evoqués, chez Gabriel Fauré, par une expression évasive et amphibolique; cette géographie pneumatique ou l'alibi estompe et brouille sans cesse le repérage univoque des lieux, elle rend fondante et fuyante toute localisation: ne disions-nous pas que la musique, phénoméne temporel, refuse en général toute spatialisation?’Google Scholar

³⁷ This snapping G is also heard within the faded serenade (in the piano part of bars 35–6) (at ‘[et] leurs molles ombres [bleues]’), where Debussy layers a future sound over a present phenomenon. The coincidence of this G, foretelling the end, with that image of ‘shadows’ is perhaps the only overt mapping of word on musical gesture in the song.Google Scholar

³⁸ Cited in Nichols, Roger and Langham-Smith, Richard, Claude Debussy: ‘Pelléas et Mélisande’ (Cambridge, 1989), 4.Google Scholar Raising the question of sound effects in opera, Friedrich Kitder notes that Debussy's music for Pelléas is derived from a new ‘optics and acoustics’ adumbrated in Tristan, one in which orchestral sounds become ‘ambivalent to the extent that the fundamental European distinction between tones and noises, music and nature, becomes blurred’; he reads Debussy's musical sound effects for Maeterlinck's play as ‘grammophonic’, and associated them with Debussy's own discussions of sound and cinema technology in his essay ‘Music in the Open Air’. See , Kittler, ‘Opera in the Light of Technology’, in Languages of Visuality: Crossings Between Science, Art, Politics, and Literature, ed. Allert, Beate (Detroit, 1996), 73–85, esp. 80–1.Google Scholar

³⁹ ‘Mes longs cheveux descendent’ was one original text of Mélisande's tower song in Maeterlinck's play, though an alternative text, ‘Les trois soeurs aveugles’ (‘The King's Three Blind Daughters’, in Jack Mackail's translation), was substituted for the première in London. This text was ultimately set in English by Gabriel Fauré for a later, 1898 London performance, where it was transposed from the Tower scene (Act III scene 2) to Act III scene 1 - not used by Debussy - where Mélisande sings while spinning, à la Gretchen;Google Scholar see Nectoux, Jean-Michel, ‘Le Pelleas de Fauré’, Revue de Musicologie, 67/2 (1981), 180. I owe thanks to Carlo Caballero for information on Faure's setting of the alternative song.CrossRefGoogle Scholar

⁴⁰ In the 1943 film L'Eternel Retour, lean Cocteau's modern re-telling of both Tristan und Isolde and Pelléas, this inversion of femininity/masculinity and covert feminisation of Pelléas/Tristan (here named ‘Patrice’) is made explicit. Patrice has an acoustic marker: he whistles and imitates birds, initially as a kind of party trick, but later also as a secret signal to the Mélisande/Isolde character. Thus the mutation of human voice into ‘birdsong’ and lingua ignota is grafted directly on to the male character, perhaps even as a private autobiographical message: Jean Marais, who played Patrice, was Cocteau's lover.Google Scholar

⁴¹ This truism constitutes a kind of psychoanalytic, literary-critical, and musicological monolith. One need only trace recent accounts, from all these domains, of the voice of The Queen of the Night, such as that of Poizat, Michel, Le Cri du diable: la jouissance Iyrique sacrée (Paris, 1991), 219–26. The Queen's voice is characterised as a figure ‘making present an imaginarised “maternal” ’, as constituting ‘a musical version of sheer high-pitched shrieking’ which bears evidence to a ‘will to destroy the Word’ (220). She is, further, a figure of ‘maternal bliss’ who is initially accorded positive value, who bans the ‘law of the Word and the Father’ from her realm (221); yet, her voice becomes the focus of a profound ambivalence whose suasions must be renounced by Tamino, as he enters subjecthood, in favour of a realm of ‘culture, architecture, and the Temple, which must be substituted for the realm of savage nature represented by Act I’ (220–1).Google Scholar

⁴² Adorno made this claim in 1928 in ‘Nadelkurven’, Musikblätter des Anbruch, 10 (02 1928), 47–50;Google Scholar‘Curves of the Needle’, trans. Levin, Thomas Y., October, 55 (Winter 1990), 48–55;Google Scholar the male voice is praised for its ability to survive recording intact, since the male subject is identical to his voice, while the female voice is diminished by subtraction when her body is erased: her voice and body are unsunderable. For leitmotivically consistent treatments of this theme, see the essays of Dunn, Leslie C. and Jones, Nancy A., eds., Embodied Voices: Representing Female Vocality in Western Culture (Cambridge, 1994), many of them scholarly offshoots from Silverman's The Acoustic Mirror, which exhaustively analyses embodiment in both the filmic habit of binding female voice to its visual origins through synchronisation, and burdening it with excess acoustic grain (in the form of accents, breathiness and the like) that reinvokes its physical origins.Google Scholar

⁴³ Maeterlinck's alternative song text, the more formally poetic ‘Les trois soeurs aveugles’ with its multiple verses, suggests the contrary: that Melisande is simply the medium for an a priori text, hence that she occupies a more conventionally feminine position.Google Scholar

⁴⁴ The Rhapsody to Mélisande's voice in Act IV scene 4 (‘on dirait que ta voix a passé sur la mer au printemps’) was the first fragment of the play set to music by Debussy; the sketch for the Rhapsody dates from late summer 1893, and the pre-composed passage was then integrated into the entire scene at a later point, retaining its character as a kind of foreign body in the completed scene. One still senses this - there is a slide into formal lyricism, almost an aria effect - when Pelléas begins to sing. This quality lends the Rhapsody a peculiar ambiguity: though it is Mélisande's voice that the poetry converts into pure grain, in the opera it is the singer's rich baritone, caught in a sudden cage of aria, that has the effect of being both fetish and an object of desire. This kind of fetishised singing voice can be differentiated from what psychoanalytic theory would call a ‘voice object’; for instance, Reinhard Meyer-Kalkus argues that the aesthetic appeal of the beautiful singing voice is a manner of veil, which shields us from anxiety about the voice-object (that is, the voice of the ‘Other’); see ‘Jacques Lacans Lehre von der Stimme als Triebobjeckt’, in Kulturelle Perspektiven auf Schrijt und Schreibprozesse, ed. Raible, Wolfgang (Tübingen, 1995), 259–307, esp. 304–7.Google Scholar

⁴⁵ La Musique et l'ineffable, 130.Google Scholar

⁴⁶ This is what Kittler implies when he writes ‘fin-de-siècle opera is already film theatre’, ‘Opera in the Light of Technology’, 81.Google Scholar