The retranslation pact: Performability and “writing forward” in Darina Al Joundi’s The Day Nina Simone Stopped Singing

ABSTRACT The term “performability” is widely acknowledged as being conceptually unsatisfactory, yet it is a key component of any translation for performance. By reconsidering performability in the context of Alvstad’s “translation pact” and Alvstad’s work with Assis Rosa on retranslation, this article develops the notion of a retranslation pact that represents a mediation between the author/text and the audience. Key to this mediation is a new theoretically grounded understanding of performability that hinges on an understanding of voice: using direct observations from my author-assisted translation of a French-language play as a case study, an exploration of key discussion points aims to contribute to an improved understanding of the negotiation inherent in the retranslation pact by framing the collaborative process as an example of what Johnston defines as “writing forward”.

process that is fictionalised in her second play Ma Marseillaise [Marseillaise My Way] (Al Joundi 2012bJoundi , 2022b, the sequel to Le Jour où Nina Simone a cessé de chanter. The titular reference to Nina Simone evokes both the secular music Noun's father wanted to be played at his funeral instead of the Qur'an, and the "griot" oral storytelling tradition that Al Joundi enters into with her monologue performances; the reference to the day Simone "stopped singing" implies a violent silencing of women's voices that Al Joundi seeks to redress through her performance. In 2007 Al Joundi first performed Le Jour où Nina Simone a cessé de chanter at the Avignon Theatre Festival: it was an instant success, attracting a large audience and critical acclaim not only at the festival, but also in several runs in Paris and tours across France. The play has also been published as a novel of the same name (Al Joundi and Kacimi 2008) and translated into English by Marjolijn de Jager (Al Joundi and Kacimi 2011). 2 Al Joundi was, however, unhappy with a number of instances of overly literal translation, the dilution of crude language, and inaccuracies in the English translation of the novel (Al Joundi, personal communication). These concerns extended to the (unpublished) first English translation of the play (also provided by de Jager around the same time the novel was published), which Al Joundi felt lost the connection between language and performance. De Jager's translation was used for Al Joundi to perform the play in the USA in 2010 and as a basis for the surtitles at a French-language performance of The Day Nina Simone Stopped Singing at the Chelsea Theatre, London (UK) on 26 and 27 October 2015. Neither event led to the play being commissioned for a run in English, and so Al Joundi approached me to prepare a new translation that would convey her "voice" (Al Joundi uses this term generally; I will engage with it more critically in the course of my analysis), which her agent would subsequently be able to use with a view to securing a London performance. This context provides a response to the "who?" and the "why?" of the retranslation, which are two of the pertinent questions formulated by Cecilia Alvstad and Alexandra Assis Rosa (2015, 9) about driving factors behind a given retranslation. Though the manuscript of The Day Nina Simone Stopped Singing and its sequel Marseillaise My Way were also prepared for publication (Naked Eye Publishing, 2022), the present discussion will focus specifically on the retranslation for performance.
The retranslation process involved extended discussions in French with Al Joundiwhich she noted had not happened with the initial translationand our exchanges resulted in sometimes significant edits to my early drafts. The previous translation of the play was not consulted during the translation process, but has been viewed in two ways: first, I observed its use in the surtitles at the French-language UK performance in 2015; secondly, I was able to access a copy of the surtitles after completing the retranslation, which has allowed me to make more engaged comparisons between the previous translation and my completed version. The early observation of the surtitles allowed me to identify a number of shifts between the text as it was performed in French and read in English by the primarily Anglophone audience, whose reception appeared more muted in comparison to that of the Francophone audience when I attended Al Joundi's performances in France in 2008 and 2012. Though this could, in part, be because a majority of the audience was responding primarily to a (slightly, though not significantly) condensed text rather than to the text-as-performance, other potential contributing factors include a tempering of the black humour, calques that distance the performer from the audience, occasional phrases that seem to contradict the overall context of the narration, and a sanitising of Noun's tone in places. The two principal considerations of the retranslation were as follows: first, to engage with the author's concerns and reflect on how the shifts noted above might have affected the reception of the translation; secondly, to consider how the retranslation might address these shifts to elicit an audience reception comparable to the one it enjoyed throughout its several runs in French.
In "Professing Translation", David Johnston claims that the theatre translator's role is "to write forward, to bring the potentials for performance that are encoded in the original explicitly into the temporal and spatial purview of the audience […] to engage on behalf of the new receiving context with the sweep and scope of the text's possible meanings" (2013,366), suggesting an agency on the part of the translator that goes beyond a single or fixed meaning of the original playtext. Numerous "potentials for performance" have emerged not only from my own reading of the original text and experience of the performance in French, but also from my dialogue with Al Joundi, which enabled me to add her perspective on the "sweep and scope" of the original playtext to my own interpretation of it. Johnston's insistence on the plurality of "the text's possible meanings" also recalls Chantal Wright's rejection of meaning as "an inherent, unchanging property of a text" (2013, 27): both Johnston and Wright see the translator's role as a mediator of meaning, arbitrating between the original (play)text and the new context into which it is brought with the (re)translation. Though Al Joundi had fixed ideas regarding certain linguistic aspects of the translation, by following Johnston's model of "writing forward", which advocates for a foregrounding of the interpretative scope in the translation, my goal for the retranslation was to reconcile this with audience expectations regarding a familiar use of lexis and syntax. This would bring insights gained from my discussions with Al Joundi into the "temporal and spatial purview of the audience" but avoid the literal translation approach that I perceived as restrictive in terms of bringing the "sweep and scope" of the text to the target audience. In the retranslation, my aim was to produce a playtext that was "simultaneously alive to the contexts of the original text […] and responsive to the cognitive and affective processing of a new audience" (Johnston 2013, 373); this dual-facing approach is at the heart of the retranslation pact, which is itself an attempt to balance potentially conflicting demands of authorial direction and target language norms. The author commissioned the retranslation owing to her sense that the existing translation had distanced her from an Anglophone audience, but she had certain ideas about how the retranslation should address this, specifically regarding syntactical emphasis and brevity of expression. These intentions, "encoded in the original" and decoded through extensive dialogue, were at times in conflict with my commitment to the audience, and the imperative to reconcile the two renders the process of "writing forward" a mediation which forms the basis of the retranslation pact.

The retranslation pact as a two-way contract
The retranslation pact builds on Alvstad's notion of the "translation pact", a term that borrows from Philippe Lejeune's (1975) much-cited "autobiographical pact". The translation pact "invit[es] the reader to read the book as if it were written only by the author" (Alvstad 2014, 271), and although Alvstad refers explicitly to literature, we can extend her analysis to theatre in that it invites the audience to watch the play as if it were written only by the author. If the translation pact breaks down in the initial translation owing to features that mark it as a translation (particularly with literal translations that do not have the same field of usage in the target language), here the retranslation pact aims to produce a text that will function successfully in English without alienating the audience from the text/performance. The retranslation pact thus becomes an implicit contract with both the author and the audience that is premised on an accurate representation of voice.
Alvstad and Assis Rosa's work on voice and retranslation thus becomes a useful frame of reference for exploring the "writing forward" inherent in retranslation. The underlying premise of their approach is that "the study of voice can enhance our understanding of retranslation, and vice versa" (Alvstad and Rosa 2015, 3). It is by focusing on voice (and the intersection between how Al Joundi understands this term and its more theoretical applications as outlined by Alvstad and Assis Rosa) that I refine my scope, as it is key to both my conceptualisation of performability and the nature of the "retranslation pact". A crucial part of the performance and its reception is the rapport between Al Joundi and her audience: invariably, she is sitting on the stage floor as the audience enters the theatre, and she frequently addresses her audience directly throughout the play. The relative immediacy of this relationship between performer and audience undoubtedly affects the reception of the play (and is made explicit in the published manuscript via stage directions). The "writing forward" approach thus enables the retranslation pact to differentiate itself from the translation pact because of its two-way responsibility to the author/text and to the audience, and rests on the importance of curating the cultural encounter between the two parties. 3 Integral to the retranslation pact's focus on both the "sweep and scope of the text" and the "new receiving context" is the notion of voice, and so I turn again to Alvstad and Assis Rosa for further clarification of this term, and in particular the distinction between textual voices and contextual voices in translation: Textual voices are part of the product (narrative voice, the voices of characters and the translator's textually manifested voice), whereas contextual voices are related to the sociological translation process and hence to the multiple agents that produce, promote and write about translations. (Alvstad and Rosa 2015, 3-4) The two main (imbricated) textual voices in the source text are Noun and Al Joundi, with Noun representing for Al Joundi: Noun's narratological function is as a mouthpiece for Al Joundi, a means to say all that she cannot say with her own voice, and as such it is important to communicate in the retranslation what Noun says (the textual voice), while remaining alert to how she is directed to speak by the contextual voices of the author and translator in dialogue with each other.
The French source text of The Day Nina Simone Stopped Singing is about a nation that historically was ruled by France (though the narrative is situated after this, taking place before and during the Lebanese civil war), but that has very different governing policies, social norms, and considerations of human rights. However, the French mandate left its mark on Lebanese culture in both evident ways (such as French still being the second language of Lebanon) and more insidious ones (for example, Al Joundi notes in Vassallo [2012] that the law allowing families to intern women in mental asylums is a French law). After her release from the mental asylum, Al Joundi saved enough money to move to the west, as her liberty was uncertain in Lebanon. The necessity of this departure is fictionalised in Marseillaise My Way, in which Noun explains that: I wasn't allowed to go out any more, I wasn't allowed to walk or laugh or dance, I didn't even want to do any of that any more, I just wanted to leave. My city, my country, my friends, my neighbourhood, my family, everything … it was all lost to me now, I had nothing left. (Al Joundi 2022b, 12) Therefore, while the source culture specificity is crucial to the narrative of The Day Nina Simone Stopped Singing (the context of the Lebanese civil war is the basis for each key feature of Noun's story), the intended audience is western. The play was originally written in French rather than Arabic and has been performed throughout France (indeed, the intended audience is increasingly evident in Marseillaise My Way, which represents a direct plea to the French Republic to offer Noun protection from the people and laws she has fled).
The retranslation pact, with its focus on the relationship between author and audience, aims to ensure that the relevance of the source culture specificity is clear in the translation into English, but without creating additional distance. Al Joundi's performance is already marked in English owing to her accent (although she speaks English fluently, she retains a strong accent that identifies her as a non-native speaker; this is far less marked in French), and the comedy and drama of her performance rest on an intimacy between her and the audience. Given that there is a contextual distance between the performer and the target audiencethe UK not having the same historical ties with Lebanon as France doesmy priority for the retranslation was to create a text that remained, in line with Johnston (2013, 366), "alive to the contexts of the original text" as outlined above, while striving to be "responsive to the cognitive and affective processing of a new audience". This moves away from Alvstad and Assis Rosa's either/or assertion that a retranslation, like any translation, "may relate differently to the source text and culture as well as to the target text and culture, in that it may be predominantly source-or target-oriented" (Alvstad and Rosa 2015, 9). Instead, Johnston's position allows for the retranslation to combine the source-oriented goal of preserving the author's voice with a target-oriented approach of bringing the text closer to the new audience. These two considerations are key to both translating voice and curating the cultural encounter between author/performer and target audience. They also come together in the notion of "performability".
Pinning down "performability": Verbal and non-verbal voices The foregrounding of voice in the retranslation pact is fundamental to my understanding of "performability" and its importance to reception. The merits of performability as a solution to the difficulties of translating for performance are explained by Susan Bassnett in the following terms: The translator of dramatic texts is expected to grapple not only with the eternal problem of "faithfulness", however that may be interpreted, but also with the problem of what the relationship between the written and the performed may be. "Performability" offers a way out of the dilemma, since it allows the translator to take greater liberties with the text than many might deem acceptable, in the interests of the end product of "performability". (1998,96) However, in her extensive discussions of performability (and particularly in her later work on the subject), Bassnett expresses understandable caution about using "performability" as shorthand to explain that something is judged to sound better, which could be used to justify any translation choice without giving any substantial or substantiated reason for this choice. My discussions with Al Joundi allowed insight into "the relationship between the written and the performed" that offered a more concrete understanding of performability in the context of this play, particularly with regard to the non-verbal elements of the monologue (the gestic text, how Al Joundi occupies the space on stage, and the limited use of props) and how they interact with the verbal delivery. Although Charlotte Bosseaux (2015) has offered a more theoretically rigorous overview of "translating performance" in the context of audiovisual translation, in the context of translating for the stage, this seems as unsatisfactorily vague as "performability", and so the analysis will now seek to better define these slippery terms.
Silvia Bigliazzi, Peter Kofler and Paola Ambrosi assert that "performability may be seen as a pragmatic instrument associated with the style, conventions, and ideology of the target cultural environment, rather than as an abstract and universal quality of the source text" (2013, 4). Although pragmatism is essential in order to produce a playtext that will recreate as closely as possible the relationship between the author/performer and the audience, considerations of performability must nonetheless still encompass a "quality of the source text", which I hope to show is neither abstract nor universal, but can be understood and substantiated via a discussion of voice.
Bassnett's critique of how "performability" has been used as a catch-all term to justify "substantial variations in the target language text, including cuts and additions" is further explained in her comment about the relationship between verbal and non-verbal voices: "the term 'performability' is also frequently used to describe the indescribable, the supposedly existent concealed gestic text within the written" (1991, 102). However, rather than performability being used as an excuse for taking greater liberties in the translation, I propose that it can be an integral part of the negotiation inherent in the retranslation pact. This negotiation is both pragmatic, "associated with the style, conventions, and ideology of the target cultural environment", and focused on a "quality of the source text" (Bigliazzi, Kofler, and Ambrosi 2013, 4)which I interpret as the "what" and "how" of Noun's voicein order to develop a retranslation that both functions as a written text and can be performed in conjunction with the gestic text transferred from the original. The author-assisted element allowed the "concealed gestic text" to emerge: as both author and sole performer, Al Joundi's input into the translation was inflected with the years spent honing the performance of the original text, which has undergone several shifts in its mise-en-scène over the years. 4 Al Joundi's understanding of performability maps onto Eva Espasa's claim that "from the point of view of theatrical practice, playability, or actability, are used as synonyms for performability" (2000,50). Al Joundi's arguments during the negotiation process generally rested on how she would deliver a given line, which determined the rhythm and emphasis of particular sections of the translated text. She also focused on the gestic features that would accompany the words, prioritising these and requesting that the syntax of the translation should fit with features of the performance. My initial focus was more linguistic, focusing on the informality and intimacy of the register, standard target language collocations, and a compensation that Al Joundi frequently felt represented the "cuts and additions" she wished to avoid. The discussions that occurred during the editing process extended beyond the text to consider intersemiotic issues that brought together the linguistic and the physical, corresponding not only to Bassnett's contention about interpreting "the relationship between the written and the performed" (1998,96) but also to Johnston's assertions that "the theatre translator is involved not just with a rereading but, in terms of performance, a re-making of text" (2013, 367). The physical aspect of Al Joundi's text-as-performance was inseparable from the words her character delivers, and so both the verbal and the non-verbal had to be negotiated in the translation. This focus on how Noun is directed to speak exemplifies the two-way retranslation pact, compelling me to take on board instances where Al Joundi was unwilling to compromise, and thinking about how these could best be rendered in terms of the encounter between author/performer and audience.
Though the original translation was not consulted as a written script during the process of retranslation, the aforementioned opportunity to observe the surtitles before beginning the retranslation allowed me to identify key moments when shifts in meaning might have affected reception or did not fully transmit the sweep and scope of the source text. The retrospective consultation of the original script offered the further possibility of interrogating these specific instances of divergence between the two versions, and thus of considering whether the retranslation pact had been successful in balancing responsibilities to author and audience in these cases. One example of a condensed surtitle not fully conveying the sweep and scope of the original comes in a key scene when Noun finally questions her father's motives; she accuses him of having used her as a laboratory rat to try out his experiments with freedom and shouts "Je crois que tu as oublié que j'étais ta fille" ("I think you forgot I was your daughter"). In the surtitles, this was rendered as "I think you forgot your daughter", which does not convey Noun's disappointment with her father's failure to consider the consequences of his actions, forgetting his duty and desire to love and protect herthat is, forgetting the relationship between them, indicated in the source text and the retranslation, rather than forgetting about her or forgetting the person, as implied in the previous translation.
There are also details in the original translation that mark it as a translation (such as "daughter of mine" for the repeated affectionate and informal use of "ma fille", and "my father was a strange bird" as a near-literal translation of "c'était un drôle d'oiseau, mon père"); though both of these literal translations ("daughter of mine" and "strange bird") exist in English, they do not have the same field of usage, and so they appear more marked in the translation. Therefore, because of the immediacy of the playtext and the relationship it is necessary to forge between performer and audience as part of the retranslation pact, my own choices are target-oriented for the following reasons: in French, "ma fille" is an idiomatic term of address; in the original translation the translator transposes the possessive from the adjective (ma) to the pronoun (of mine), yet the result is not an equivalent idiomatic term of address in English. As a marker in conversation, "ma fille" is used as a brief interjection that does not carry much emphasis, but that indicates a closeness between interlocutors. "Daughter of mine" is longer, more deliberate, and not necessarily a term of endearment, as it can also have a formal and admonitory presupposed meaning. The original translation prioritises the grammatical possessive over the meaning or function of the interjection; even Al Joundi's pronunciation of "daughter of mine" makes the phrase cumbersome. This risks creating a distance between the actor and the audience that both undermines the intimacy of the mise-en-scène and fails to mediate between the interpretative potential of the source text and the "cognitive and affective processing" of the target audience (Johnston 2013, 373). The retranslation seeks to redress this: in the retranslation, "ma fille" is translated variously with a term of endearment such as "sweetheart" or "darling", when its use in the source text implies the close relationship between father and daughter, or simply "Noun", the character's name, for more general pronouncements.
Similarly, "drôle d'oiseau" is a more common collocation in French than "strange bird" in English, and I proposed many suggestions for the term until Al Joundi explained that in addition to its primary meaning of not fitting in with the norm, she also meant to indicate that her father would not be restrained or labelled. This discussion of meaning results in two phrases, each representing one aspect of Al Joundi's use of "drôle d'oiseau": "My father never ran with the crowd. He was a free spirit." The brevity of these phrases respects Noun's repeated use of short sentences and was a rare instance of addition in the retranslation, justified in terms of the audience's cognitive load: the use of two phrases creates an emphasis that compensates for the shift in word order in the original (where the subject is placed at the end of the clause for emphasis), and the choices bring the performance closer to the audience by conveying the plurality of meaning in the original phrase with expressions which are not marked in the target language. To return to Johnston's terms, this negotiation "engage[s] on behalf of the new receiving context with the sweep and scope of the text's possible meanings" (Johnston 2013, 366, my emphasis), highlighting the importance of the two-way pact.
The "what" and the "how" of voice: Humour, rhythm and rhyme In addition to the kind of stylistic details analysed above, there was a particular passage in the previous translation that Al Joundi felt did not fully convey Noun's voice: a lighthearted section in which she describes rhythmically and humorously the different men's genitals she has seen. This section crystallises Noun's notion of "freedom", her poetic irreverence and her crude humour; an essential part of the retranslation pact was to communicate the "what" and "how" of Noun's voice here. The "what" is the succinct list and the representation of each image; the "how" is, firstly, the comical and suggestive turns of phrase (as the humourand therefore the audience responserests on the crescendo) and, secondly, the regular metre and rhyme (for the comic effect that this lends to the list). I first noted down a literal translation of each word, then focused on representing the images they imply in a series of nouns, adjectives, and references recognisable in the target culture, prioritising metre and rhyme over the order of the original list: J'en ai vu des tordues, des ventrues, des qui avaient comme un torticolis, des moussues, des étiques, des rachitiques, des longilignes, des sous Cellophane, des camemberts, des Carambar, des fondants sans chocolat. Des fèves pour la galette des rois, des entêtées, des pics-verts, des hérissons, des polissons, des burins, des Haribo, des scoubidous, des zigomars, des soufflés au fromage, des marteux-piqueurs, des ratons laveurs, des arlésiennes, et des rappeurs [I've seen twisted ones, pot-bellied ones, ones that looked like they had a stiff neck, short and curly ones, shrivelled ones, scrawny ones, slender ones, ones wrapped in Cellophane, camemberts, Carambars, fondants without chocolate. Beans for the Epiphany cake, stubborn ones, green woodpeckers, hedgehogs, rascals, chisels, Haribos, friendship bracelets, Zigomars, cheese soufflés, pneumatic drills, raccoons, elusive ones and rappers] The original translation uses alliteration rather than rhyme, which is achieved by adding adjectives to some of the nouns and omitting others; the section therefore has a predominance of adjectives or adjective-noun combinations, rather than the nouns (or adjectives used as nouns) of the original: I've seen kinky cocks, limp noodles, loose linguine, flabby ones, twisted ones, stiff-necked, foamy, wiggly worms, furry, fuzzy, wilted willies, droopy dicks, pesky pricks, woody peckers, slimy snakes, hairy hoses, prickly pears, jolly johnsons, marinated meat, dirty dicks, sorry soufflés, fluffy marshmallows, melted M & Ms, yummy Mars bars and wibbly wobblers.
There is an attempt to convey the humour (for example, in the wordplay of "woody peckers") and to create a crescendo by increasing the gastronomical metaphors that represent one recurring feature of the original text. The animal references are included, but with different images (worms and snakes rather than hedgehogs and raccoons); some references are repeated (limp/wilted/droopy; loose/flabby), while many are omitted and new ones introduced. There is one mistranslation, where "moussues", an adjective meaning "short and curly" is conflated with the noun "mousse", translated here as "foamy", and the absence of rhyme means that the section does not have the cumulative comedy of the original.
French offers greater possibilities for rhyme, because of its limited range of noun endings and the way in which past participles follow set patterns. To achieve a similar effect, I used adjectives plus the plural pronoun "ones" to open up more possibilities for rhyme, and working from images rather than words (so "des ventrues/pot-bellied ones" becomes "puffed-up ones", and "des sous Cellophane/ones wrapped in Cellophane" becomes "trussed-up ones"). The order of the images is abandoned in favour of the rhythm and rhyme, so "bracelets, gimlets, giblets, ringlets" represents "des scoubidous/friendship bracelets", "des burins/chisels", "des rachitiques/scrawny ones" and "des moussues/short and curly ones", and relies on the similarity of the word endings ("-lets") and the assonance of the last three ("gimlets, giblets, ringlets") to recreate the comedy of Noun rattling off her list of conquests. Physical descriptions might be represented by either physical characteristics or personality traits ("peaky ones, cheeky ones, floppy ones, stroppy ones" represent "des étiques/puny ones", "des polissons/rascals", "des arlésiennes/elusive ones" and "des entêtées/stubborn ones"). The French culture-specific food references are not as immediately recognisable to the target audience, and so these were transposed into more general references that could still describe the shape or function of a penis: for example, the "fondants sans chocolat" is a play on words with the popular French dessert fondant au chocolat, rendered in the original translation with the reference to "melted M&Ms" (which does not have a direct association with the image it is supposed to evoke). In the retranslation, the culture-specific "fondants au chocolat" become the more universal "fudge cakes", the disappointment of the sexual encounter emphasised with the sexually loaded caveat "without the fudge". Similarly, the "fèves pour la galette des rois", the figurines (literally "beans") hidden in the frangipane cake eaten on Epiphany become the more generic "jelly beans", which also broadly encompasses references to the confectionary brands "Carambar" and "Haribo". This left possibilities open for other additions where an image in the French was difficult to replicate in one word. For example, "Zigomar" is a fictional masked anti-hero, whose qualities are conveyed in both "wolverines" and "caped crusaders": the connotative meanings of dark conflicted superheroes (the X-Men character Wolverine and the recent reboot of Batman, the caped crusader) combine with the image of "ratons laveurs" [raccoons] in "wolverines" and of a condom-clad penis in "caped crusaders". These choices, which are more instantly recognisable to the target audienceand thus crucial to the translation of the text's humourare more domesticating than the equivalent section of the original translation: the "quality of the source text" (Bigliazzi, Kofler, and Ambrosi 2013, 4) is thus communicated in a way that brings the text closer to the target audience, curating the cultural encounter in a manner compatible with the two-way responsibility of the retranslation pact. This is achieved through an attempt to remain "simultaneously alive to the contexts of the original text" by focusing on what the textual voice conveys (the list of images) and "responsive to the cognitive and affective processing of a new audience" (Johnston 2013, 373) by taking account of how the voice is expressed: the combination of humour, rhythm, rhyme and imagery in the delivery of the lines.

Understanding the gestic text: Balancing the linguistic and the extralinguistic
The relationship between the linguistic (the text) and the extra-linguistic (the performance) can be further clarified by what Al Joundi describes as the "musicality" of the text, which also falls under the "how" of voice (in its verbal and non-verbal expression). Al Joundi's understanding of musicality was the basis for much of her negotiation, and maps onto notions of rhythm, scansion and word order. At several junctures, Al Joundi insisted that a particular word should remain at the end of a sentence or section for dramatic emphasis. For example, the first scene closes on Noun recounting how hordes of weeping womenmostly former mistresseshad come to mourn her father's death. The women kiss his hands, face and feet greedily, and Noun suggests that one of them should try kissing his penis, remarking drolly that: "On ne sait jamais, elle aurait pu te ressusciter. Elle aurait joué Jésus et toi, Lazare!" ["You never know, she could have resuscitated you. She could have played Jesus and you, Lazarus!"] My original suggestion for the retranslation was "you never know, there could have been some life in the old dog yet. She might have raised you from the dead!", substituting the image of resuscitation with a colloquial phrase to imply the father's sexual voracity, and omitting the direct reference to Jesus and Lazarus with the rationale that "raising from the dead" is enough to evoke the story of Lazarus (and, with its biblical connotations, more directly so than the more generic phrase used in the original translation "she might have brought you back to life"). Al Joundi insisted that she wanted to retain the explicit reference to Lazarus, at the end of the sentence, with a pause before the word. This was not for ideological reasons (although it does bear out the father's understanding of "secular" as equivalent to "not Muslim", and his admiration of Jesus as a figure comparable to Che Guevara). Rather, it was because of the relationship between verbal and non-verbal voice: in the performance, Al Joundi would raise her hand upwards dramatically as she said the word "Lazarus", and the lights would go down. This precluded any formulation such as "she could have been Jesus to your Lazarus", as the dramatic pause would be uncomfortable between the possessive adjective "your" and the proper noun "Lazarus". Here Noun's voice, particularly her irreverence for religion and her ability to laugh in the face of absurdity and sorrow, is entwined with the gestic text, as the physical gesture that accompanies the final word and the pause that precedes it are crucial to the humour of the moment. This negotiation is not only indicative of the ways in which the verbal and non-verbal are both key to the "how" of translating the textual voice, but also bears out Espasa's assertion that "[t]heatre translation, embodied in the performers' bodies, voices, and attire, is very visible and physical" (2013,39). The visible and physical aspects are unknowable from the written text, but the punctuation in the retranslation helps to hint at them: the agreed version "she could have been Jesus and you … Lazarus!" creates a pause and emphasis through the ellipsis (which was a comma in the French script). In terms of cognitive processing and grammatical expectations, the omission of a verb before the proper noun is compensated by the pause (and particularly by the ellipsis in the written text, which offers the suggestion that a word is missing). This solution respects the interpretative layers of the text as expressed through "the performer's bod[y and] voice" (Espasa 2013, 39), thereby allowing for the same gestic features to accompany the phrase. Retention of both aspects of the original translation ("you never know, there could have been some life in the old dog yet. She might have raised you from the dead!"), as a partial addition prior to the climactic ending on the proper noun "Lazarus", also ensures that the irreverent tone, as well as the biblical context, is clearly communicated within the text.
A further example of negotiation comes towards the end of the text: Noun calls out in anger to her father, saying "J'ai pensé à toutes ces années, à tes rêves, à ta folie, à tes vérités, à tes défaites" [I('ve) thought about all these/those years, your dreams, your madness, your truths, your defeats]. The French perfect tense can be translated by either the perfect or the simple past, and the demonstrative adjective can be more immediate ("these") or distanced ("those"); in each case, de Jager and I made different choices, but the main concern here is the second half of the section. This comprises a series of nouns which cannot be rendered literally in English, as "defeats" is rarely attributed with a possessive but rather one is defeated by something/someone, and "truths" in the plural has a much reduced field of usage (the original translation here is literal, but adds a preposition and a conjunction: "I've thought about all these years, about your dreams, your madness, your truths and your defeats", my italics). In addition to these grammatical considerations, "folie" has a wider field of meaning in French, covering both madness and folly; the former is a potentially loaded term given Noun's incarceration in the asylum owing to her adherence to her father's beliefs. The first draft of the retranslation converted the nouns into noun phrases or clauses that made the context clearer ("I thought about all those years, your dreams, your crazy ideas, the truth that was so important to you, and how you lost in the end"), but Al Joundi pointed out that the brevity was important, as with each noun she strikes the floor to intensify the drama. This combines the "what" of the voice (the list of nouns) with the "how" (the concise list, representative of the pared-down way in which Noun speaks, and the gestic text, striking the floor in time with the list) to give substance to Noun's voice and its importance for performability. As such, the final version focused on finding nouns which work in the plural and with the possessive without needing to be part of a noun phrase or noun clause, each of which would reflect the corresponding element in the source text sentence. Thus, this sentence became "I thought about all those years, your dreams, your delusions, your principles, your failures", communicating both the "what" and "how" of voice because of its consideration of both the linguistic and the extra-linguistic.
These decisionsboth the first, that was rejected by the author, and the second, which was approvedshow the importance of negotiation in fulfilling the retranslation pact, as the four nouns needed not only to work in the plural and with the possessive, but also to convey the sweep and scope of a source text that encompasses the relationship between the protagonist and her father, the importance of his idealism against state barbarism, the impossibility of "freedom" in a country that was not free, and the horror of finding that every struggle for liberty and equality had been in vain. For the retranslation to work in terms of cognitive and affective processing, this dialogue was crucial precisely because of the extra-linguistic considerations that an audience would see and hear, but which are not evident in a written manuscript. Thus, the retranslation pact here exemplifies Johnston's definition of "writing forward" as "bring[ing] the potentials for performance that are encoded in the original explicitly into the temporal and spatial purview of the audience" (Johnston 2013, 366).
Conclusion: A two-way pact to retranslate by "writing forward" The retranslation pact gives substance to the notion of performability through its focus on the mediation of voice, which is explored as a balance between meaning, textual possibilities, and audience reception. Though there are many reasons why a retranslation might take place, the retranslator's role is always likely to involve balancing multiple responsibilities as they negotiate both the voice of the source text and the "new receiving context" (Johnston 2013, 366). Above all, the focus on retranslation as both a practice and a concept advances the "fostering of dialogue" promoted by Cristina Marinetti, Manuela Perteghella and Roger Baines with regard to theatre translation (2011, 2). Dialogue is key to the negotiation inherent to the retranslation pact, reconciling theory and practice, the "consciousness of translation" and the "dynamics of performance" (Johnston 2013, 373). To apply this to Al Joundi's case, we need to take into account not only the personal story and historical accuracy (the "fixed" contextual categories), but also verbal and non-verbal voices, articulation for a non-native speaker of English, delivery of the playtext as performance, and the non-verbal elements of the text as detailed above. These are brought together in the focus on performability (the "what" and the "how" of voice). It is in and through Noun's voice that all the key issues of the play are transmitted, and so by focusing on performability and voice, a translation can emerge that curates the cultural encounter between actor/author and audience more successfully.
Building on Alvstad's theorisation of the translation pact, I have suggested that my own approach was based on a retranslation pact, where the main aim of the retranslation was to offer in the target language a playtext comparable to the original, by focusing on its performabilitya term I have reformulated through a discussion of voice. It would be false to suggest that greater author involvement in the translation process is the sole factor responsible for a more "successful" retranslation; rather, the mediation between authorial direction and responsibility to the target audience came together in the retranslation. The direct processes of negotiation and creative collaboration have thus enhanced the retranslation pact: using the retranslation as a way of reading and understanding the textmore particularly, to understand both the textual and contextual voiceshas helped to define the scope of the retranslation pact as a means of "writing forward" to curate a new cultural encounter between the author/performer and the target audience.

Notes
1. This applies to all languages Al Joundi speaks (French, Arabic, English), though the play has been translated and performed in other languages. 2. The relationship between Al Joundi and Kacimi is strained, and Al Joundi has obtained legal confirmation that the playtext is entirely her own work. 3. I borrow the term "curating cultural encounters" from Esther Allen and Susan Bernofsky, who describe the translator as "an intellectual figure empowered with agency and sensibility who produces knowledge by curating cultural encounters" (2013, xix). Allen and Bernofksy's focus on the translator's agency allows me to move beyond the invisibility implicit in the "translation pact" and focus instead on the mediation between author and audience. 4. Specifically, these changes have been about how Al Joundi occupies the space on stagethe "gestic features" that accompany the text, and how she delivers particular lines for maximum comic or tragic effect.

Disclosure statement
No potential conflict of interest was reported by the author(s). This study did not generate any new data.

Note on contributor
Helen Vassallo is a Senior Lecturer in French at the University of Exeter. Her primary research interests are in translated literature and feminism, with a focus on contemporary women's writing and theory. Her research project "Translating Women" explores women's writing in translation and challenges gender bias in the UK independent publishing industry and is disseminated widely via a popular blog (blogs.exeter.ac.uk/translatingwomen) and Twitter account (@translatewomen).