Published online by Cambridge University Press: 27 August 2008
‘Quoi du reste’, to paraphrase Derrida on Hegel in Glas, ‘ici, maintenant, d'une Carmen?’ What's left of Carmen here and now? Aside from its intrinsic interest, the question seems worth asking in light of a bias that recent treatments of opera – particularly those influenced by poststrucuralist theory – seem to betray. The most prominent, Catherine Clément's Opera, or The Undoing of Women, Michel Poizat's The Angel's Cry and Jeremy Tambling's Opera, Ideology, and Film, regard opera as an institution rather than as a body of texts. Each of the authors, to my mind at least, allows a prior structure or structures – the systemic presence of male domination and its construction of women in society, a quasi-Lacanian understanding of unconscious fantasy, the bourgeois construction of operatic experience – to constrain what operas, and opera, can mean. They thus produce what are in effect reception studies or analyses of the audience, which is perhaps why they operate at some distance from the detail of the texts, musical or verbal, of the operas they analyse.
1 Godard, Jean-Luc, ‘Faire un film comme on joue un quatuor’, interview with Jacques Drillon, April 1983, Jean-Luc Godard par Jean-Luc Godard, ed. Bergala, Alain (Paris, 1985), 574.Google Scholar
2 Opera, or The Undoing of Women, trans. Wing, Betsy (Minneapolis, 1988)Google Scholar; The Angel's Cry: Beyond the Pleasure Principle in Opera, trans. Denver, Arthur (Ithaca, 1992)Google Scholar; Opera, Ideology, and Film (New York, 1987).Google Scholar The most important, though necessarily partial, exceptions to this generalisation are Groos, Arthur and Parker, Roger, eds., Reading Opera (Princeton, 1988)CrossRefGoogle Scholar, and Abbate, Carolyn and Parker, Roger, eds., Analyzing Opera: Verdi and Wagner (Berkeley, 1989).Google Scholar
3 Tambling, whose methodology is closer to my own, and whose materialist perspective allows for a greater appreciation of historical differences in the social and political meaning of an opera, is an apparent exception here, but he sees little progressive social use for the modem institution of opera, and his textual readings are almost all readings of films. His basic thesis is that film has the potential to ironise opera and make it available for progressive political reflection, in a way he seems to believe opera – as a bourgeois institution – cannot do for itself.
4 The distinction between the last two is only relative. Both Clément and Poizat explore the implications for gender construction of a psychoanalytic perspective on opera, and both, especially Poizat, project a generic imaginary for the operatic subject that is in effect essentialist.
5 These views thus reinscribe in their account of opera itself the very image of femininity they criticise opera for perpetuating.
6 See the full listing in Tambling, chapter 1, supplemented by McClary, Susan, George Bizet: ‘Carmen’, Cambridge Opera Handbooks (Cambridge, 1992), 130–46.CrossRefGoogle Scholar
7 On this, see McClary, , 130.Google Scholar
8 The term is Lacan's. See ‘The Subversion of the Subject and the Dialectic of Desire in the Freudian Unconscious’, Ecrits, A Selection, trans. Sheridan, Alan (New York, 1977), 292–325, especially 303.Google Scholar The best gloss on this article is Zizek, Slavoj, The Sublime Object of Ideology (New York, 1989), 87–129.Google Scholar
9 It is perhaps worth pointing out that the camera does not pre-empt the looking of the actual, as opposed to the virtual, spectator. That looking remains located in the body – it is only that what it looks at, the filmed image, has been pre-visualised, so that now one looks at a looking.
10 See Doane, Mary Ann, ‘Ideology and the Practice of Sound Editing and Mixing’, in Film Sound Theory and Practice, ed. Weis, Elisabeth and Belton, John (New York, 1985), 54–62.Google Scholar
11 This also seems to be why she keeps calling Joseph ‘Joe’, because she wants to fit him into the version of Carmen Jones she is trying out. He as constantly corrects her and insists on ‘Joseph’, in resistance to being captured by her or her story, though in doing so he succeeds only in putting himself into Godard's own Pierrot le fou instead, where this struggle over naming is a running gag.
12 ‘Faire un film’, 576.Google Scholar
13 Joseph tacks this meaning down at one point when he speaks to the Micaëla-figure, Claire, of his feelings for Carmen as a ‘tide’ (she drily remarks that tides go out, too), and comes back to it at the end. Like Saura, who uses the Act III prelude for this purpose, and like Brook, who uses only José's half of the Act II love duet, Godard seems to feel the lack of a love-theme for Carmen and José in Bizet, and supplies it with the sea and the gulls.
14 See the scenario, ‘Séquence 17’, Godard par Godard, 566.Google Scholar
15 As several commentators have noted, Prénom Carmen is closely related in plot to Pierrot le fou (see n. 11). The vaguely revolutionary criminality of Carmen's gang – the way one cannot tell if their motives are criminal, political or just a hunger for excitement – is more than a little reminiscent of Bande à part, and especially of La Chinoise. In the larger sweep of Godard's career, especially after his ‘return’ to such themes in Sauve qui peut (la vie) and Passion, Prénom Carmen clearly looks back to the fascination with cultural representations (especially images) of women that informs the trilogy Une femme est une femme (itself a film about the musical accompaniment of everyday life), La femme mariée and Deux ou trois choses que je sais d'elle, and forward to Je vous salue, Marie as part of a duet of meditations on the stock figures of the whore and the virgin. See Mulvey, Laura, ‘Godard's Iconography of Women’, in Jean-Luc Godard's ‘Hail Mary’. Woman and the Sacred in Film, ed. Locke, Maryel and Warren, Charles (Carbondale and Edwardsville, 1993), 45–6.Google Scholar
16 The use of this word as a noun is now common in post-Lacanian theory. The word in French, as used by Bachelard, Merleau-Ponty and Sartre before Lacan, and by Barthes as well, is imaginaire, and it means not the mental faculty of picture-making but the repertory of images used by that faculty.
17 This account is, with individual inflections, the organising framework of recent discussions of Spanish film, and especially of the Nuevo tine erpañol with which Saura is associated. See D'Lugo, Marvin, The Films of Carlos Saura. The Practice of seeing (Princeton, 1991)Google Scholar; Higginbottom, Virginia, Spanish Film Under Franco (Austin, 1988)Google Scholar; Kinder, Marsha, Blood Cinema: The Reconstruction of National Identity in Spain (Berkeley, 1993).Google Scholar
18 See McClary, , 136.Google Scholar
19 It is, however, by no means the only thing to do so. As the Film proceeds, Saura increasingly uses the trick (also common in Godard) of having the soundtrack bridge a cut between scenes, so that the same sounds, whether Bizet, flamenco or even the noise of stamping feet, are now diegetic (produced in imagination or reality by the characters), now extra-diegetic (unheard by the characters, functioning as background or commentary).
20 D'Lugo, , 205.Google Scholar
21 McClary, , 137.Google Scholar
22 From this point of view, the nearest analogy in operatic terms to ‘flamenco’ as the film represents it is verismo, the attempt in Italian opera at the turn of the century to give new purchase to the intensity of operatic emotions by relocating them in low or popular culture. The strategy in both cases is to justify and protect the purity, simplicity and directness of feelings by ascribing them to a society more innocent than the contemporary audience. The kinship of the other two ballets by the Gades troupe filmed by Saura, Blood Wedding and El amor brujo, to works like Pagliacci and Cavalleria rusticana is evident.
23 As in Prénom Carmen, the characters in Saura's film do not know the opera well, and are willing to take their reading of it on faith from Antonio – who prefers Mérimée.
24 McClary, , 135–6.Google Scholar
25 The closest analogue in Sauna's career is the ending of his 1970 Garden of Delights. The family of an industrialist who is amnesiac and wheelchair-bound spends most of the movie acting out scenes from his past for him, in an effort to get him to remember the number of his Swiss bank-account and the combination of his safe. The failure of this project (and its reversal) is conveyed in the last frames by an image of the industrialist in his wheelchair in the middle of an open field, surrounded by the entire family circling around him in wheelchairs of their own. The mise-en-scéne departs from the generally realistic mode of the rest of the movie in order to sum up in a single surreal image: the underlying identity between the family superficially in possession of their faculties and the crippled victim of repression at the centre of their attention.
26 I am indebted to Dion Farquhar for this point.
27 Brook, Peter, ‘The Art of Noise’, in The Shifting Point. Theatre, Film, Opera, 1946–1987 (New York, 1987), 170.Google Scholar
28 McClary, , 137–8.Google Scholar
29 McClary, , 140.Google Scholar
30 Mérimée, Prosper, ‘Carmen’, in Lokis et autres contes (Paris, 1964), 240.Google Scholar Carmen's last words in the film, sung before she knows Escamillo is dead, are Bizet's version of the last sentence, ‘Libre elle est née’.
31 Tambling's rather scattered discussion, Opera, Ideology, and Film, 33–5Google Scholar, is perhaps something of an exception.
32 Wills, David, ‘Carmen: Sound/Effect’, Cinema Journal, 25 (1986), 34.CrossRefGoogle Scholar
33 Dickenson, Frances, in Milne, Tom, ed., The Time Out Film Guide, 3rd edn (New York, 1993), 730.Google Scholar
34 McClary, , 138, 141.Google Scholar
35 ‘When I had directed [Marat/Sade] for the stage, I had not attempted to impose my own point of view on the work: on the contrary, I tried to make it as many-sided as I could. As a result, the spectators were continually free to choose, in each scene and at every moment, the points which interested them most. Of course, I too had my preferences, and in the film I did what a film director cannot avoid, which is to show what his own eyes see. In the theatre, a thousand spectators see the same thing with a thousand pairs of eyes, but also at the same time they enter into a composite, collective vision. This is what makes the two experiences so different.’ ‘Filming a Play’, The Sh Ring Point, 190.Google Scholar
36 I think in particular of the end of Robert Aldrich's Vera Cruz, in which Burt Lancaster's final on fight with Gary Cooper ends with him repeating the broad On and juggler-like fancy holstering of his on that he has practised after each of his victories in the movie, and then falling dead.
37 See, for example, ‘The Goddess and the Jeep’, The Shfng Point, 164–5.Google Scholar
38 This description would apply more to Brook's version (despite its genuinely cinematic dimension), where the major ‘new’ effects are in the staging, the revised libretto and the newly orchestrated score. From this point of view, Brook's film is more an to Regieoper, the directorial appropriation of traditional operas through productions that run aggressively against the grain not only of traditional performance values, but of the text itself, a practice common in contemporary Germany, and most familiar in the U.S. in Peter Sellars' versions of Mozart.
39 The project was initiated by Gaumont, the French film consortium, in the wake of the spectacular success of their production of Losey's, JosephDon Giovanni (1979).Google Scholar According to D'Lugo, it was first offered to Saura as ‘a film version of the Bizet opera to be made in France with a maximum of fidelity to the original text’, The Films of Carlos Saura, 202.Google Scholar
40 See McClary, , 25–6Google Scholar, and her notes for a summary with references to the Oeser-Dean controversy and the state of the text.
41 The locus classicus for this discussion is Derrida, Jacques, Of Grammatotogy, trans. Spivak, Gayatri Chakravorty (Baltimore, 1976), 141–64.Google Scholar The same considerations of course apply to the libretto, and here the open possibilities of the reste are even more pressing, since Rosi does not begin to make use of all the available words. See, for example, the libretto as printed in the ENO/Royal Opera Carmen (London and New York, 1982)Google Scholar, which gives a full, if unscholarly, account of variant versions, material frequently cut in performance, etc., more valuable than a more ‘disciplined’ editio variorum precisely because of its witness to practice.
42 One of the most characteristic shots in the film, repeated many times, is a medium-high-angle travelling shot in which the camera also pans, moving across and rotating in a field in deep focus. One function of this complex compound shot is to pick out and move with a given character while continuing to emphasise the fullness and bustle of the world around him or her. An example is the opening scene at the bullfight, in which the camera moves around the arena from the stands so as to follow Escamillo's triumphal parade in the ring while also surveying a sizeable chunk of the spectators. Many of the shots that delineate the ordinary social world of Seville work in the same way, especially those that present Carmen outdoors, i.e., in Acts I and III.
43 This is another example of the technique described above, using the extended duration of scenic music as a way to get more information and additional narrative into the film.
44 McClary notes the way Bizet revises the function of the traditional opéra-canigve chorus to make this social point: ‘Chameleon-like, Carmen's choruses represent a variety of social groups … Carmen … always fits into and is supported by the community as represented by the various choruses’, 46–7.
45 Compare the account of Carmen's response to the flower song, given above.
46 See McClary, , 15–16 and 44–7Google Scholar, for a brief, pointed account of the class associations of this musical style.
47 Un ballo in maschera, with its interest in shifting alliances, is full of this effect, as is, for similar reasons, Don Carlos.
48 A fuller reading would need to take into account the way Rosi and Bizet position this aria as a contrast to Carmen's style of reflection on her situation in the card song, which almost immediately precedes this scene. In both Bizet and Rosi, we are invited to notice the same difference between an essay at self-definition – Carmen's – conducted in the presence of an audience, and one like Mcaëla's, whose surface humility serves as a pretext for expanding the self to appropriate an empty landscape. Once again the fact of singing in a ‘realistic’ mise-en-scène tags an attitude towards the world, and in this case the likelihood that this amount of noise would be sure to attract attention questions the song's message of secret fear and identifies the underlying aggressivity typical of Mcaëla, who never shrinks from asserting herself when she can blame it on a mother or God.
49 This is shorthand for something that could be elaborated in terms of Rosi's own themes and polemic, that is, an analysis of a movie that enlists powerful (and spectacularly naturalistic) images in support of what he would surely regard as lies about the possibilities for women to transcend class boundaries. In this context the film is politically ‘bad’ from Rosi's point of view, whatever else it may be.
50 The suggestion is not entirely facetious. Given Bizet's announced fix on Boieldieu's opera as the enemy Caren was intended to slay, a serious investigation is needed into the verbal and musical intertextualities of the two works, starting with such things as the continual use in Carmen of forms of the expression prendre garde, the signature of Boieldieu's heroine: ‘Prenez garde, prenez garde, / La Dame Blanche vous regarde. ‘ Prenez garde, prenez garde, / La Dame Blanche vous entend!’
51 See Wills, , 34.Google Scholar
52 The choreography for the film was done by the same Antonio Gades who stars in Saura's Carmen.
53 Rosi thus projects an estimate of the ‘toreador song’ close to Bizet's own, who ‘seems to have regarded [it] as a sop for the Opéra-Comique audience: he is reported to have said concerning this number, “so they want trash [de l'ordure]? All right; I'll give them trash”’; McClary, , 46.Google Scholar
54 See Dahlhaus, Carl, Nineteenth-Century Music, trans. Robinson, J. Bradford (Berkeley, 1989), 290–2Google Scholar, and McClary, , 107.Google Scholar
55 Dahlhaus, , 280–2.Google Scholar
56 As I tried to suggest above, something of the sort can cling to male-male duets as well, as in Escamillo's and José's duet in Act III.
57 This oscillation, insofar as it represents a refusal to be contained by a single structure or metaphor of relationship, may explain the unresolved debate in the literature on the meaning of the bullfight in BiZet's Caren. The critics are agreed that the bullfight is proposed as a master-image by the opening of the film, and note that its end is structured (as is the opera, musically) by an insistent parallel cross-cutting between Escamillo's bullfight and the final confrontation between José and Carmen. The problem arises in trying to work out the parallel in relation to the characters – who is the matador, who the bull, and where is the third person in the triangle to be located in the bullfight? Even Wills's dazzling suggestion that Carmen is the cape (borne out, as David Rosen pointed out to me, by Rosi's continual play with the movements of Carmen's dress), does not solve all the difficulties. I think the reason for the problem is that by the end of the film Carmen and her relationships have come to evade or refuse such master-images so that they are no longer capable of containing and dismissing her. Instead, she reads them: Carmen will tell you more about the bullfight (about Escamillo, for whom Carmen might well seem to be the cape, something he manipulates in order to beat José) than the bullfight will tell you about Carmen. See Wills, , 36–7Google Scholar, Tambling, , 35–7Google Scholar, McClary, , 142.Google Scholar
58 See Furman, Nelly, ‘The Languages of Love in Carmen’, in Reading Opera, 168–83.Google Scholar