¹ Such an idea is by no means restricted to feminist writers. See, for example, Conrad, Peter, A Song of Love and Death: The Meaning of Opera (New York, 1987), 359–60.Google Scholar

² Clément, Catherine, Opera, or the Undoing of Women, trans. Wing, Betsy (Minneapolis, 1988), 89.Google Scholar

³ Madness and Civilization (New York, 1965).Google Scholar

⁴ see Kolodny, Annette, ‘A Map for Rereading: Gender and the Interpretation of Literary Texts’, The New Feminist Criticism, ed. Showalter, Elaine (New York, 1985), 50–4Google Scholar; and Gilben, Sandra M. and Gubar, Susan, The Madwoman in the Attic: The Woman Writer and the Nineteenth-Century Literary Imagination (New Haven, 1979).Google Scholar

⁵ McClary's work has had a considerable influence on scholars wishing to address issues of gender in classical music. As the first to assert the relevance of gender to the previously ‘unisex’ (read male) preserve of musicology, she both deserves accolades and demands serious responses. I would hope that feminist musicologists inspired by McClary will test some of her theories on specific repertories, and that the second phase of the endeavour of feminist musicology will emphasise depth over breadth. We are still awaiting examples of this type of serious response to McClary's project; the recent outburst of reaction offers little in the way of clear-thinking, even-handed response.

⁶ Feminine Endings (Minneapolis, 1991), 90–9.Google Scholar

⁷ The interpretation of Lucia's madness as resistance predates McClary and feminist theory by more than a century: in Madame Bovary, Flaubert has Emma wonder at a performance of Lucie de Lammermoor, ‘Oh, why had not she, like this woman, resisted?’ Emma's plaint is quoted both by McClary as support for the notion of Lucia as a feminist heroine and by Elaine Showalter, who takes the opposite position, arguing that operatic madwomen ultimately are victims. See McClary, , Feminine Endings, 99Google Scholar; and Showalter, , The Female Malady (New York, 1985), 17.Google Scholar Emma's reaction has generally been interpreted as applying to the mad scene; it will become significant later that the idea in fact comes to Emma during the opera's second act, and that she leaves the opera with her husband and lover just as the mad scene is beginning.

⁸ Female Malady (see n. 7). For a more general history of madness in the same period, see Porter, Roy, Mind-Forg'd Manacles (London, 1987).Google Scholar

⁹ This painting is discussed in Showalter, , Female Malady, 2Google Scholar; for a more extensive study of visual representations of madness, see Gilman, Sander L., Seeing the Insane: A Cultural History of Psychiatric Illustration (New York, 1976).Google Scholar

¹⁰ Mulvey, Laura, ‘Visual Pleasure and Narrative Cinema’Google Scholar, and ‘Afterthoughts on “Visual Pleasure and Narrative Cinema” inspired by King Vidor's Duel in the Sun (1946)’, in Visual and Other Pleasures (Bloomington, 1990), 14–28 and 29–38.Google Scholar For an analysis of the nude in European painting as a commodity presented for the pleasure of the male spectator, see Berger, John, Ways of Seeing (London, 1972), 45–64.Google Scholar

¹¹ The Bride of Lammermoor (New York, 1973), 323.Google ScholarMcClary, (Feminine Endings, 91)Google Scholar points out that Lucy's use of Scottish dialect at this point suggests a rsistance to her family's aristocratic pretensions. The roughness of Lucy's speech may also owe something to the bawdy language of hte mad Ophelia.

¹² The notion that the mere presence of voice endows operatic heroines with power despite their victimisation by plot has been explored by Paul Robinson, who has proposed in response to Clément that ‘perhaps the single most important musical fact about opera's female victims is that they sing with an authority equal to that of their male oppressors.… This fundamental vocal fact means that women in opera are rarely experienced as victims. Rather, they seem subversive presences in a patriarchal culture, since they so manifestly contain the promise – or rather the threat – of women's full equality.‘ See ‘It's Not Over Till the Soprano Dies‘, New York Times Book Review (1 01 1989).Google Scholar Robinson's argument has been developed by Abbate, Carolyn in Unsung Voices (Princeton, 1991), ixGoogle Scholar, although in this context the plot/voice debate is only a prelude to a broader, more personal concept of voice evolved as the book progresses.

¹³ This aestheticisation has ensured that Lucia could indeed captivate an audience: enthusiasm for her mad scene has been long-lived. Although the première of Lucia had a comparatively lukewarm reception, later audiences flocked to Nellie Melba‘s performances of the mad scene as a concert piece, and when the opera was performed complete, Edgardo‘s suicide was usually omitted so that the opera could end with the mad scene. On the early reception of Lucia, see Ashbrook, William, ‘Popular Success, the Critics and Fame: The Early Careers of Lucia di Lammermoor and Belisario‘, this journal, 2 (1990), 65–81.Google Scholar

¹⁴ On the power of Ophelia's songs as music, see Rosand, Ellen, ‘Operatic Madness: A Challenge to Convention’, in Music and Text: Critical Inquiries, ed. Scher, Steven Paul (Cambridge, 1992), 241–87CrossRefGoogle Scholar, and Leslie Dunn, ‘Ophelia's Songs in Hamlet: Music, Madness and the Feminine’, unpublished paper delivered at the conference ‘Feminism and Music: Toward a Common Language’, Minneapolis, 1991.

¹⁵ Raby, Peter, Fair Ophelia (Cambridge, 1982).Google Scholar Smithson's interpretation continued an already existing tradition of stage madness; the costuming of madwomen in white was already firmly enough established towards the end of the eighteenth century for the practice to occasion satire in Sheridan's The Critic (1777).Google Scholar see Barish, Jonas, ‘Madness, Hallucination, and Sleepwalking’, in Verdi's Macbeth. A Sourcebook, ed. Rosen, David and Porter, Andrew (Cambridge, 1984), 149–55.Google Scholar

¹⁶ Female Malady (see n. 7), 10–17Google Scholar

²⁰ This recalls Michel Poizat's Lacanian theory of the significance of high notes, or, as he calls them, ‘les Notes Bleues’. Poiat argues that the power of certain moments in opera, when words and music are subsumed inot an inarticualte cry, derives from the universal memory of hte infant's cry, which is irrecoverable. Poizat, Opéra, ou Le cri de l'ange; excerpts translated by Denner, Arthur in this journal, 3 (1991), 195–211.Google Scholar

²¹ ‘Opera and Homosexuality: Seven arias’, Yale Journal of critiscism, 5 (1991), 244.Google Scholar

²² It is well known that the flute was a late addition to the score, replacing the far more uncanny glass harmonica that Donizetti originally wanted. In the autograph, the completed glass harmonica part is crossed out and a flute part added, suggesting that the change was made fairly late in composition (see Fig. 3). The reasons are unknown, although the unavailability of a glass harmonica player in Naples seems plausible. see Ashbrook, William, Donizetti (London, 1965), 417.Google Scholar

²³ On the origin of the cadenza, see Ashbrook, , Donizetti and his Operas (Cambridge, 1982), 376CrossRefGoogle Scholar; and Barblan, Guglielmo, L'opera di Donizetti nell'età romantica (Bergamo, 1948), 125.Google Scholar

²⁴ ‘Conventional double aria structure’ consists of four movements: the scena (in recitative), a lyrical slow movement, the tempo di mezzo and the cabaletta, often a virtuosic showpiece. The text of the scena is written in the unrhymed, free verse known as versi sciolti, while the other three movements are in versi lirici or metred verse.

²⁵ The originality of the cavatina compared to the mad scene again recalls Emma Bovary's remark about Lucia after hearing Act I. Perhaps it is no accident that Emma admires Lucia's resistance after this cavatina and not after the mad scene. However, history suggests otherwise: the French version of the opera that Flaubert was writing about probably did not include ‘Regnava nel’, but rather used the substitute aria from Rosmonda d'Inghilterra, ‘Que n'avons nous des ailes’. See Ashbrook, , Donizetti and his Operas (n. 23), 381.Google Scholar

²⁶ A well-known example of this type is Leonora's narrative from Act I of Il trovatore, ‘Tacea la notte’.

²⁷ The melodic pattern of lyric form could be schematised as a a′ b a′′ (or c), with each phrase lasting four bars and setting two lines of text. In the minor mode, such a pattern is often modified, ending with a new phrase in the parallel major.

²⁸ Other examples occur in Donizetti's Linda di Chamounix and Bellini's Il pirata, while the related convention of recalling a song from the past, though not one heard earlier in the opera, is employed in Donizetti's Anna Bolena and Maria Padilla, and again in Bellini's I Puritani. See Döhring, Sieghart, ‘Die Wahnsinnszene‘, in Die ‘couleur locale‘ in der Oper des 19. Jahrhunderts, ed. Becker, Heinz (Regensburg, 1976), 279–310.Google Scholar

²⁹ Ashbrook, , Donizetti and his Operas (see n. 23), 379Google Scholar

³⁰ Another dimension of the power invested in the orchestra by quotation lies in the fact that quotations may have the power to question the conventional status of operatic characters and their orchestra. Although we accept the use of quotation as conventional in itself, hearing the orchestra recall a theme without its being signalled in the sung text calls into question the subordinate role we generally assign to ‘accompaniment’. This often entails a shift in the audience's understanding of where the composer's voice is situated – a temporary consciousness of the apparent power of characters and orchestra to structure their own discourse, separate from the overall controlling influence of the composer. See Cone, Edward T., ‘The World of Opera and its Inhabitants’, in Music: A View from Delft (Chicago, 1989), 125–38.Google Scholar

³¹ Madness and Civilization (see n. 3), 100 and 107.Google Scholar