[In the following essay, Forestier discusses Racine's conception of characterization for the stage, focusing on his innovative contribution to the portrayal of tragic heroes.]

It is evident from a reading of Subligny's La Folle querelle ou la Critique d'Andromaque (1668) that the debate surrounding Andromaque immediately provoked contention regarding whether or not Racine was worthy of being considered Corneille's equal.¹ As far as is known, it was never seriously considered that Quinault, despite his success, should even be compared to Corneille. In the case of Racine, however, the public genuinely felt that the young author had opened up a new path between Cornelian tragedy and la tragédie galante. Racine's new form, based on the tragedy of passionate love, was superior to la tragédie galante and could reasonably be compared to Cornelian tragedy. It is this aspect of passionate love that the first part of the preface to Andromaque (the extract from the Aeneid) set out to defend, and it does so by evoking the highest authority, Virgil. However, the second part of the preface, i.e. the preface proper, written by Racine, is entirely and exclusively devoted to the manner in which tragic roles are recreated for the stage, proving that contemporaries had clearly recognized that Andromaque's revolutionary quality also lay in its conception of characterisation.

Twentieth-century criticism of Racine, however, has generally concentrated on the preface purely as a basis for analysing and examining Racine's insincerity and his adroit deflection of charges against him by discussing issues that may not even have been controversial. Consequently, it is felt that Racine pretends to ignore the main reproach against him, namely that Pyrrhus's conduct towards Andromaque is no different to that of a courteous hero of a fashionable novel of the time. Instead, Racine deals with a niggling objection, one that may indeed never have been raised, and defends himself for having portrayed Pyrrhus as a little too brutal. Whence the argument judged to be supremely polemical:

J'avoue qu'il n'est pas assez résigné à la volonté de sa Maîtresse, et que Céladon a mieux connu que lui le parfait Amour. Mais que faire? Pyrrhus n'avait pas lu nos Romans. Il était violent de son naturel. Et tous les héros ne sont pas faits pour être des Céladons.

[I must admit that he is not sufficiently submissive to the wishes of his Mistress, and that Celadon was better versed than [Pyrrhus] in Perfect Love. But what can be done about it? Pyrrhus had not read our [French] Novels. He was of a naturally violent temperament. And not all heroes are cut out to be Céladons.]²

To which it is easy for one of the characters of La Folle Querelle to retort: ‘Je lui soutiens, moi, que Pyrrhus avait lu la Clélie’ [But I insist that Pyrrhus had read Clélie].³ It is true that Racine was accused of having portrayed a Pyrrhus for whom, ‘l'amour est l'âme de toutes ses actions’ [love is the motive for all his actions], a quality judged to be ‘indigne des grands caractères’ [unworthy of great characters].⁴ But he was also accused of the opposite, as the preface to La Folle Querelle illustrates:

M. Corneille, dis-je, […] aurait conservé le caractère violent et farouche de Pyrrhus, sans qu'il cessât d'être honnête homme, parce qu'on peut être honnête homme dans toutes sortes de tempéraments: et donnant moins d'horreur qu'il ne donne des faiblesses de ce Prince qui sont de pures lâchetés, il aurait empêché le spectateur de désirer qu'Hermione en fût vengée, au lieu de le craindre pour lui.

[I maintain that M. Corneille […] would have preserved the wild and violent character of Pyrrhus, without him failing to remain an ‘honnête homme’ because one can remain an ‘honnête homme’ with different kinds of temperament; and by giving less horror to the role than he accords to the weaknesses of this Prince, which are morally reprehensible, he would have avoided the spectators' desire to see Hermione get her own back on him, instead of fearing for his safety.]⁵

Thus it is evident that Pyrrhus was accused of being excessively ‘violent et farouche’ [violent and wild], qualities entirely inappropriate for a king, who must remain an ‘honnête homme’.

The first question raised by these contradictory criticisms concerns their interpretation. Can Pyrrhus be at once too galant and too violent? In effect, the contradiction is invaluable, as it prompts us to reflect upon Racine's work and to understand his innovative contribution to the portrayal of tragic heroes. Secondly, why does his argument deal only with one of the two criticisms levelled at him? Is it simply the manœuvre of a good polemicist? In fact, the question of whether it is necessary or not, in Racine's words, to transform ‘tous les héros de l'Antiquité pour en faire des Héros parfaits’ [all the heroes of Antiquity to make shining Heroes of them all],⁶ touches on one of the essential points of the poetics of tragedy: the relationship between bienséance (propriety) and the ressemblance (verisimilitude) of characters, a point which had proved to be an encumbrance for Corneille, but which Racine completely revolutionizes. In this way, the two issues become so closely linked that they form a single question.

In the seventeenth century, bringing a tragic role to the stage meant adhering to a complete series of criteria relative to his/her caractère (or what was referred to in the seventeenth century as his/her mœurs). These criteria had been defined by Aristotle and had been unanimously espoused by dramatists of the modern age, including Corneille.⁷ Based on the rhetorical typology of caractère, this series of specifications had proved itself over two thousand years, in judicial prosecutions as well as in the major literary genres, and none would have thought to question it. Three of these conventions were considered essential: convenance or bienséance, ressemblance and constance (consistency).⁸

Firstly, a tragic role must be bienséant, that is to say it had to conform to the human or social type to which it belonged. In chapter eight of his Poétique (1639), La Mesnardière wrote that it was necessary to

faire les Héros généreux, les Philosophes prudents, les Femmes douces et modestes, les Filles pleines de pudeur, les Ambassadeurs hardis, les Espions téméraires et peu soucieux de la vie, les Valets grossiers et fidèles: et ainsi des autres personnes, chacune selon sa fortune, son âge et sa condition.

[make Heroes magnanimous, Philosophers cautious, Women gentle and modest, Maidens full of modesty, Ambassadors bold, Spies hardy and reckless of life, Valets ungainly and faithful; and so on with other characters, according to their wealth, age, and condition.]⁹

Tragicomedy had progressively yielded to this convention since the controversy over Le Cid and criticism of the character not only of Chimène, whose conduct in front of Rodrigue was deemed inappropriate for a ‘fille pleine de pudeur’, but also of King Fernand whose perfunctory reaction, when informed of the threatened attack by the Moors, was considered unkingly. Consequently, in Andromaque, Oreste and Pyrrhus, respectively King of Argos and King of Epirus, must act in accordance with their royal status, which in the seventeenth century meant assuring the welfare of their kingdom, being généreux, and also being galant in love. As for Hermione, daughter of Ménélas and betrothed to Pyrrhus, she must act in accordance with the dignity pertaining to the status of princess. Secondly, a character must resemble the image which tradition has bestowed upon him/her. Thus, Oreste must be melancholy; Hermione, Racine's major borrowing from Euripides' Andromache, must be jealous and vindictive; and Pyrrhus, as Racine indicates in his preface, referring to Horace's Ars poetica, must be ‘farouche, inexorable, violent’ [wild, unyielding, violent].¹⁰ Needless to say, the ideal situation was one where a character was both bienséant and ressemblant, a situation which allows Andromaque to act freely both as a dignified and submissive captive princess, and also as the loyal and tearful widow of tradition. However, the extent to which these two conventions may prove to be contradictory in the case of the other three roles soon becomes evident. How can Pyrrhus, for example, a character on whom contemporary critics of Racine concentrated, be at one and the same time généreux and unyielding, galant and violent?

In his first Discours, Corneille had underlined the contradictory nature of bienséance and ressemblance, and judging it impossible to render them compatible, the only solution he found was to establish a system whereby these conventions were adhered to differently according to the type of subject:

Ainsi ces deux qualités, dont quelques interprètes ont beaucoup de peine à trouver la différence qu'Aristote veut qui soit entre elles sans la désigner, s'accordent aisément, pourvu qu'on les sépare, et qu'on donne celle de convenables aux personnes imaginées qui n'ont jamais eu d'être que dans l'esprit du poète, en reservant l'autre pour celles qui sont connues par l'histoire, ou par la fable, comme je le viens de dire.

[And so these two requirements, between which several interpreters are unable to distinguish despite Aristotle's uninformative insistence that a distinction should be observed, can easily be reconciled, but on the specific condition that they are separated from each other, and that the seemly roles be given to imaginary characters, whose existence was only ever in the mind of the poet, with the other requirement being matched by those who have been identified by history, or legend, as I have said.]¹¹

Corneille's clever solution which, as we shall see, he did not apply in his own work, seems, in fact, to be the only possible solution. This is because the third Aristotelian convention, which demands that a character also be constant, or consistent, for the duration of the play, usually makes it impossible to solve the contradiction by alternating between bienséance and ressemblance. On this subject the most influential European theorist, the Dutch thinker Daniel Heinsius, explains: ‘If one begins by portraying a character as hard, cruel and volatile, he must remain in this disposition until the end of the tragedy’, unless, as he adds, the dramatist introduces a character who is by nature inconsistent, or unless, as in the case of Ajax, the character, during the course of the play, should temporarily lapse into madness before returning to reason.¹²

With all the above in mind, the extent of Racine's daring in the writing of Andromaque becomes obvious and the criticisms levelled at him much clearer. He dared to portray a Pyrrhus généreux and galant like a king of French tragedy, yet inexorable like his classical model. His Hermione is dignified and proud and never demeans herself by wishing for Andromaque's death, unlike the sterile, neglected and jealous matron in Euripides' play. However, like the Euripidean model, Racine's Hermione is also vindictive and jealous. Racine's Oreste conforms to his royal status and hesitates until the very end before assassinating another sovereign, but he also conforms to the tristis Orestes of legend, caught up in the throes of violent melancholy, which leads him ultimately to accept the idea of murder.¹³ Far from favouring bienséance over ressemblance, as most of his contemporaries did and which he himself had done in Alexandre le Grand, where few of the historical characteristics of Alexandre are evident, Racine did not hesitate to combine these two conventions despite their glaring contradiction. However, this was only possible by representing the third criterion of constance in a novel way. Given that the nature of passionate love is its irresistibility, its function becomes akin to that of Ajax's madness in that it permits a temporary rupture in the character's consistency. Thus Pyrrhus can be alternatively généreux and inexorable, galant and violent, in other words, inconstant, or inconsistent as dictated by the rhythm of passion's aberrations.

With his success in staging two-sided characters (i.e. variable and contradictory), Racine had forcefully challenged one of the pillars of dramatic poetics, the principle of bienséance, undoubtedly the most important remaining principle at a time when the question of the rules of unity had been resolved. Consequently, as he himself took pleasure in underlining, Racine had also run headlong into Cornelian and courtly conceptions of the perfect hero, which had derived from the epic and the heroic novel, and to which he had conformed in his first two tragedies:

Quoi qu'il en soit, le Public m'a été trop favorable, pour m'embarrasser du chagrin particulier de deux ou trois personnes, qui voudraient qu'on reformât tous les Héros de l'Antiquité, pour en faire des Héros parfaits. Je trouve leur intention fort bonne, de vouloir qu'on ne mette sur la scène que des hommes impeccables. Mais je les prie de se souvenir, que c'est n'est pas à moi de changer les règles du Théâtre. Horace nous recommande de dépeindre Achille, farouche, inexorable, violent, tel qu'il était, et tel qu'on dépeint son Fils.

[Whatever the case may be, the Public has been too indulgent to me for me to be seriously concerned by the distress caused to two or three particular people, who would like all the Heroes of Antiquity to be reformed into shining Heroes of today. I find their intention to put only perfect examples of behaviour on the stage entirely laudable, but I beg them to remember that it is not for me to change the rules of the Theatre. Horace urges us to portray Achilles as fierce, unmerciful, and violent, as he really was, and as his own son portrayed him.]¹⁴

While the Cornelian conception of the perfect hero is directly inherited from the hero of the epic (and of the novel and tragi-comedy), following the controversy over Le Cid, it was nevertheless validated by the pre-eminence accorded to the principle of bienséance of character. For a long time prior to his elegant 1660 theoretical solution to the articulation of bienséance and ressemblance (quoted above), Corneille had, in practice, privileged the principle of bienséance over ressemblance, whilst at the same time pushing the convention of constance to its limits. In addition, it was only in the case of historically virtuous or historically monstrous characters that Corneille applied his proposal for a theoretical distribution of conventions. However, each time one of his plots required a hero whose character was based on a doubtful past, bienséance took precedence over resemblance. The most striking example of this is the case of Nicomède, the parricidal king, whom Corneille depicts as a perfect prince, and who, contrary to historical fact, chooses to be led to his death rather than revolt against his father.¹⁵ As regards the manner in which he justifies his portrayal of Sophonisbe in the Avis au lecteur of the play of the same title, it is extraordinarily ambiguous (whilst at the same time revealing the same tendency).¹⁶ It is clear that alone amongst his predecessors and contemporaries, Corneille respects the principle of ressemblance. But in the name of what? In the name of what a seventeenth-century image of a Carthaginian heroine might be, in other words, in the name of bienséance!

In order to understand fully the extent of the break with the Cornelian and galant aesthetic, one need only examine the character of Hermione in Andromaque, in particular in the scene at the end of Act IV where Hermione and Pyrrhus meet. Hermione appears firstly as a typical Cornelian princess, not only in her refusal to lower herself by offering the slightest reproach to him who announces her rejection, but also in her constant ironic tone; she seems to have been directly modelled on the character of Éryxe from Corneille's Sophonisbe. However, from the moment that Pyrrhus answers her as though she were a Cornelian heroine (in summary, ‘forgive me for having being presumptious enough to think you loved me’), Hermione breaks down and becomes reminiscent of another model: i.e. that of Ovid's Heroides which is evoked in the informal address, insults, degradation, honest admission, jealousy, ultimate supplication, and veiled threat.¹⁷ More tellingly, when Hermione learns of Pyrrhus's death, Racine denies her any sense of avenged gloire, even if it were only to temporarily assuage her extreme distress. Subligny was precisely sensitive enough to this fact to criticize it by comparing it with the Cornelian model:

Enfin, [Corneille] aurait modéré l'emportement d'Hermione, ou du moins il l'aurait rendu sensible pour quelque temps au plaisir d'être vengée. Car il n'est pas possible qu'après avoir été outragée jusqu'au bout, qu'après n'avoir pu obtenir seulement que Pyrrhus dissimulât à ses yeux le mépris qu'il faisait d'elle: qu'après qu'il l'a congédiée, sans pitié, sans douleur du moins étudiée, et qu'elle a perdu toute espérance de le voir revenir à elle, puisqu'il a épousé sa rivale; il n'est, dis-je, pas possible qu'en cet état elle ne goûte un peu sa vengeance

[At least [Corneille] would have moderated the frenzy of Hermione, or rendered her vulnerable for a short time to the pleasure of being avenged. Because it is simply not possible after being insulted to the limit, after succeeding only in having Pyrrhus mask his contempt when actually talking to her; after being rejected by him without pity or remorse, after she has lost all hope, for her to see him return to her as the spouse of her rival; I repeat that it is simply not possible for her in this state not to savour to some extent her revenge.]¹⁸

It is clear from the first staging of the play that critics mooted the question of divergence from the Cornelian model, and therefore from the dominant norms, of the tragic heroine.

All this brings us to an analysis of the preface to Britannicus, where Racine continuously shelters behind Tacitus and at the same time justifies his portrayal of the character of Néron by the fact that ‘il ne s'agit pas dans [ma] Tragédie des affaires du dehors. Néron est ici dans son particulier et dans sa famille’ [my Tragedy is not about external matters. Nero is here in his private capacity, in the intimacy of his family].¹⁹ This has usually been perceived as bringing us to the crux of the rupture between the Cornelian and the Racinian aesthetic, not in relation to the concept of character, but in relation to the question of history and politics, as if Racine wished to underline that he was not trying to paint vast historical frescoes or to depict major political conflicts. Needless to say, if I deem it necessary to return to this well-worn interpretation it is because it seems to me to be based on a misinterpretation.

There is no doubt that there is a break with Corneille, but not in regard to the question of history and politics, not at least in Britannicus which is the play under consideration here. Once again, it is the conception of the tragic character which is at stake. This is immediately obvious when one compares Othon, Corneille's most recent tragedy, with Britannicus, its exact counterpoint. In both cases, according to Corneille ‘ce ne sont qu'intrigues de cabinet qui se détruisent les unes les autres’ [they are only closet conspiracies, which are mutually destructive].²⁰ Also, in both cases power has been usurped, while clearly the place attributed to the ‘épisode amoureux’ is less ‘épisodique’ in Othon than it is in Britannicus. In short, a comparison of the two plays illustrates that maintaining the endless opposition between the two authors cannot be justified. The difference between the two poets, therefore, lies elsewhere.

Corneille constructed his characters from the image created by external circumstances. Othon, a debauched character of some notoriety, who had shared his wife with Nero, had subsequently presented an outward impression of being a virtuous provincial governor and also (for a short while) of being a righteous emperor. This allowed Corneille to apply the principle of bienséance without hindrance and to represent Othon as a perfect hero. Racine, however, constructs his heroes in the opposite way. Like Othon's final years, the early years of Nero's rule had created the image of a good emperor. But it was an outward appearance based on actions inspired by good counsellors, and the sequence of events tragically revealed that it was, in fact, just an image. Racine obviously uses this image, but he does so in order to contrast it to Néron's ‘true character’, which he could only reveal in his private surroundings, and the lengthy description of which in Tacitus provides the starting part for Racine.

As we have said, far from ignoring Néron's external image, inspired by Sénèque and Burrhus, the virtuous governors of the young emperor, Racine uses it as a counterpoint when Néron's ‘true’ character (the character as revealed through his ‘actions’) is revealed. The admirable scene in Act IV, where the tyrant, on the brink of committing a crime, yields to the tears of Burrhus, undoubtedly constitutes the most perfect illustration of the way the Racinian ethic works. The exceptional power of the scene derives from the fact that the ‘true Néron’ is placed face to face with his external image, which Burrhus tries to reflect, like a mirror, as though it were the true image of Néron's private self. Hence the ambivalence of this two-sided character, whom Racine characterizes in his celebrated formula: ‘c'est ici un monstre naissant’. This represents the essential characteristic which posterity has described as Racinian ‘psychology’: the dramatic principle of the two-sided character (which of course conforms to the ‘psychology’ of the latter half of the seventeenth century, based on the dialectic of mask and face).

We therefore have good reason to take seriously the conclusion to the preface to Andromaque, where Racine confirms his desire to apply to the letter the Aristotelian theory of the tragic flaw, which is linked to the conception of the imperfect hero. The tragic hero must be a virtuous man but capable of weakness. He is thus susceptible to committing an error, which provokes a calamity, and thus arouses fear and pity in the spectator. Clearly a subtle argument: Racine justifies his bending of the Aristotelian system of ‘character’ in the name of Aristotle. It is as though abandoning the perfect hero in favour of the two-sided character had allowed him to access the secret of the perfect tragic hero. Perhaps this preface was an a posteriori justification destined to silence all those who had criticized the behaviour of his principal characters. Whatever the explanation, the accidental merging of the two-sided character with the imperfect tragic hero would subsequently be exploited deliberately by Racine. From Britannicus it would become one of the bases of his formula for tragedy.

The subject of Britannicus, which is based on the subjugation of an innocent victim by an omnipotent tyrant, was the type of subject for which Aristotle had posed the problem of the tragic hero. According to Aristotle, pity is unquestionably the emotion elicited when the spectator is confronted with the undeserved suffering of an individual. However, at the same time, the vision of misfortune suffered by a completely innocent victim creates such an impression of injustice that it prevents all identification, on the part of the spectator, and gives rise to feelings which surpass fear and pity: repulsion and horror, that is to say a feeling that completely destroys the tragic pleasure. When Corneille dramatized the executioner versus victim plot (as in Rodogune, Héraclius, Théodore, Attila, for example), he had surmounted this difficulty by means of his dramatic art of the perfect hero. The heroic (and stoic) response offered by his heroes to oppression managed to replace repulsion by the modern feeling of admiration, in itself a combination of terror and pity. For Saint-Évremond, this was one of the reasons for modern tragedy's superiority over Greek tragedy.²¹

Having turned his back on the Cornelian principles of dramatic art in Andromaque, Racine was necessarily led, in choosing the subject of Britannicus, to reflect upon the way in which to oppose the executioner and the innocent victim. The technique of the two-sided character, introduced for the first time in Andromaque, allowed him firstly to mitigate the horror aroused by an outright tyrant, such as Tristan L'Hermite's Néron in La Mort de Sénèque. Hence the irony of his preface to Britannicus, in which Racine dismisses, one after the other, criticisms which held that his Néron had been depicted as too good or as too cruel. Hence the role he attributes to passionate love in the progressive unveiling of the monstrous nature of a character, who until then had carried out only virtuous actions, even if under the guidance of his good counsellors. At the same time, however, it was also necessary to moderate the extreme innocence of Britannicus: his excesses of love, fervour, candour and blindness result in him carrying out actions which, although innocently motivated, provoke the murderous wrath of the ‘monstre naissant’.²² In short, Britannicus is a hero who does not deserve his misfortune, but who is nonetheless partly responsible for it.

Apart from the specific experiment of Bérénice, this idea of the victim's part in his/her tragedy would consequently determine the representation of a significant number of Racine's heroes. It explains why, at the end of Bajazet, the innocent Atalide assumes complete responsibility for the tragic catastrophe before killing herself. It also explains, in the same tragedy, the variant which emphasizes Bajazet's guilt in his attitude towards Roxane. In the first performance of the play (1672), before he even considers himself ‘barbare, injuste, criminel’ (Act III. 4, l. 995), Bajazet declares:

Et je serais heureux, si je pouvais goûter
Quelque bonheur, au prix qu'il vient de m'en coûter.

(Bajazet, Act III. 4, ll. 943-4)

Later, in the 1676 and subsequent editions this became:

Et je serais heureux, si la foi, si l'honneur
Ne me reprochait point mon injuste bonheur.

This notion also explains why Mithridate, one of Racine's most admirable two-sided characters (apart from Phèdre), perishes because of the flaw of his own blindness. It explains the tragic nature of Agamemnon's role in Iphigénie, and finally, it explains the intense satisfaction of the poet who, in his preface of 1677, would describe Phèdre as the character ‘le plus raisonnable’ that he had put on stage.

However, in order to gauge exactly what Racine meant by ‘reasonable’,²³ one must read what comes next:

Je ne suis point étonné que ce Caractère ait eu un succès si hereux du temps d'Euripide, et qu'il ait encore si bien réussi dans notre siècle, puisqu'il a toutes les qualités qu'Aristote demande dans le Héros de la Tragédie, et qui sont propres à exciter la Compassion et la Terreur. En effet Phèdre n'est ni tout à fait coupable, ni tout à fait innocente.

[I am not all surprised that this character had such success in the time of Euripides, and that it should have succeeded in our own time, because it has all the qualities that Aristotle demands of the tragic Hero, and which are conducive to the emotions of pity and terror. In fact, Phèdre is neither completely guilty nor completely innocent.]²⁴

Phèdre effectively marks the end of a quest, a quest for the ideal tragic hero, who allows for the representation of the entire gamut of tragic emotions within the context of a tragic conflict. In short, Phèdre marks the accomplishment of the two-sided character, who emerged from the break with the system based on the rhetorical conception of ‘character’. And at the same time, Phèdre marks the apogee of the ‘Racinian psychology’ of contradiction, which for three centuries has been considered more natural than its Cornelian equivalent.

In conclusion, it is possible now to understand why, in his preface to Andromaque, Racine ignored criticism of Pyrrhus's galanterie—after all, this was only criticism regarding the degree of galanterie, since all tragic kings, as we saw, need to be courteous and galant according to the laws of bienséance. It also becomes clear why most of this preface is exclusively devoted to the problem of ressemblance. In fact, this issue sums up all the implications of Racine's revolution. It was a risk he could justify only in the name of the principle of fidelity to sources:

Mais véritablement mes Personnages sont si fameux dans l'Antiquité, que pour peu qu'on la connaisse, on verra fort bien que je les ai rendus tels, que les anciens Poètes nous les ont donnés. Aussi n'ai-je pas pensé qu'il me fût permis de rien changé à leurs mœurs.

[But to tell the truth, my characters are so well known from Antiquity, that if one is already familiar with them, it will easily be seen that I have presented them just as the poets of old did. Therefore I felt it was not legitimate to change anything in their behaviour.]²⁵

In this respect Racine had support, dangerous support, which before long would be at the centre of another more general debate, known as the querelle des Anciens et des Modernes. But that is another story.

Notes

1. Subligny, Preface to La Folle Querelle, in Georges Forestier (ed.). Racine Oeuvres Complètes, I, Théâtre-Poesie, coll. de la Pléiade (Paris: Gallimard, 1999).

2. Preface to Andromaque, in Forestier, Œuvres, p. 197. [The reference to Céladon is to the ‘héros galant’, and model of the ‘parfait amant’ in the famous novel Astrée (1607-28) by Honoré d'Urfé (1567-1625), completed by his secretary Balthazar Baro.]

3. Subligny, La Folle Querelle, Act II. 9, in Forestier, Œuvres, p. 280. [The reference to Clélie is to Madeleine de Scudéry's ten-volume novel, published between 1654 and 1660.]

4. Ibid. (All italics are in the original).

5. La Folle Querelle, in Forestier, Œuvres, p. 262.

6. Preface to Andromaque, in Forestier, Œuvres, p. 197.

7. Aristotle, Poetics, Ch. 15, ed. Russell, pp. 110-11. For a French edition, see R. Dupont-Roc and J. Lallot (eds), La Poétique (Paris: Editions du Seuil, 1980), pp. 84-5; Corneille, Discours de l'utilité et des parties du poème dramatique, in Couton, Corneille, 3; see especially pp. 129-33. Ch.15 of the Poetics is the one which Racine translated with most diligence. See Picard, Œuvres, 2, pp. 927-9.

8. The fourth criterion, la bonté, will not be discussed here. For an analysis of its characteristics, see my Essai de génétique théâtrale: Corneille à l'œuvre (Paris: Klincksieck, 1996), pp. 222-3.

9. La Mesnardière, Poétique (Paris: Antoine de Sommaville, 1640; repr. Geneva: Slatkine, 1972) p. 140. [First edition 1639].

10. Preface to Andromaque, in Forestier, Œuvres, p. 197.

11. Couton, Corneille, 3, p. 132.

12. Daniel Heinsius, De Tragœdiae constitutione (Leiden: Jean Baudoin, 1610), Ch. 1.

13. Even when his passion has led Oreste to accept the idea of murder, Cléone's account of events in Act V. 2 reveals him to be unresolved until the very end. Furthermore his own account of events indicates that the Greeks set upon Pyrrhus before Oreste himself could intervene.

14. Preface to Andromaque, in Forestier, Œuvres, p. 197. The reference to Horace is to his Ars Poetica, ll. 120-2. See Horace, The Art of Poetry, in D. A. Russell and M. Winterbottom (eds), Ancient Literary Criticism. The Principal Texts in New Translations, trans. by Donald A. Russell (Oxford: Clarendon, 1972), p. 282.

15. See also the elegant solution proposed in Cinna. The tyrannical and bloodthirsty image of Auguste is rejected in the accounts of the past that the conspirators give: each time Auguste is on stage (in fact from the moment he appears on stage), he presents an image of a perfect sovereign. For further analysis of the characteristics of Cornelian heroes, see my Essai de génétique théâtrale, Ch. 4, pp. 198-270. (For Auguste and Cinna, see pp. 212-15 and 225-9).

16. Couton, Corneille, 3, pp. 382ff.

17. See my article, ‘Écrire Andromaque. Quelques hypothèses génétiques’, Revue d'histoire littéraire de la France, 1 (1998), 43-63. See also my introduction to Andromaque in the Pléiade edition, pp. 136-8.

18. La folle Querelle, in Forestier, Œuvres, p. 262.

19. Preface to Britannicus, in Forestier, Œuvres, p. 372.

20. Corneille, Preface to Othon, in Couton, Corneille, 3, p. 462.

21. ‘J'aime à voir plaindre l'infortune d'un grand homme malheureux; j'aime qu'il s'attire de la compassion, et qu'il se rende quelquefois maître de nos larmes; mais je veux que ces larmes tendres et généreuses regardent ensemble ses malheurs et ses vertus, et qu'avec le triste sentiment de la pitié nous ayons celui d'une admiration animée, qui fasse naître en notre âme comme un amoureux désir de l'imiter. [I like to see the misfortune of a great man pitied; I like to see him draw compassion and tears from us, but I want those generous and understanding tears to take account of his misfortune and his qualities, and that with the melancholy feeling of pity we also have that of admiration, creating within us something like an affectionate desire to emulate him], Saint-Évremond, De la tragédie ancienne et moderne (1674), in Œuvres en prose, ed. by R. Ternois, 4 vols (Paris: S.T.F.M., 1962-9), 4, pp. 179-80.

22. Love, fervour and blindness are among the main characteristics of the behaviour of the young, according to the traditional typology; these same characteristics, when excessive, can be the source of the ‘tragic flaw’, as Racine explains in his preface. For further detail, see my introduction to Britannicus, in Forestier, Œuvres, p. 1418.

23. Here ‘reasonable’ means appropriate (for tragedy).

24. Preface to Phèdre, in Forestier, Œuvres, p. 817.

25. Preface to Andromaque, in Forestier, Œuvres, p. 197.