PFSCL XXXVIII, 75 (2011) The Death of Camille in Corneille’s Horace: Performance, Print, Theory M ICHAEL H AWCROFT (Keble College, Oxford) When, in 1660, Pierre Corneille published an edition of his plays to date, he famously prefaced each of the three volumes with a discourse considering his own theatrical practice in the light of Aristotle’s Poetics and the text of each play with an “Examen”, a critical evaluation of its strengths and weaknesses as he perceived them to be in retrospect. “ [ J’ ] en dirai mes pensées tout simplement”, says Corneille at the beginning of the first discourse. 1 This claim is deceptive. It is true that Corneille’s theoretical writing, when compared to La Mesnardière’s La Poétique (1639) or d’Aubignac’s La Pratique du théâtre (1657), cultivates concision and avoids extensive explicit engagement with theorists other than Aristotle and dramatists other than himself. But concision and implication often lead to a style marked by irony and indirection so that Corneille’s apparently simple assessments of his dramatic practice can provoke much critical discussion. The “Examen” he wrote for Horace, a play first performed in 1640 and published in 1641, is an example of the provocative nature of his own critical practice. 2 Few critics fail to engage with Corneille’s own critical account of his perceived reasons for the play’s unsuccessful ending. He agrees with his seventeenth-century critics that the death of Camille spoils the ending, but not, as most of them allegedly thought, because her death 1 Pierre Corneille, Œuvres complètes, 3 vols, ed. Georges Couton (Paris: Gallimard, 1980-87), III, 119. All references to Corneille are to this edition. Although quotations from the Discours and “Examens” are taken from this edition, I acknowledge the extremely helpful annotation in the following two editions: Pierre Corneille, Writings on the Theatre, ed. H. T. Barnwell (Oxford: Blackwell, 1965); Trois Discours sur le poème dramatique, ed. Bénédicte Louvat and Marc Escola (Paris: Flammarion, 1999). 2 For a discussion of some of its complexities, see David Maskell, “Corneille’s ‘Examens’ Examined: The Case of Horace”, French Studies, 51 (1997), 267-80. Michael Hawcroft 44 was visible on stage (first paragraph), rather for a combination of three other reasons: because Horace’s murder of Camille occurs without adequate preparation (second paragraph); because there is no necessary link between Horace’s successful fight against the Albans and the murder of his sister, so the play has two actions rather than one (third paragraph); and because it is unsatisfactory that Camille suddenly becomes a major character in the fourth act, having been a minor one in the previous three (third and fourth paragraphs). Much modern critical discussion focusses on these opening paragraphs of the “Examen” as a way of exploring the unity of the play’s action. Many critics find that Corneille was too harsh on himself and find ways of restoring unity, usually by supplying a psychological motivation that seeks to make compelling sense of Horace’s triumph and his sister’s murder. 3 Others take Corneille’s critical insights seriously and explore the text of the play in the light of Corneille’s engagement with Aristotle and French critics of his time. 4 Yet others read Corneille’s bold retention of Horace’s murder of his sister for its polemical, judicial and political resonances. 5 What all the modern critics do is to skirt over the implications of the first paragraph of Corneille’s “Examen”. In one sense Merlin-Kajman is right that its topic is of secondary importance to Corneille: “L’autocritique de 1660 ne concerne donc pas vraiment le meurtre de Camille, mais le péril où ce meurtre met Horace. La question de savoir s’il faut ensanglanter la scène ou ne pas l’ensanglanter apparaît de ce point de vue secondaire” (“Réécriture cornélienne”, 104). Yet, in its apparent simplicity, the opening paragraph raises important questions about the editions of Corneille’s plays, the status of the printed form of a work of theatre, performance practices, and, above all, issues of dramatic theory as they were understood at the 3 See, for instance, the two articles by Christopher Gossip: “The Unity of Horace”, Modern Language Review, 93 (1998), 345-55; “Héros trop magnanime: le crime d’Horace et son châtiment”, Papers on French Seventeenth-Century Literature, XXXXVI-71 (2009), 543-55. 4 See, for instance, Georges Forestier, Essai de génétique théâtrale: Corneille à l’œuvre (Paris: Klincksieck, 1996), pp. 74-79; and Marc Escola, “Récrire Horace”, Dixseptième siècle, 216 (2002), 445-67, and his edition of Horace (Paris: Garnier Flammarion, 2001). 5 See, for instance, Hélène Merlin-Kajman, “Ré-écriture cornélienne du crime: le cas d’Horace”, Littératures classiques, 67 (2009), 101-14, and her “Horace, l’équivoque et la dédicace”, Dix-septième siècle, 182 (1994), 121-34; and Alain Brunn and Tiphaine Karsenti, “Pourquoi Horace s’enfuit-il? La bienséance, rapport ou limite”, Dix-septième siècle, 223 (2004), 199-212. The Death of Camille in Corneille’s Horace 44 time and as they came to be understood later. 6 These issues can only be teased out by a patient reading of the text, a critical engagement with its modern commentators and an attentive ear to its intertexts. This will lead us to a better understanding of what is at stake in the opening paragraph of the “Examen”, a more accurate grasp of the nature of dramatic theory in the middle of the seventeenth century and another possible reason to account for the dramatic problems posed by Horace’s murder of Camille. This is the whole of the first paragraph of the “Examen”: C’est une croyance assez générale que cette Pièce pourrait passer pour la plus belle des miennes, si les derniers Actes répondaient aux premiers. Tous veulent que la mort de Camille en gâte la fin, et j’en demeure d’accord: mais je ne sais si tous en savent la raison. On l’attribue communément à ce qu’on voit cette mort sur la Scène, ce qui serait plutôt la faute de l’Actrice que la mienne, parce que quand elle voit son frère mettre l’épée à la main, la frayeur si naturelle au sexe lui doit faire prendre la fuite, et recevoir le coup derrière le théâtre, comme je le marque dans cette impression. D’ailleurs, si c’est une règle de ne le point ensanglanter, elle n’est pas du temps d’Aristote, qui nous apprend que pour émouvoir puissamment, il faut de grands déplaisirs, des blessures, et des morts en spectacle. Horace [the Latin writer on the art of poetry] ne veut pas que nous y hasardions les événements trop dénaturés, comme de Médée qui tue ses enfants, mais je ne vois pas qu’il en fasse une Règle générale pour toutes sortes de morts, ni que l’emportement d’un homme passionné pour sa Patrie, contre une sœur qui la maudit en sa présence avec des imprécations horribles, soit de même nature que la cruauté de cette mère. Sénèque l’expose aux yeux du Peuple, en dépit d’Horace; et chez Sophocle, Ajax ne se cache point au Spectateur lorsqu’il se tue. L’adoucissement que j’apporte dans le second de ces Discours pour rectifier la mort de Clytemnestre ne peut être propre ici à celle de Camille. Quand elle s’enferrerait d’elle-même par désespoir en voyant son frère l’épée à la main, ce frère n’y laisserait pas d’y être criminel de l’avoir tirée contre elle, puisqu’il n’y a point de troisième personne sur le Théâtre à qui il pût adresser le coup qu’elle recevrait, comme peut faire Oreste à Egisthe. D’ailleurs, l’Histoire est trop connue, pour retrancher le péril qu’il court d’une mort infâme après l’avoir tuée: et la défense que lui prête son père pour obtenir sa grâce n’aurait plus de lieu, s’il demeurait innocent. Quoi qu’il en soit, voyons si cette action n’a pu causer la chute de 6 One of the issues concerns the significance of the terms bienséance and bienséances in the mid-seventeenth century, since they are often applied by modern critics, as if in an explanatory way, to elucidations of the first paragraph of the “Examen”. I deal with this issue in a separate, though related article: “The Bienséances and their Irrelevance to the Death of Camille in Corneille’s Horace”, see pp. 463-477 in this issue. Michael Hawcroft 44 ce Poème que par là, et si elle n’a point d’autre irrégularité que de blesser les yeux. (I, 839-40) Seventeenth-Century Reactions Corneille’s starting point is a discussion of generally held views about his play. It was thought to be a good play, but spoilt by the death of Camille because the actress playing her was struck dead on stage. We have some evidence for contemporary reaction, but it does not entirely concur with Corneille’s summary of it. 7 The first private performance of the play was on 9 March 1640; the first public performance in May of the same year. But Corneille had already consulted Chapelain, who wrote to Balzac on 19 February: “il y a une quantité de belles choses et du même esprit du Cid. Néanmoins je voudrais pour sa perfection, qu’il eût inventé et disposé autrement qu’il n’a fait” (quoted by Couton, I, 1535). These comments do not tell us any specific observations that Chapelain might have made at this stage. We know too, however, that Corneille read his play to a group that included Chapelain and d’Aubignac and sought their advice. D’Aubignac tells us this many years later in the dissertation he wrote attacking Corneille’s Œdipe (1663), and reports tartly that “il ne voulut pas suivre l’avis que j’avais ouvert”. 8 The “avis” in question might, or might not, be the same piece of advice that d’Aubignac mentions in the Pratique du théâtre (to which I shall return later). It concerns the death of Camille, but has nothing to do with its visibility on stage. 9 There is further evidence of Chapelain’s views in a letter to Balzac of 18 November 1640, detailing observations he had made to Corneille many months before. They explicitly concern the fifth act, but might have implications for the treatment of the death of Camille: “Dès l’année passée, je lui dis qu’il fallait changer son cinquième acte des Horaces, et lui dis par le menu comment; à quoi il avait toujours résisté depuis, quoi que tout le monde lui criât que sa fin était brutale et froide et qu’il en devait passer par mon avis” (quoted by Couton, I, 1537). Whatever the details of Chapelain’s advice, Corneille did not act on it. 7 The evidence is presented in the “Notice” of Couton’s edition (I, 1535-39). 8 Abbé d’Aubignac, Dissertations contre Corneille, ed. Nicholas Hammond and Michael Hawcroft (Exeter University Press, 1995), p. 75. 9 Abbé d’Aubignac, La Pratique du théâtre, ed. Hélène Baby (Paris: Champion, 2001), pp. 113-14. The Death of Camille in Corneille’s Horace 44 It is clear that Corneille knew that the ending of the play was subject to criticism, but it is not clear from the available evidence that any of this criticism was focussed on the visibility of Camille’s death on stage. For the purposes of following his argument in the “Examen”, we shall have to accept Corneille’s word that that was indeed the focus of some of the criticism. Corneille and the Actress Having suggested that his contemporaries blamed the visibility of Camille’s death on stage for the play’s allegedly weak ending, Corneille is keen to disculpate himself and to blame the actress. It is particularly interesting to observe his assumptions about an actress’s skills. We do not know what was indicated in the manuscript of the play that Corneille sold to the troupe of actors at the point at which Horace loses his patience with Camille for turning against him in his moment of triumph. Nor do we know what he might have told the actors during rehearsal, if anything. Certainly the words spoken by the characters say nothing explicit about a drawn sword or movement off stage. At one extreme, it is possible that Corneille’s manuscript gave no direction at this point other than what can be inferred from the characters’ words: Horace loses patience and threatens to kill Camille (“Va dedans les Enfers plaindre ton Curiace! ” (4.5, 1320)); he commits the deed when she shouts “Ah, traître! ” (1321); and he offers it as an example of summary justice: “Ainsi reçoive un châtiment soudain / Quiconque ose pleurer un ennemi romain! ” (1321-22). The next line is spoken by Procule; the printed page indicates that a new scene has begun; and Camille is not present. If this is all the actors had to go by and if there were no further directions in the manuscript and if Corneille did not attend a rehearsal to give instructions about how this should be performed, the words of the “Examen” make it clear that the dramatist still thinks the actress is to blame if spectators see the death on stage. He expects the actress to interpret and perform her role in a stereotypical way: women are naturally fearful; so as soon as Horace threatens violence, she should automatically take flight and so be off stage by the time the blow is struck. Some modern critical discussion seems to be based straightforwardly on the assumption that Camille’s death was indeed visible on stage. Merlin- Kajman says that Horace’s crime “se trouve théâtralement détaché, et comme exposé en tant que tel, dans sa nudité, aux yeux du spectateur” (“Ré-écriture”, p. 103). Forestier refers to it as “le meurtre délibéré et sous les yeux du public de Camille par Horace” (Essai de génétique théâtrale, p. 75). Gossip discusses “the staged killing of Camille” (“The Unity of Horace”, Michael Hawcroft 44 p. 347). This approach is all very well, as long as it is also admitted that Camille’s murder was only ever seen in this fashion against the express instructions of the dramatist, when the actress was performing badly (by the dramatist’s lights) and that Corneille takes pains to say that this is not what he wanted the spectators to see. So Corneille intended the spectators to imagine Camille receiving the blow off stage. And it is as if he is using the printed form of his play in 1660 to make this intention plain (in case it had not been plain earlier) when he adds to his explanation the phrase “comme je le marque dans cette impression”. Stage Directions and Early Editions Now, the phrase “comme je le marque dans cette impression” is problematic. It is sometimes understood to refer to an explanatory stage direction or stage directions, which might have been introduced in the 1660 edition of the dramatist’s plays, i.e. the edition in which the “Examen” first appeared. This is certainly Georges Couton’s interpretation of this phrase in his Pléiade edition of Corneille’s works: “En effet, à partir de 1660, Corneille a modifié ses indications scéniques à la fin de la scène v de l’acte IV, pour bien préciser que le coup mortel est donné à Camille dans la coulisse” (I, p. 1560, n. 2). Corneille seems to imply that for the 1660 edition he has decided to make it absolutely clear that the audience should not see Camille being struck on stage. Yet this was already made perfectly clear in every preceding edition of the play. Every preceding edition, including the first, contains the stage direction “blessée derrière le théâtre”, which accompanies Camille’s cry of “Ah, traître! ”. The blow has to be imagined as being inflicted off stage. Moreover, all preceding editions have the stage direction “revenant sur le théâtre” as Horace utters the final lines of the scene. It is impossible to imagine him having left the stage and returning on to it without also imagining that what he has done off stage is kill Camille. If we imagine him killing Camille on stage, the stage direction would then seem very bizarre, because there would be no need for him to leave the stage and then come straight back on. If both these crucial stage directions, which make Corneille’s intentions clear to readers of the play, were present not only in 1660, but also in all preceding editions of the play, did he actually make any modifications to the 1660 text, as the “Examen” seems to imply and as Couton seems to confirm? There is, in fact, only one uncontested variant, and that is in the stage direction that accompanies Horace’s expression of exasperation and The Death of Camille in Corneille’s Horace 44 threat of violence. Here, in 1660, Corneille inserts the conjunction “et” between “mettant la main à l’épée” and “poursuivant sa sœur qui s’enfuit”. I cannot think that the addition of the word “et” makes any difference to the envisaged movements on and off stage. There is one further, and more problematic, variant. Both Marty- Laveaux 10 and Couton say that that editions up to 1655 give the first part of this stage direction as “mettant l’épée à la main” as opposed to the later “mettant la main à l’épée”. It could be that there is a real difference in meaning between these two formulations. The earlier formulation “Mettant l’épée à la main” might be more graphic, more suggestive of imminent violence, and so more likely to mislead the reader (and indeed the actress, if the stage direction were present in the troupe’s manuscript) into envisaging violence on stage. The later variant, “mettant la main à l’épée”, figuring in the 1660 edition, might imply that Horace is only starting to draw the sword from its sheath as he speaks words of menace, and this might therefore be less likely to mislead readers as to what they might envisage being visible on stage. However, if Marty-Laveaux and Couton had inspected more copies of the earlier editions, they would have discovered that the supposedly later variant “mettant la main à l’épée” already existed in some copies of the first edition of the play published in 1641. 11 Since the only straightforwardly new variant to the stage directions in 1660 is the insertion of the word “et” into one of them, I am inclined to think that when Corneille says “comme je le marque dans cette impression” he is simply drawing the attention of readers of the “Examen” to the printed stage directions which clarify his intentions about the movements on and off stage. Pace Couton, Corneille is not signalling that there is any 10 Pierre Corneille, Œuvres, ed. C. Marty-Laveaux, 12 vols (Paris: Hachette, 1862-68), III, 340; Couton, I, 1569. 11 I have consulted 7 copies of the 1641 editions of the play (all published in Paris by Augustin Courbé). Most contain the wording ‘mettant l’épée à la main’: in the Bibliothèque Nationale de France, RES-YF-673 (in-4°), RES-YF-3904 (in-12°), RES- YF-3905 (in-12°), SMITH-LESOUEF-R-3515 (in-12°); in the Département des Arts du Spectacle on the Richelieu site of the BNF, 8-RF-1940(RES) (in-12°). Two contain the allegedly later variant “mettant la main à l’épée”: in the BNF, RES-YF- 635; and in the Département des Arts du Spectacle 8-RF-1939(RES), both in-4°. Emile Picot’s Bibliographie cornélienne (Paris: Auguste Fontaine, 1876) identifies three editions of Horace in 1641, two in-4° and one in-12°, but this is almost certainly an oversimplification in the light of my inspection of multiple copies. Picot does not identify individual copies consulted. The 1660 edition is very rare. I have consulted only two copies (Paris: Augustin Courbé and Guillaume de Luyne): in the BNF, RES-YF-2984-86; in the Département des Arts du Spectacle, 8-RF- 1701(RES, 1-3). Michael Hawcroft 4 significantly new stage direction in the 1660 edition. Even from the first editions, he has used the printed form of his play, in particular the stage directions, to make it clear to readers that Camille is struck off stage. But for whose benefit is Corneille being so attentive when he writes, and slightly modifies, his stage directions? If it is for that of an actress playing Camille, it is unlikely that she would have been reading any printed edition of the play. The original troupe would obviously have worked from a manuscript, which might or might not have contained these stage directions. We might speculate that Corneille’s manuscript did not have sufficiently explanatory stage directions, which is why he would take the trouble, in the “Examen”, of referring specially to his wording in the printed form of the play. It is true, however, that, once a play was in print, any troupe could perform it, and such troupes would almost certainly create their part roles by getting their copyist to copy out speeches from the printed form of the play. Indeed, in his third discourse, Corneille specifically envisages stage directions in the printed text as a benefit to provincial troupes: “l’impression met nos pièces entre les mains des comédiens, qui courent les provinces, que nous ne pouvons avertir que par là de ce qu’ils ont à faire, et qui feraient d’étranges contretemps, si nous ne leur aidions par ces notes” (III, 182). So perhaps Corneille uses stage directions in printed editions of his play in the hope that they would filter their way through to provincial actresses who might not otherwise perform with stereotypical feminine fear and run off stage at the sight of Horace’s sword. In attending to his stage directions, he might also be thinking of his readers and seeking to ensure that they have a clear picture of what he wants to be visible (or not) on stage. If he is thinking of readers generally, it must be in a peculiarly pedantic spirit of wanting to help them envisage a precise stage picture. If we remember that the point at issue is whether or not seeing Camille’s death on stage spoils the end of the play, we might think that it cannot make much difference to readers whether they envisage the sword striking Camille on stage or off stage, since readers see nothing in either case. So if the stage directions here are for the benefit of readers, their purpose is not to enhance readers’ aesthetic experience, but to defend the dramatist in the eyes of any readers who might have seen, or be ready to envisage, a performance in which Camille dies on stage and who might feel that that is somehow inappropriate. It could be, in fact, that Corneille’s stage directions were intended to defend him against specific critical readers; that the stage directions had sought to do this in all printed editions of the play; and that they were deliberately reinforced in the 1660 edition by the discussion in the “Examen” which draws attention to them. The Death of Camille in Corneille’s Horace 4 Violence on Stage in Theory Scherer’s gloss on Corneille’s stage directions here and on his discussion of them in the “Examen” will lead us helpfully to identify two key issues in mid-seventeenth-century dramatic theory to which the dramatist’s reflexions give rise: Dans Horace, Corneille a soin d’indiquer dès l’édition originale que Camille est “blessée derrière le théâtre” par son frère Horace, qui, “revenant sur le théâtre” après avoir infligé à sa sœur cette blessure mortelle, prononce encore quelques vers; l’attitude d’Horace, “mettant l’épée à la main”, comme le disent les premières éditions, impliquait que Camille prît la fuite, pour pouvoir être frappée “derrière le théâtre”; malgré ces précautions, l’actrice qui jouait Camille mourait en scène, et mécontentait ainsi les spectateurs attachés aux bienséances; Corneille doit donc renforcer ses indications scéniques; au lieu d’Horace “mettant l’épée à la main”, il nous montre, plus explicitement, Horace “mettant la main à l’épée, et poursuivant sa sœur qui s’enfuit” à partir de l’édition collective de 1655 [ … ] et dans l’Examen de 1660, Corneille revient sur la nécessité de “prendre la fuite et recevoir le coup derrière le théâtre” pour l’actrice qui joue le rôle de Camille. Les indications scéniques de tous ces auteurs montrent bien qu’on s’ingénie à ne pas “ensanglanter la scène” par des meurtres. 12 Leaving aside Scherer’s small inaccuracies in connection with the variants, I want to draw attention to his main argument, which is generally shared by other critics, but which seems to me an inaccurate account of Corneille’s concern in the “Examen” and a misleading presentation of seventeenthcentury dramatic theory. For Scherer, Corneille is at pains to ensure that the murder of Camille takes place off stage, because spectators, attached to the “bienséances”, would be displeased to witness it, the “bienséances” not allowing murders to be visible on stage. In his edition of Horace, Marc Escola expresses the same argument more succinctly: ‘Corneille s’assujettissait donc aux bienséances, en évitant d’“ensanglanter le théâtre”. Voir Examen” (p. 122, note h). These modern critical comments raise two distinct but related issues. One is: what is the value of the bienséances (or bienséance) as a critical notion likely to explain Corneille’s preoccupation in the “Examen” with the visibility of Camille’s death on stage? I argue elsewhere that in 1660 neither Corneille nor anybody else would have thought of bienséances as an appropriate concept in this context. 13 In 1660 it was a concept that occupied 12 Jacques Scherer, La Dramaturgie classique en France (Paris: Nizet, 1950), p. 418. 13 “The Bienséances and their Irrelevance”. See also Brunn and Karsenti “Pourquoi Horace”, pp. 206-09. Michael Hawcroft 45 theorists less than posterity has suggested; it was a strictly poetic concept; and it was entirely unrelated to the question of physical violence on stage. The other issue to which Scherer’s and Escola’s comments give rise is that of an alleged rule against the portrayal of violence on stage, or “bloodying the stage”. Much of Corneille’s argumentation in the first paragraph of the “Examen” is concerned with this alleged rule. His discussion is far from straightforward and calls for close scrutiny. Many modern critics tend to assume that there was indeed a rule against bloodying the stage in the seventeenth century. 14 If we look at the sentence in which Corneille himself uses the phrase, we see that he clearly does not want readers to think that there is any such prohibitive rule. His use of the conditional form “si c’est une règle de ne [ point ensanglanter le théâtre ] ” certainly suggests that Corneille has doubts about whether or not there is such a rule. His main clause increases these doubts by citing Aristotle as an authority in favour of depicting murder on stage. In fact the bulk of this paragraph deals with the question of the representation of violence on stage in such a way as to imply that Corneille is sceptical about there being any rule against it, whilst also allowing him to reserve judgement. In strictly logical terms, this might strike readers of the “Examen” as an odd development. His previous sentence had implied that he accepted that it was wrong for Camille’s death to be visible on stage, that this was not what he had wanted, and that he had ensured that stage 14 “Ces situations de violence extrême [ … ] n’ont pas droit de cité sur la scène” (Bénédicte Louvat, Poétique de la tragédie classique (Paris: SEDES, 1997), p. 80); “The bienséances also rules out the depiction on stage of all forms of violent action such as duels, battles, and murders” (John Lough, Seventeenth-Century French Drama: The Background (Oxford University Press, 1979), p. 112). Pierre Pasquier envisages Camille’s murder on stage as “transgressant l’un des interdits de la dramaturgie classique” in “Les Apartés d’Icare. Eléments pour une théorie de la convention classique”, Littératures classiques, 16 (1992), 79-101 (p. 80). Such claims are endlessly repeated. For a revisionist view, based on scrutiny of a wide range of theoretical texts, see Emmanuelle Hénin, “Faut-il ensanglanter la scène? Les enjeux d’une controverse classique”, Littératures classiques, 67 (2009), 13-32. Her account needs to be treated with some caution, however: partly because of the uncritical importance she attaches to the notion of the bienséances (it is not helpful, for instance, to claim that “le moment des bienséances coïncide avec la Querelle du Cid” (p. 21), when bienséance had only been used as a critical concept by Chapelain in the singular and was not to be used in the plural until 1639 by La Mesnardière, and in neither case in connection with violence on stage); and partly because her discussion of the murder of Camille is marred by bibilographical error (she says that Corneille has introduced new clarificatory stage directions into the 1660 edition, which is simply not the case (p. 21 and note 41)). The Death of Camille in Corneille’s Horace 45 directions in the printed version of his play made it clear that he did not wish her death to be visible on stage. We might have expected that Corneille was saying all this because he believed that it is in principle wrong to depict death on stage; but he now confounds us by suggesting that he is not persuaded that there is any ban on depicting death on stage. This should make us wonder why, in that case, he is so insistent that Camille’s death should not be visible. But he makes us wait; and by the end of his first paragraph we shall have to work out an answer to that question for ourselves. If Corneille uses the doubting formulation “si c’est une règle de ne [ point ensanglanter le théâtre ] ”, it is probably as a riposte to his hostile rival Scudéry, who had so virulently attacked Le Cid. In his Observations sur le Cid Scudéry had used the phrase “cette autre regle qui deffend d’ensanglanter le Theatre”. 15 It would no doubt have given Corneille great pleasure to quote Scudéry’s words back to him and to follow them immediately with the contradictory recommendation of Aristotle. But things are not entirely straightforward. According to Corneille, Aristotle “nous apprend que pour émouvoir puissamment il faut de grands déplaisirs, des blessures et des morts en spectacle”. Corneille is referring here to the third of three possible parts of a plot as identified by Aristotle: (1) peripeteia, (2) recognition, (3) pathetic event. This is a modern French translation of Aristotle’s account of the pathetic event: “quant à l’événement pathétique, c’est une action qui provoque destruction ou douleur, comme les agonies présentées sur la scène, les douleurs très vives, les blessures et toutes les choses du même genre” (my italics). 16 One reason why Corneille might have chosen to reserve his judgement overall in this passage, whilst wanting to maintain some scepticism about there being a rule, is that he probably knew that there was no clear agreement over the precise meaning of Aristotle’s phrase “en to phanero”, translated above as “présentées sur la scène”. Castelvetro, for instance, did not believe that Aristotle’s phrase referred to acts of violence being performed on stage. He thought that Aristotle was referring to other ways in which violent actions could be made manifest: bringing a corpse on stage; shrieks off stage; or even a messenger speech. 17 15 Jean-Marc Civardi, La Querelle du Cid (1637-1638) (Paris: Champion, 2004), p. 385. 16 Aristote, Poétique, trans. and ed. Michel Magnien (Paris: Le Livre de Poche, 1990), XI, 1452 b, p. 102. 17 Castelvetro, Poetica d’Aristotele vulgarizzata e sposta, 2 vols, ed. Werther Romani (Rome, Bari: Laterza, 1978-79), I, pp. 337-39) (cited by Hénin, “Faut-il ensanglanter”, pp. 15-16). Michael Hawcroft 45 Moreover, Corneille also knew that Horace, in the Ars Poetica, appears to say the opposite of Aristotle. Indeed Corneille seems to concede this point in his next sentence, giving one of Horace’s key examples, that of Medea murdering her children. The line “Ne pueros coram populo Medea trucidet” is much quoted and glossed (in the early modern period as well as among modern critics) as Horace’s alleged opposition to the depiction of violence on stage. But it is crucial to consider Horace’s line in its proper context. Just as Aristotle’s apparent sanctioning of on-stage violence is not absolutely clear, neither is Horace’s apparent interdiction. Here are all the relevant lines from Horace (lines 179-88): Aut agitur res in scaenis aut acta refertur. Segnius inritant animos demissa per aurem quam quae sunt oculis subiecta fidelibus et quae ipse sibi tradit spectator; non tamen intus digna geri promes in scaenam multaque tolles ex oculis, quae mox narret facundia praesens. Ne pueros coram populo Medea trucidet, aut humana palam coquat exta nefarius Atreus, aut in auem Procne uertatur, Cadmus in anguem. Quodcumque ostendis mihi sic, incredulus odi. 18 (Either the action happens on stage or it is narrated. We are less moved by what we hear than by what we witness with our own trustworthy eyes. None the less you will not show on stage things that might better happen off stage and you will keep many things from our gaze, which an eye witness can then narrate eloquently. Do not let Medea slaughter her children in public, do not let the wicked Atreus cook human flesh in front of the audience, do not let Procne turn into a bird, or Cadmus into a snake. My disbelief will make me reject anything you show me of that nature. (my translation)) In conceding that Horace does not want a dramatist to show, on stage, Medea’s slaughter of her children, Corneille explains that this is a suspicion about depicting “des événements trop dénaturés”. But he does not, here, explain Horace’s thinking further than that. This might be because he has already done so in his second discourse, first published in the same volume of the 1660 edition as the “Examen” of Horace: “Horace ne veut pas que Médée tue ses enfants, ni qu’Atrée fasse rôtir ceux de Thyeste à la vue du peuple. L’horreur de ces actions engendre une répugnance à les croire, aussi bien que la métamorphose de Progné en oiseau et de Cadmus en serpent, dont la représentation presque impossible excite la même incrédulité, quand 18 Horace, Epîtres, trans. and ed. François Villeneuve (Paris: Les Belles Lettres, 1934), p. 212 (Pour l’Art poétique, pp. 202-26). The Death of Camille in Corneille’s Horace 45 on la hasarde aux yeux du spectateur: Quaecumque ostendis mihi sic, incredulus odi” (III, 159-60, Corneille’s italics). Corneille is clear that the objection to the depiction of certain acts of violence is that they cannot be performed without endangering the spectators’ suspension of disbelief. The fact that he mentions all Horace’s examples (and not simply that of Medea) makes his interpretation even more transparent. Horace says that Procne’s metamorphosis into a swallow and Cadmus’s into a snake should not be performed on stage for the same reason that Medea’s and Atreus’s bloody crimes should be kept off stage. The audience would not believe what they were seeing. In short, Horace can be interpreted - and Corneille clearly saw this - as condemning, not on-stage violence as such, but any on-stage action that is difficult to perform credibly. It is this same argument that leads Castelvetro to oppose the depiction of on-stage violence in particular. He goes further than Horace in spelling out the inappropriate reaction such depictions can provoke in the audience: “l’esperienza ha mostrato che simili crudeltà e orribilità non si possono verisimilmente far vedere in atto, e che fanno anzi ridere che piangere, e che producono non effeto di tragedie, ma di comedia” (Poetica d’Aristotele, I, 379). The illusion broken, the audience risks laughing at the discrepancy between what they see and what it is supposed to represent. This claim by Castelvetro is quoted by La Mesnardière, 19 who generally disagrees with his Italian predecessor. Indeed La Mesnardière is inclined to disagree with Castelvetro on this very point and to adopt a more Aristotelian view. If, he argues, the dramatist’s purpose is to maximize the arousal of pity and fear, “il luy seroit plus facile de tirer de la Pitié d’une large effusion de sang, & de provoquer la Terreur par la montre des coupables qui endureroient les supplices, que d’exciter ces sentimens par la seule relation des maux qu’ils viendroient de souffrir” (p. 201). But he goes on to develop a more nuanced position, based on the identification of three different types of death. Certain deaths, like the dismemberment of Hippolytus, cannot be shown on stage, because they would put the actor’s life in peril (p. 202). Horrible deaths like Medea’s murder of her children should not be shown for two different reasons. One reason echoes Castelvetro’s. They might be laughable in performance: “il est tres-difficile d’imiter ces bourrelleries sans que la feinte en soit grossiere, & par consequent ridicule” (p. 205). The other reason is that they might evoke the wrong emotion in the audience, horror, or as La Mesnardière puts it, “un transissement odieux & une horreur desagreable, qui surmontent infiniment la Terreur & la Pitié qui doivent regner l’une et l’autre, & toutes deux s’il est possible, dans la parfaite 19 La Mesnardière, La Poétique (Paris: Antoine de Sommaville, 1640; Geneva: Slatkine Reprints, 1972), p. 200. Michael Hawcroft 45 Tragedie” (p. 204). His third type, generous deaths, were, however, very acceptable on stage as far as La Mesnardière was concerned. He thinks particularly of the suicides of characters who have done wrong and who express repentence by their action. Such characters and their suicides would be very suitable, in his view, for the arousal of pity and fear (p. 206). That said, La Mesnardière acknowledges that ancient dramatists showed a much wider range of deaths on stage, and he glosses Horace’s line about not showing Medea committing murder on stage in a relatively permissive way: “C’est le sens de ce Précepte, qui n’abolit pas tous les meurtres, mais qui bannit de la Scene tous les meurtres odieux” (p. 210). Even if La Mesnardière reads a moral preoccupation into the Horatian tag, he does not interpret Horace’s lines as being as universally prohibitive as some modern critics do. Taking together the glosses of Castelvetro, Corneille and La Mesnardière, we can see that there was an awareness in the early modern period that Aristotle and Horace appeared to have said contradictory things and that what kind of violence could be shown on stage was a matter for debate, the criteria being linked to the illusion of reality and the emotional response of the audience. If “ensanglanter le théâtre / la scène” was a phrase that seventeenth-century French theorists and dramatists had recourse to, no clear and unanimous definition was ever attached to it. Nobody was ever clear what, if anything, was actually forbidden or why. 20 There is another reason why Corneille, in the “Examen”, might not want to repeat his own explanation of the Latin writer’s lines (already given in the second discourse): his purpose in the “Examen” is not an exegesis of Horace, but an allusion to Horace that will be sufficient to defend his own dramatic practice in Horace. He wants to show that he knows what Horace said so that he can then claim with authority that he also knows that Horace did not issue a blanket ban on the depiction of death on stage: “je ne vois pas qu’il en fasse une règle générale pour toutes sortes de morts”. Corneille was on safe ground in making this claim, which is not inconsistent with La Mesnardière’s reading of Horace. But then Corneille goes further. He draws a distinction between Medea’s murder of her children and Horace’s murder of Camille in such a way as to imply that Camille’s murder would not fall into the category of those that should not be shown on stage: “ [ je ne vois pas ] que l’emportement d’un homme passionné pour sa patrie, contre une sœur qui la maudit en sa pré- 20 Or as Emmanuelle Hénin puts it, “tout en défendant inlassablement d’“ensanglanter la scène”, ils évitent soigneusement de préciser ce qui est interdit” (p. 19). Her first phrase seems to me to take too little account of the scepticism of a writer like Corneille. The Death of Camille in Corneille’s Horace 45 sence avec des imprécations horribles, soit de même nature que la cruauté de cette mère”. Yet again, we might be puzzled by Corneille’s line of argument. On the one hand he says that he has taken pains not to allow the murder of Camille to be visible on stage; on the other, he refers to Aristotle in terms which would support the representation of her death on stage, and he now refers to Horace’s famous interdiction of Medea’s murderous act in such a way as to exclude Camille’s murder from its terms and as to imply that it would be perfectly acceptable for her death to be visible on stage. Violence on the Ancient Stage Corneille’s intellectual game with his readers goes on. He seems to pursue an argument about the acceptability of seeing Camille’s death on stage by adducing three examples of violent on-stage deaths to be found in ancient plays. The first is particularly forceful, because it is Seneca’s depiction in his Medea of the heroine slaughtering her children on stage, even though this is precisely an example of violent action that Horace would have banned. And even Corneille in the second discourse had seemed to endorse Horace’s view that such a murder could not have been performed without straining the audience’s belief in the theatrical illusion. 21 His second example is of Ajax’s on-stage suicide in Sophocles’ Ajax. There is a technical problem with this example, which he addresses in the third discourse in connection with the unity of place: “Sophocle toutefois ne l’a pas observée dans son Ajax, qui sort du théâtre afin de trouver un lieu écarté pour se tuer, et s’y tue à la vue du peuple; ce qui fait juger aisément que celui où il se tue n’est pas le même que celui d’où on l’a vu sortir, puisqu’il n’en est sorti, que pour en choisir un autre” (III, 188). How the suicide of Ajax was represented on the Greek stage was a matter of controversy amongst early modern critics, and is still controversial today. 22 None the less, in presenting it as an on-stage death, Corneille would have had the support of the likes of Heinsius (“alia [fieri] possunt [...] ut Ajacem sibi manus inferre” (“other things can be shown like Ajax taking a sword to himself”). 23 Corneille develops his third example more fully, that of Clytemnestra’s death in the two Electra plays by Sophocles and Euripides; and he relates 21 In his own play Médée, the heroine kills the children off stage before appearing on stage to boast about it (see Act 5, scene 5). 22 See Hénin, “Faut-il ensanglanter”, pp. 16-19, for a summary of the different views. 23 Daniel Heinsius, De Constitutione Tragoediae, ed. Anne Duprat (Geneva: Droz, 2001), p. 239. Michael Hawcroft 45 this example more directly to the case in hand by discussing (without mentioning him by name) d’Aubignac’s comments in the Pratique du théâtre on the death of Camille. Corneille’s discussion of Clytemnestra in the “Examen” only makes sense in the light of his fuller discussion in the second discourse, to which he makes explicit reference. Clytemnestra is no doubt invoked as a third example to support Corneille’s case about the presence of on-stage death in ancient tragedy. But Corneille’s account of Clytemnestra’s death and its implications for his own argument about the death of Camille in Horace suddenly become very complex. Corneille discusses Clytemnestra’s on-stage death at two points in the second discourse, both times to disapprove of it, but for a different reason on each occasion. His first reason is to do with performance and credibility. He thinks it such a cruel murder that it is difficult to perform it credibly: “Il faut examiner en même temps si elle n’est point si cruelle, ou si difficile à représenter, qu’elle puisse diminuer quelque chose de la croyance que l’auditeur doit à l’histoire [ … ] . Lorsque cet inconvénient est à craindre, il est bon de cacher l’événement à la vue, et de faire savoir par un récit qui frappe moins que le spectacle et nous impose plus aisément” (III, 159). This leads Corneille directly into his discussion of Horace’s views on Medea and the depiction of incredible events on stage, and Corneille’s views here tally with those he expresses in the “Examen”: that it is possible to depict on-stage death as long as it does not damage the illusion of reality. Shortly afterwards Corneille returns to the death of Clytemnestra as depicted by Sophocles and advances a second reason for disapproving of it. He appeals to a notion that Chapelain had used to criticize Chimène in Le Cid: she lacks consistency, because, having been introduced as virtuous, she too promptly agrees to marry her father’s murderer. The desired kind of character consistency is what Chapelain called bienséance. 24 Corneille does not use the term bienséance, but his anxiety concerns precisely a sudden inconsistency in the presentation of the character of Electra. Electra is suddenly too inhumane and Orestes too cruelly vengeful for the normal emotions of pity and fear to be aroused: “Je ne puis souffrir chez Sophocle que ce fils la poignarde de dessein formé, cependant qu’elle est à genoux devant lui, et le conjure de lui laisser la vie. Je ne puis même pardonner à Electre, qui passe pour une vertueuse opprimée dans le reste de la pièce, l’inhumanité dont elle encourage son frère à ce parricide. C’est un fils qui venge son père, mais c’est sur sa mère qu’il le venge” (III, 161). The overall 24 “La bienséance des mœurs d’une fille introduite vertueuse n’y est pas gardée par le poète lorsqu’elle se résout à épouser celui qui a tué son père” (Chapelain, Opuscules critiques, ed. Anne Duprat (Geneva: Droz, 2007), p. 288). For a fuller discussion see my “The Bienséances and their Irrelevance”. The Death of Camille in Corneille’s Horace 45 context of Corneille’s discussion of this example is whether or not a dramatist can change a well-known event in a story, in this case Clytemnestra’s murder by her son. Corneille’s view is that the well-known event has to be retained, but it can be presented in such a way as to make it more appropriate for performance on stage. Accordingly he ventures to suggest a re-writing of the ancient text which would diminish the cruelty of Orestes by turning the murder of his mother into an accident that occurs whilst he is attempting to murder her adulterous lover Aegisthe: “Pour rectifier ce sujet à notre mode, il faudrait qu’Oreste n’eût dessein que contre Egisthe, qu’un reste de tendresse respectueuse pour sa mère lui en fît remettre la punition aux Dieux, que cette reine s’opiniâtrât à la protection de son adultère, et qu’elle se mît entre son fils et lui si malheureusement, qu’elle reçût le coup que ce prince voudrait porter à cet assassin de son père. Ainsi elle mourrait de la main de son fils, comme le veut Aristote, sans que la barbarie d’Oreste nous fît horreur, comme dans Sophocle, ni que son action méritât des Furies vengeresses pour le tourmenter, puisqu’il demeurerait innocent” (III, 161). Sophocles’ presentation is likely, in Corneille’s view, to arouse horror rather than pity or fear. Corneille clearly agrees with La Mesnardière that there is a kind of horror that distracts from the appropriate emotional impact of tragedy. 25 Corneille rewrites Sophocles and d’Aubignac rewrites Corneille It is important to bear all this discussion of Clytemnestra in mind when reading the “Examen” of Horace. Corneille’s argument swerves from being an apparently straightforward demonstration of the evidence of ancient dramatists depicting death on stage to a consideration of the limits of the appropriateness of such depictions. He is therefore approaching his topic indirectly, but with a clear purpose in mind. That purpose is to refute d’Aubignac, who had ventured to suggest that Horace’s murder of Camille had breached the limits of appropriateness; d’Aubignac had even proposed a reworking of the scene both at the time of the first performance in 1640 and again in the Pratique du théâtre in 1657. Again without naming him, Corneille wishes to make clear in what way d’Aubignac’s proposed reworking 25 Corneille’s discussion of Clytemnestra’s visible death is based on a misunderstanding about the performance conditions in the Greek theatre. Her death would not have been visible, though it would have been heard as if it had been taking place in the skene, the building on stage supposed to represent the royal palace. Corneille’s mistake is a common one for the time. See also La Mesnardière: “Clytemnestre est poignardée sur la Scene” (p. 208). Michael Hawcroft 4 of the scene would be unsuccessful by contrasting it with his own proposed reworking of the Sophoclean scene. This is the sense of Corneille’s very allusive sentence: “L’adoucissement que j’apporte dans le second de ces discours pour rectifier la mort de Clytemnestre ne peut être propre ici à celle de Camille”. This sentence is the key to understanding the drift of Corneille’s argument, containing as it does a subtle but significant allusion to d’Aubignac. It is important to note that Corneille’s suggested re-working (“adoucissement”) of Sophocles leaves the stage every bit as bloody as Sophocles’ version does. Corneille’s concern with Sophocles is not about the presence of murder on stage; it is about the manner of the murder, given the characters involved, and the likely audience response, which Corneille wants to make appropriately tragic. Although he is acutely keen to distinguish between his proposed reworking of Sophocles and d’Aubignac’s proposed reworking of Camille’s murder in Horace, what both reworked versions have in common is that they leave death on the stage and they are both motivated by a desire to channel the audience’s emotions in ways deemed to be more appropriate. Corneille goes on to evoke, without approval, a proposed alternative treatment of Camille’s death, and it is precisely the one which d’Aubignac suggests in the Pratique: Horace would draw his sword, but before he could take any further action, Camille would thrust herself onto it in an act of suicide caused by despair. Corneille does not bother to spell out what the unnamed proponent of this version sees as its advantage. For that we need to turn to the Pratique. This, as it happens, also leads us to one of d’Aubignac’s rare uses of the word bienséance in the Pratique, and the only time in any critical writing of the period that the word bienséance is used in connection with this scene (bienséances (in the plural) is not used by d’Aubignac and is never, in the period, used in connection with this scene). D’Aubignac’s use of the word is entirely poetic, and not at all connected with the fact of violence on stage. It is important to grasp the general critical context in which d’Aubignac discusses the scene: “La Scène ne donne point les choses comme elles ont été, mais comme elles devaient être [ … ] C’est pourquoi la mort de Camille par la main d’Horace son frère, n’a pas été approuvée au Théâtre, bien que ce soit une aventure véritable, et j’avais été d’avis, pour sauver en quelque sorte l’Histoire, et tout ensemble la bienséance de la Scène, que cette fille désespérée voyant son frère l’épée à la main, se fût précipitée dessus: ainsi elle fût morte de la main d’Horace, et lui eût été digne de compassion, comme un malheureux Innocent, l’Histoire et le Théâtre auraient été d’accord” (pp. 113-14). D’Aubignac’s topic is whether dramatists should depict the truth or a modified version of the truth. In order to understand The Death of Camille in Corneille’s Horace 4 his use of the term “bienséance”, we need properly to assess the sigificance of the tense in the phrase “comme elles devaient être”. D’Aubignac is not saying that the dramatist should modify the truth in order to create a more morally acceptable version of events for the audience, a version that would have been the truth if only the world were a better place. He believes the dramatist should modify the truth when a modified version would better suit the requirements of theatrical performance and better help the dramatist to achieve the desired impact on the audience. It is true that Horace killed his sister. But d’Aubignac clearly feels that this act, in the play, diminishes the pity which the dramatist should be arousing for his main character and which is a proper tragic emotion. Hence his proposed rewriting, which aims to reconcile truth (“histoire”) and dramatic requirements (“bienséance”). Camille would still die on Horace’s sword (so the truth would thereby be acknowledged); but Horace’s responsibility would be diminished, because he would not actually be striking a blow. In this way he would emerge as a more successful tragic character “digne de compassion, comme un malheureux Innocent”. So when D’Aubignac is concerned with the “bienséance” of this scene, it is not to do with whether or not Camille dies on stage. D’Aubignac’s version would show her death on stage. His concern is the same as that of Chapelain for Chimène and Corneille for Electra. He is concerned with the consistent presentation of a character, Horace, who, in order to arouse tragic emotions, cannot, in his view, suddenly switch from triumphant patriot to murderer of his sister without forfeiting the audience’s sympathy. If Corneille is pleased with his own re-writing of Sophocles, why is he not willing to accept d’Aubignac’s re-writing of his own scene, prompted by the same motivation of channelling the audience’s emotional response? The lines in the “Examen” adumbrate Corneille’s objections (still without mentioning d’Aubignac by name). First, Horace would not appear to the audience as an unfortunate innocent, because he would still have drawn his sword against his sister and so would still appear criminal. By contrast, in Corneille’s proposed reworking of Sophocles, Orestes’ guilt would genuinely be diminished because he would be drawing his sword against Aegisthus and would only accidentally kill Clytemnestra. Secondly (and this is the major recurrent disagreement between Corneille and d’Aubignac), the historical record says that Horatius ran his sword through his sister (“transfigit puellam” according to Livy, as quoted by Corneille in certain editions (I, 837)) and this is too well known to be changed; the dramatist has to work with it. Corneille’s rewritten version of Sophocles, on the other hand, retains the image of Orestes running his sword through his mother, so Corneille is not suggesting any modification to that famous truth. And Michael Hawcroft 46 thirdly, in d’Aubignac’s rewritten version, there would also be major consequences for the trial of Horace in Act 5 and for the father’s defence of his son: that too is historically true, but would be imperilled if Horace were not actually to run a sword through Camille. So Corneille rejects d’Aubignac’s proposed reworking as an unsatisfactory attempt to deal with what the dramatist still acknowledges is a problem in his play: the death of Camille. Corneille’s final sentence sums up his first paragraph and anticipates the next ones. How? He had started by admitting that the death of Camille spoils the end of the play. He had reported that people generally thought that this was because Camille’s death was visible on stage. He then said that it was never his intention that the murder should be visible on stage, whilst, in any case, making it clear that he thinks death and murders may sometimes be shown on stage. At the end of the pararaph his argumentation might seem abruptly cavalier: “Quoi qu’il en soit, voyons si cette action n’a pu causer la chute de ce poème que par là [ i.e. by the accidental visibility of her death on stage ] , et si elle n’a point d’autre irrégularité que de blesser les yeux”. With the phrase “quoi qu’il en soit”, it is as if Corneille cannot be bothered to pursue further the issues to do with the visibility of Camille’s death on stage. And the phrase “blesser les yeux” is, I think, ironic, given that he clearly has no objection in principle to the visibility of death on stage. But it is as if he is content to have proved that death can be acceptable on stage and that if, therefore, by accident, spectators were to see Camille die on stage, that could not be the cause of the admitted weakness of the play’s ending. So he wishes now to move on to explore other reasons why Camille’s death proved problematic. This is what he goes on to do; and he does of course attach the blame to the lack of preparation, to the introduction of a second peril (when one peril would have been enough), and to the unacceptable change in importance of the character of Camille. The Problem of Camille’s Death and Corneille’s Incomplete Argument In moving on to these other arguments, Corneille has left one question hanging. Why, given that there are circumstances when death is clearly acceptable on stage, did he take pains to ensure that the audience should not see the moment of Camille’s death? and why does he not explain his motivation? It is easy to imagine why he does not explain his motivation. The point is not to his purpose, which is to explore what he thinks are the problems caused by Camille’s death in the dramatic structure of the play, having first of all swept aside the view of others that this might have something to do with the visibility of her death on stage. There is another The Death of Camille in Corneille’s Horace 46 reason why he might not have wished to explore his motivation for trying to keep her death off stage: to do so might have involved giving some credit to d’Aubignac. The issue that d’Aubignac identifies in connection with the death of Camille is that of the extent of Horace’s guilt and the need to minimize his guilt so that the audience can respond to him with appropriate tragic sympathy as an “innocent malheureux”. Hence d’Aubignac’s suggestion that Camille should impale herself on his sword. I think it is possible to deduce that Corneille had already envisaged a similar problem to do with the audience’s emotional response to Horace. His reasoning is found not in the “Examen” of Horace, but in the closing lines of the “Examen” of Le Cid, his immediately preceding play. He closes his comments on Le Cid with a reminder of Horace’s view in the Ars Poetica, when he expresses his preference for keeping Medea’s slaughter of her children off stage, that “ce qu’on expose à la vue touche bien plus que ce qu’on n’apprend que par un récit” (I, 706). 26 Corneille clearly thinks that the audience’s emotional reaction to what it sees is greater than to what it hears, and he explains how this belief determined his presentation, in Le Cid, of two moments of physical violence: “C’est sur quoi [ i.e. Horace’s precept ] je me suis fondé pour faire voir le soufflet que reçoit Don Diègue, et cacher aux yeux la mort du Comte, afin d’acquérir et conserver à mon premier Acteur l’amitié des Auditeurs, si nécessaire pour réussir au Théâtre” (I, 706). Spectators, seeing the count strike the old man, will be full of sympathy for Don Diègue and for Rodrigue who defends him. Meanwhile, ensuring that Rodrigue’s murder of the Count is not performed on stage will prevent the audience from feeling sympathy for the Count and aversion for Rodrigue. It is possible to speculate that exactly the same consideration about the emotional power of what is seen led Corneille to want to keep the moment of Camille’s murder off stage (whilst insisting on the option to represent any other physical violence that suited his emotional aims as a dramatist). Corneille therefore had anticipated the very problem that strikes d’Aubignac when he sees a bad performance in which the actress allows herself to be struck in full view of the audience. But believing that a well-known historical event has to be preserved by the dramatist, Corneille had already anticipated a more satisfactory solution than d’Aubignac’s: reducing the negative feelings that the murder might arouse in the audience towards Horace by ensuring that the blow is struck out of sight. It is only a solution, however, if the actors perform according to the dramatist’s wishes. 26 The allusion is to lines 180-81 of the Ars Poetica, already quoted above. Michael Hawcroft 46 Conclusion The opening paragraph of the “Examen” of Horace is rich, complex, and allusive. To interpret it, we have to consider Aristotle, Horace, the Greek and Roman tragedians, Castelvetro and Heinsius, the French theorists of that crucial period of dramatic fermentation 1630-60, as well as their modern commentators, whose accounts need to be treated with caution when, with retrospective knowledge, they attach a specious synthetic coherence to concepts that at the time were tentative and unformed. In accounting for Corneille’s insistence that Camille should be struck off stage and that his stage directions make this clear, we have had to verify and correct the frequently misrepresented evidence of the presence of stage directions for Camille’s murder in early editions of the play. Corneille had in fact always, in all his editions, included stage directions that made his wishes about Camille’s death clear (his words in the “Examen” have been misinterpreted by modern readers as implying new stage directions in the 1660 edition). Moreover, modern commentators’ claims that Corneille wanted to keep Camille’s death off stage in order to respect the bienséances are undermined both by a survey of the precise and limited senses attached to the term (in either its singular or plural forms) in theoretical writing of the period up to 1660 27 and by the fact that there was no straightforward prohibition against depicting violence on stage. Indeed the “Examen” makes it clear that Corneille reserved the right to depict death on stage, and suggests that the limits imposed on depicting physical violence should be related on the one hand to the practicalities of performance (can it be performed credibly? ) and on the other the need to manage the audience’s tragic response (will the violence enhance or endanger the arousal of pity and fear? ). The accidental visibility, in at least some performances, of Camille’s death on stage is a preliminary aside in Corneille’s overall argument in the “Examen”, but the issues raised by his discussion of it take us to the heart of the tragic dramatist’s creative task: how to deal with wellknown but intractable material and how to present character and action so as to maximize the audience’s feelings of pity or fear. 27 See my “The Bienséances and their Irrelevance”. I am grateful to Emma Herdman and David Maskell for their comments on a draft of this article.