‘En nous humblement requerant’: Crime Narrations and Rhetorical Strategies in Late Medieval Pardon Letters

The charters of pardon or ‘remission letters’ granted by the king of France and the duke of Burgundy in the late Middle Ages have often been interpreted as a valuable source material to access to the lives, memories, and even the ‘voices’ of the ordinary people who did not produce any other writing. The preamble of each letter was indeed a copy of the supplication submitted by the future pardon beneficiary, in which he narrated his crime and begged for the sovereign’s mercy. However, as with many other documents preserved in court and chancery records, remission letters engage historians with a series of methodological questions due to the nature of the documents and the context in which they were produced. Because they hoped to be granted the monarch’s mercy, petitioners used a series of legal and rhetorical techniques to describe themselves and elaborate their narrations, following the advice of the clerks and lawyers who helped them to compose their petitions. This article explores some of these strategies used by French and Burgundian pardon beneficiaries and compares them to those found in English petitions for pardon. It argues that rather than being considered as an obstacle to accessing to the truth behind the sources, these strategies should be analysed as the testimony of the legal and administrative practices of the time.


Introduction
On an unspecified day around the year 1435, Ywain Voet, a Flemish messenger of the ducal equerry of Philip the Good, ' accidentally encountered' ('d'aventure rencontra') another messenger named Master Jehan and beat him to death. After fleeing from the county of Flanders to escape justice, Ywain submitted a petition for pardon to the duke of Burgundy, explaining the reasons for his fatal gesture. One year before this affair, Master Jehan, who was such a close friend to Ywain that they 'had fraternal company as if they were brothers' ('avoit compaignie fraternelle comme a son frere'), seduced Ywain's wife and eloped with her, taking some of Ywain's properties along with her. Thus, when Ywain Voet unexpectedly met Jehan, he was immediately 'moved by sentiments' ('esmeu de couraige'), remembering the betrayal of his former friend, and could not refrain from killing him. In the end, the messenger was fortunate enough, because Philip the Good and his counsellors were convinced by his arguments and granted him a pardon in the form of a remission letter in December 1438, which contained a copy of Ywain's petition. 1 Unlike the duke of Burgundy and the members of his council, a contemporary reader of Ywain Voet's petition may be sceptical about his story. How can we be sure that what is presented as an unpremeditated brawl wouldn't have been an ambush in reality? After all, the document tells us very little about the context of Ywain's crime. The petitioner was apparently unarmed when he met Master Jehan, but he also admitted in the petition that he was accompanied by two other men whose participation in the fight remains uncertain. Moreover, the place where they fought is not mentioned: was it in Nieuwpoort, the place of residence of Ywain and Jehan, or in another city of Flanders? Did they meet on the street, or maybe in a tavern, a popular place for travellers and messengers like them? Were there other people present during the scene or did it take place out of sight? In fact, most of the petition is dedicated to Jehan's betrayal and Ywain's feelings about this, as if the petitioner wanted his motivations to overshadow the circumstances of his crime. Ywain may have had good reasons to elaborate his story that way, as he was probably aware that 1 Achives départementales du Nord [hereafter ADN], B 1682 fol. 34r-v. See also Arnade and Prevenier (2015: 97-8), from whom I borrow the English translation of this case.
Verreycken: 'En nous humblement requerant' 3 admitting premeditation would jeopardize his demand for a remission letter. Indeed, even though there was theoretically no real limit to the ducal power to pardon in the 15 th century, it was common knowledge that 'there are causes or motives because of which there is a greater tendency to grant a pardon' ('y a des causes ou motiz pour quoy on est plus enclin de passer la grace'), as wrote Odart Morchesne, a notary and secretary at the French royal chancery, in the formulary he achieved in 1426 (O.S.) according to its dating -that is to say, between 31 March 1426 and 20 April 1427 (N.S.). Homicide, in particular, was more excusable 'when the case was committed by anger rather than ambush or malice' ('quant le cas est advenu de chaude cole et non pas d'aguet appensé ne de mauvais malice') (Guyotjeannin, Lusignan, and Frunzeanu, 2005: 424). Although it is very unlikely that Ywain Voet had ever heard about Morchesne's formulary, he may have benefited from the help of legal advisors who were familiars with the rules of the ducal chancery to draft his petition, adapting his testimony to the criteria for remissibility recommended by Morchesne. Ultimately, whether Ywain Voet deliberately tried to dissimulate some aspects of the circumstances of his crime or not, it is clear that he elaborated a narrative oriented toward one single purpose: to demonstrate that his case was excusable and that he deserved to be pardoned by the duke.
This speculative reflection on the sincerity of Ywain Voet's petition illustrates what Shannon McSheffrey (2008: 65) calls the ' epistemic doubt' experienced by historians when working on archival documents, particularly legal records. Not only are these sources often incomplete or fragmented, but they are also not designed to offer an unbiased record of the past. It is a common statement that the critical interpretation of documents is an integral part of the historical method. Yet the use of critical method itself has long been considered as a preliminary step for historical research, while the disciplines that examined documents were assigned to the role of ' auxiliary sciences' of history. The impact of the linguistic turn on the Humanities in the 1960s and 1970s represented a first change in this regard, compelling historians to develop a new hermeneutic of historical texts as the elements and products of ' discourses', a concept famously developed by Michel Foucault (1972 [1969] (1996) is commonly considered as a theoretical initiator for this ' archival turn ' (De Vivo, 2013;Walsham, 2016). For Derrida, archives are the product of political power through the process of archiving, and this archival knowledge itself generates new forms of power over us, by deciding which information remains accessible to us. The 'power of the archives' therefore structures and preconditions the historical truth itself (Schenk, 2013). In the domain of the history of crimes and criminal justice, the linguistic and archival turns have been important steps for the understanding of legal and court records. Since the development of social history that provoked a new interest in these sources, scholars were used to read judicial records ' against the grain', in order to extract information about the society in which they were produced, which has little to do with the reasons why these documents were recorded -what Dietmar Schenk (2013: 41) calls ' outwitting the archive'. Emmanuel Le Roy Ladurie's (1980 [1975]), Carlo Ginzburg's The Cheese and the Worms (1980Worms ( [1976), or Natalie Zemon Davis' The Return of Martin Guerre (1983) are some examples of classical microhistory masterpieces. In contrast, the linguistic and archival turns led scholars to reconsider these sources as a matter of study on their own, revealing the internal logic of the institutions that produced them. This new epistemological approach of criminal records ' along the grain' (Stoler, 2009) therefore offered a better understanding of the discourses, ideologies, and rituals operating within the judicial process (Gauvard and Jacob, 2000;Arnold, 2001;Goodich, 2006).

Montaillou
Going back to the case of Ywain Voet, the source that preserved the story of his crime is a charter of pardon or 'remission letter' granted by the Duke of Burgundy, Philip the Good. Remission letters ('lettres de rémission' in French or 'brieven van remissie' in Dutch) were legal documents in which a king or a prince granted pardon to criminals who petitioned for his mercy. For the late medieval period, the most numerous and Verreycken: 'En nous humblement requerant' 5 homogenous series of remission letters copied in chancery records are those granted by the kings of France, registered in the Trésor des chartes, and those granted by the dukes of Burgundy, preserved in the registres de l'audience (François, 1942;Gauvard, 1991: 59-109;Verreycken, 2014: 45-65). 2 The power to pardon criminals was in fact a major feature of sovereignty in the late Middle Ages and many lay rulers began to deliver charters of pardon from the 13 th century onwards (Verreycken, 2019). One of the earliest surviving series of royal pardon letters are those granted by the English Crown and recorded in a register of letters patent called the patent rolls, which was created under King John in 1201 (Hurnard, 1969;Lacey, 2009). Yet while late medieval charters of pardon -including those granted by the English kings -were usually brief and formulaic, French and Burgundian remission letters are remarkable documents because of their rich narrative and descriptive content. The preamble of each letter indeed developed an argumentation to support the monarch's decision, which was in fact a copy of the petition or supplication ('supplique') submitted by the pardon recipient, who narrated his crime and explained the reasons why he deserved to be pardoned.
Because of the abundance of their narrative content, remission letters are considered as a source of major importance for the history of late medieval justice, violence, and social conflicts: they allow medievalists to study the transcripts of petitions produced by the 'silent masses' (Würgler, 2002). Remission letters first encountered the interest of social historians who saw in these sources ' a spontaneous, living, and sometimes confused testimony, whose human authenticity appears when reading it' (Bourin and Chevalier, 1981: 246). Natalie Zemon Davis' Fiction in the Archives (1987) has however been decisive for the development of a critical attitude vis-à-vis the content of remission letters. The crime stories preserved in remission letters were copied from the petitions submitted by the future pardon beneficiaries.
These petitions were not neutral observations, but rather constructed narratives combining the self-indulgent testimonies of the petitioners with the legal and rhetorical expertise of chancery clerks and lawyers who helped them to elaborate their supplications. For Davis, these narratives therefore had fictional qualities in the sense that they were embellished stories designed to catch the monarch's pity and demonstrate that the crime of the petitioner was remissible. Following Davis' example, criminal justice historians thereafter developed a source-critical approach regarding the narratives of pardon letters, even though they have not given up on analysing these sources ' against the grain', in order to access the 'stories of ordinary people's habits, cultural assumptions, and social practices' as well as the ' customs, behaviors, and actions of often semiliterate people who left us no other forms of writing' (Arnade and Prevenier, 2015: 5). Scholars also developed a growing interest for the petitioning practices in general (Heerma van Voss, 2002;Dodd, 2007;Dodd, Musson, and Ormrod, 2009;Ormrod, 2009), examining the strategies of persuasion adopted by the petitioners who asked for a favour or the redress of an injustice (Dodd, 2011b;Smith and Killick, 2018). In many aspects, these recent works on petitioning are in line with the debates among criminal justice historians regarding the 'voices' of criminals preserved in legal records. They also offer some interesting perspectives on the construction of multiple discourses in the sources, resulting from the interactions between the petitioners and the clerks and legal advisors who helped them to elaborate their demand.
The aim of this article is to offer a re-examination of the crime stories preserved in late medieval pardon letters, considering the theoretical framework of the last 20 years on interpreting petitions and legal records. The research is based on a corpus of almost 1,000 remission letters granted by the king of France and the duke of Burgundy during the second part of the 15 th century (Verreycken, 2018b), whose narrative preambles will be analysed to underline the petitioning strategies developed by the supplicants and their legal advisors in order to elaborate a narration supporting the demand for pardon. I will also compare these strategies with those adopted by the petitioners for pardon in late medieval England, in order to illustrate how the legal, institutional, and political context influenced the elaboration of the crime narratives. The writing of petitions: A plurality of voices Surely, the 14 th and 15 th centuries were ' an age of the petition par excellence' (Dodd and Petit-Renaud, 2015: 240), as every subject had the right to beg the sovereign for a grace, a favour, or the redressing of an injustice, whereas most decisions taken by the monarchs were initiated by a written request. Despite the popularity of petitioning in the late Middle Ages, a major obstacle for the study of these documents is that they only occasionally survived in the records. At the chanceries of the king of France and the duke of Burgundy, every petition for pardon that was accepted by the monarch was transcribed in the narrative preamble of the remission letter, after the formula of notification 'let it be known to all those present and to come that we have received the humble supplication of…' ('savoir faisons a tous presens et a venir nous avoir receu l'umble supplicacion de…'). However, no original petition survived for the medieval period, which suggests that it was destroyed after the charter of pardon was sealed and recorded in the Trésor des chartes or the registres de l'audience, or when the demand of the petitioner was rejected. It is only from the 16 th century onwards that some petitions for pardon can be found in the Low Countries, in the archives of the Habsburg Privy Council and the Council of Brabant (Rousseaux and Mertens de Wilmars, 1999;Vrolijk, 2004). In comparison, the situation in the royal chancery of late medieval England was the opposite. Unlike French and Burgundian remission letters, English pardon letters did not contain a preamble based on the petition submitted by the supplicant. Every charter immediately began with the dispositive clauses of pardon: 'let it be known that by our special grace and our free will we have pardoned, remitted and acquitted […]' ('sciatis quod nos de gratia nostra speciali ac ex certa scientia et mero motu nostris pardonavimus, remisimus et relaxavimus'). However, Helen Lacey (2018: 40) estimates that approximatively 800 petitions for (or relating to) pardon from the 14 th century survived in the so-called 'Ancient Petitions series' in the National Archives, 3 and there are at least one hundred more to be found in the 15 th century (Verreycken, 2018b: 140). In total, 900 petitions is a modest number 3 For more information on the Ancient Petitions series, see Dodd (2009).

Verreycken: 'En nous humblement requerant' 8
considering that thousands of pardons were granted by the Crown every decade, but it allows to make some hypotheses on how the petitions were elaborated.
A petition for pardon had hypothetically at least two authors: the petitioner or supplicant, who elaborated a narrative of his crime based on his experience and his own interests, and a literate man, who was familiar with the norms of the chancery and transformed the supplicant's testimony into a written petition. The petitioner might also have sought the advice of lawyers or notaries to elaborate the arguments of his supplication (Davis, 1987: 15-25;Gauvard, 1997;Texier, 2014: 200-4). The study of original petitions in England and the Low Countries reveals that most of them were written by various experienced hands. Sometimes they also featured dialectal expressions, which indicate that they were not exclusively drafted by chancery clerks but also by local scriveners or notaries who were consulted by the supplicant (Haskett, 1993;Vrolijk, 2004: 310-36;Killick, 2018). The language used in the petitions is a complex question that could not be discussed in the available space here, but it should be noted that a petition was not necessarily written in the native language of the petitioner, which again suggests that he could not elaborate The multiplication of contributors in the process of writing a petition raises a question of authorship regarding who elaborated the narrative of the crime. The first person responsible for composing a petition was the supplicant. In a society of orality and public performances of narrations, even the most unlearned petitioners Verreycken: 'En nous humblement requerant' 9 had some storytelling experience that allowed them to elaborate their narratives (Davis, 1987: 18-19;Texier, 2014: 202). The occasional presence of insults such as 'son of a bitch' ('fils de putain') in some remission letters also illustrates the willingness of the petitioners to offer a direct account of the words they heard or pronounced, thus reinforcing the credibility of their testimony (Haemers and Demeyer, 2019). On the contrary, one can suppose that the legal vocabulary and the chancery formulae frequently used in the petitions were the results of the intervention of the clerks and scriveners who assisted the supplicant. However, it may be possible that some of these technical arguments had been suggested by the petitioners themselves. Michael Clanchy's study of English records has challenged the idea of a strict separation between literacy and illiteracy in the late medieval period by showing that the ability to write was not limited to clergymen: knights, merchants, and even peasants were progressively implied in the 'written business', and so they developed some literacy skills (Clanchy, 2013(Clanchy, [1979). Similarly, Anthony Musson has shown that there were various occasions when subjects could acquire some elementary knowledge of law (2001). As a letter patent, a pardon or remission letter was supposed to be read by anyone who asked to see the proof that the petitioner had been pardoned. Some charters may also have been displayed or read publicly (Petrowiste, 2004: 330;Lecuppre-Desjardin, 2010: 158). Therefore, the publicity of petitions and charters of pardon should not be neglected neither.
On the other hand, the clerks and scriveners who received the testimonies of the supplicants and transcribed them may also have played an important role in the shaping of the crime narratives contained in petitions (Hurnard, 1969: 230-32;Gauvard, 1997). They were experienced writers, trained to the ars dictaminis -the rhetorical art of composing a letter -and they had access to chancery formularies It is usually admitted by historiography that pardoned criminality was mainly masculine: women rarely represented more than one percent of the petitioners (Arnade and Prevenier, 2015: 305;Gauvard, 1991: 300;Muchembled, 1989: 19;Paresys, 1998: 17). Though any subject from every strata of the medieval society could obtain a charter of pardon, most beneficiaries were labourers, artisans, or merchants. [during the last wars in the marches of Picardy, the aforesaid supplicant was always in our service in the company of his master; and in our service he 14 ADB, B 1693 fol. 92r-93r. 15 ADN, B 1703 fol. 34v-35. loyally went to the wars in such a way that, when he was in the company of his master, he was hurt in different parts of his body, including an arrow shot in his head by those who led the enemy party, because of which his person was in great danger.] 17 Yet how could we trust these discourses on the self, considering they were elaborated by the petitioners and their advisors to support a demand for pardon? As Arnold argues, individuals who confessed and elaborated self-reflexive discourses in the context of a judicial process consciously or subconsciously endorsed the role attributed by the institution they faced (1998: 384). This does not mean that these discourses were a complete invention, because the individuals possessed a certain degree of agency in producing their speeches, but they nevertheless acted within the legal and moral categories of the institution (Laurendeau, 2010: 18-21). In the case of the pardoning process, petitioning before a king or a prince was first and foremost an act of submission before the monarch's authority in order to obtain his mercy. For Claude Gauvard, the 'self' described in a supplication was a sujet ideal, an idealized version of the petitioner as a humble and repentant subject ('humblement requerant', as it is frequently written at the end of the preambles of French remission letters) (Gauvard, 1991: 849-93). This concept is similar to what literary theorists call the 'persona', the figure of the author as it is portrayed in self-descriptive texts (Cherry, 1998). For the petitioner, adopting the stereotype of the good subject to shape their personae was an absolute necessity to secure the monarch's sympathy and to make his case remissible. Petitioners who evoked their good reputation, poverty, youth, or military service knew that these were some of the qualities that facilitated the grant of a pardon (Beaulant, 2018: 306-13;Verreycken, 2014: 103-4, 228-33). A good example of this strategy of 'self-identification' (Brubaker, 2004: 41-42) to positive stereotypes is the remission letter obtained by the squire Jehan de Thiboutot in 1481.
At the beginning of his supplication, de Thiboutot insisted on the fact that since an early age he had fought for the king and the res publica ('pour le bien et utilité de nous et de la chose publique'). 18 In late medieval France, the res publica was a fundamental concept in the political theory of royal government, as an equivalent of the emerging state (Petit-Renaud, 2001: 72-79). But whereas the king regularly evoked the res publica in the preamble of his legislation, 15 th -century pardon recipients rarely employed this concept. In the case of de Thiboutot's supplication, the use of this concept from royal propaganda followed the squire's agenda: he begged to be pardoned for the insults he pronounced against the king six years ago, which was assimilated to a crime of lèse-majesté (Hoareau-Dodinau, 1984). Did de Thiboutot truly view himself as a servant of the state and the community or was this a purely rhetorical argument to create a positive persona? Did the squire have the idea to mention the public thing in his petition or was this suggested by a clerk or legal advisor? Probably no one can answer these questions. What seems clear, however, is the fact that in this petition, the reference to the service to the res publica was a discourse of ' calculated conformity' (Scott, 1985: 241) used to balance the offence committed against royal majesty.

Narrating the crime: Making violence acceptable
Not all the crimes committed by late medieval people necessarily involved physical violence or bloodshed. Theft, insults, money counterfeiting, and even debt were condemned by law and justice and are well-documented in court and chancery records (Claustre, 2007;Prétou, 2011;Toureille, 2006). In England, there have been several attempts by the Commons to restraint the royal pardon, as when the 1328 statute of Northampton restricted royal pardon to homicides committed in selfdefence or by accident, but three other statutes promulgated in 1330, 1336, and 1340 deplored that this measure was not enforced (The Statutes of the Realm, vol. 1, 1810: 257, 264, 275, 286). Similarly, in France, after Jean II was captured at the battle of Poitiers in 1356, the General Estates pressed the Dauphin to promulgate the 1357 royal ordinance in which he declared that he would no longer grant remission letters for murder, mutilation, ravishment or rape, arson, or breach of peace and safeguard (Ordonnances des roys de France de la troisième race, vol. 3, 1732: 129). Jean and his successors ignored this prescription and continued to pardon these crimes until the middle of the 16 th century (Gauvard, 1991: 75). Even though there was theoretically no limit to the power to pardon, it appears that in late medieval France and the Burgundian Low Countries most pardoned crimes were homicides and assaults, and their presence in the records continuously increased until they exceeded 80 percent of cases from the 1450s onward (Muchembled, 1989;Gauvard, 1991;Paresys, 1998).
Homicides also predominated in pardon letters in 13 th -and 14 th -century England, but it is more difficult to have a general picture of the pardoned criminality in the 15 th century, for the Crown began to regularly grant pardons for general offences without specifying the crime committed by the petitioner in the charter. 19 There have been numerous discussions among scholars on the reasons why late medieval monarchs privileged to pardon homicides rather than other categories of crimes. The most frequent explanation is that the power to pardon was used as a remedy for the 'rigor of justice' ('voulans […] misericorde preferer a rigueur de justice' is a recurrent formula in the dispositive of remission letters). While legislation and customary law stipulated that any sort of homicide should be punished by death, those committed under mitigating circumstances such as self-defence or accident were considered more acceptable. The king or the prince, unlike his justices, could base his judgment on the notions of equity and mercy or be inspired by the principles of Roman law according to which the will to harm (the 'dol') predominates. Therefore, he preferably granted his pardon for accidental homicide or in self-defence because it avoided an injustice caused by an overly strict application of law. 20 This procedure was particularly routinized in England, for royal justice had the possibility to condemn someone for excusable homicide and then recommend him for a pardon 'de cursu' or 19 These 'global' pardons represented around 39 percent of individual charters of pardon in the time of Edward IV (Verreycken, 2018b: 371). They continued to predominate under Henry VII and until the early years of the reign of Henry VIII. According to Kesselring (2003: 76), they represented 16 percent of individual charters of pardon during the total Tudor period (1485-1503). 20 On the notion of equity as an exception to common law at the English court of Chancery, see also Klinck (2010: 13-40).
Verreycken: 'En nous humblement requerant' 17 ' of course' (Green, 1976;Lacey, 2009: 21-22). The monarchs thus used their power to pardon as an instrument of social control in order to distinguish acceptable forms of violence that were legally condemned but socially accepted, and unacceptable forms of violence that were considered as inexcusable by both law and society (Gauvard, 2005: 264-82).
We have seen that petitions for pardon are careful constructed narratives complying both legal and diplomatic requirements. The first obligation for any supplicant was to provide an accurate description of the crime for which he or she asked forgiveness. In France, the ordonnance cabochienne of 1413 stipulated that petitioners 'must express to us and our counsel the case of their request […] [and] the manner, quality, circumstance of the offences' ('ilz expriment bien au long à nous et à nostre conseil le cas duquel ils nous feront la requeste, […] la manière, qualité, circonstance des délitz') (Coville, 1891: 142). French and Burgundian petitions for pardon therefore usually offer a detailed account of the context of the crime, following the basic three-act structure of a story.  In comparison to French and Burgundian remission letters, English pardon letters were formulaic documents that did not contain any narrative preamble.
However, they employed a rich vocabulary to qualify the crime committed by the petitioner -murder (murder), homicide (homicidium), trespass (transgressio), offence (offensa), misprision (mesprisio), forfeiture (forisfactura), etc. In contrast, French and Burgundian remission letters usually referred to the offence itself simply as 'the aforesaid case'. This difference was due to specificity of the English common law and legal system, wherein a significant part of the criminal procedure was dedicated to the legal qualification of the nature of the infraction itself. As such, the Crown's decision to grant or not to grant a pardon letter was partially based on a legal analysis of the category of crime committed by the supplicant (Green, 1976;Texier, 2014 Although French, Burgundian, and English petitions for pardon offered embellished crime stories shaped by rhetorical strategies, one should not conclude that these documents were entirely fictional. When the royal or ducal council received a petition for pardon, they sometimes commissioned a local officer to make an inquiry in order to check the supplicant's statements when it was suspicious. Moreover, in France and in the Burgundian principalities, once a remission letter was granted, it had to be endorsed by a local justice. The 'entérinement' or ratification trial led to a second inquiry and the victim's party could refute the contents of the remission letter. A letter that was found 'subreptice et obreptice' (fallacious and concealing) was rejected by the court, the pardon was nullified, and the criminal prosecution of the supplicant could start again. 36 In that sense, 34 TNA, SC 8/227/11320. 35 TNA, SC 8/251/12527. 36 For examples of ratification trials in which the victim party or the witnesses contested the content of a remission letter, see Arnade and Prevenier (2015: 173-221) and Beaulant (2018: 485-90). A similar procedure existed in late medieval England, where every pardon letter should be proclaimed by a local court, after the beneficiary had made an arrangement with the victim's party. However, the justice had no power to reject the royal pardon except when it suspected the charter to be falsified or the beneficiary to be an impostor (Hurnard, 1969: 59-66;Lacey, 2009: 25, 91).
Verreycken: 'En nous humblement requerant' 21 people who petitioned for mercy could embellish their stories in order to make their cases pardonable, but they had no interest in making a false testimony when they wrote their supplications, because they risked seeing their demands rejected, or their remission letter cancelled. Therefore, we can consider with Pierre Braun that the narratives in the remission letters were a ' dressing of the truth' that fit with the social and legal norms of pardonable violence, but was also provable in court (Braun, 1984: 209;see also Beaulant, 2018: 309). Once a remission letter was granted and ratified, the petitioner's testimony became the judicial truth. From the criminal historian's point of view, this story, even if it was not completely accurate, contained information about how social interactions were perceived through a certain culture of violence and honour. As a performative legal document, its content can also be interpreted as an indication of how the supplicant elaborated a discourse to exonerate himself while facing royal or princely justice. The 'voice' of the petitioner, affected by the intervention of clerks and lawyers who gave him advice, was the product of a legal context and a formulaic language that must be studied on its own.

Conclusion
'Crimes and criminals, as well as their policing and detection, are themselves rooted firmly in narrative', criminal historians Anne-Marie Kilday and David S. Nash (2017: 3) have recently written. This statement is appealing for scholars investigating pardon letters and petitions for pardon: these sources are supplications that required the petitioners to develop a storytelling based on legal and social stereotypes in order to make their cases pardonable. The preamble of the French and Burgundian remission letters, as well as the English petitions from the Ancient Petition series, are the products of the codified interactions between rulers and ruled. As such, they were not the neutral depiction of the deeds of the supplicants, but a complex argumentation based on rhetorical and legal strategies inspired by epistolary manuals and chancery formularies. Reading these sources ' against the grain', through the filter of a legally constructed narrative, allows historians to partially access to the social realities of the past. But examining how and why this narrative was elaborated also offers a chance: to see as far as possible how people worked their way through dimensions of norms and relationships, through conflicting demands, ambivalent fears and other emotions, how men and women gave these meaning, what narrative forms this took and what this meant in a particular context (Fulbrook and Rublack, 2010: 271).
In that sense, I agree with the conclusions of Tomislav Popić (2019) in his contribution to the present Special Collection about the simplification of realities in late medieval civic court records from Dalmatian cities, which is to say that scholars must consider the interests of both power holders and litigants when reading legal records. The discourses they preserve are the products of the interactions between individuals and institutions. They tell us as much about the agency of the individuals who elaborated them as the legal, administrative, and ideological context in which they were produced.