Revisiting ‘Part One’ of the so-called Très ancien coutumier of Normandy: comments in advance of a new edition and translation of the text[1]

William Eves

A new edition and English translation of the Latin text of ‘Part One’ of the so-called Très ancien coutumier of Normandy is to be published by the Jersey and Guernsey Law Review. In advance of this publication, this article discusses the manuscript tradition of the Très ancien coutumier and how this collection of legal material has been treated by editors in the past. It then examines the treatise comprising ‘Part One’ of the Coutumier in more detail, before providing an overview of the new edition and translation of this text.

1 The importance of custom as a source of Guernsey and Jersey law is well-known, as is the fact that this custom can be traced back to the law of the medieval Duchy of Normandy and the norms and practices which crystallised in the ducal era of the province. Texts which provide a direct witness to law and legal practice in Normandy in this period are, however, somewhat scarce. They are therefore of great historical and legal significance.[2] Among the most important of these is the work which forms the first part of the collection of legal material now known as the Très ancien coutumier.

2 It has been widely held that the so-called Très ancien coutumier contains two separate works. The first, and earlier part is a law-book commonly thought to have been produced at the very end of the twelfth, or at the very beginning of the thirteenth century. The second part of the collection is generally held to be a separate treatise, dating to c.1220. Both works were originally composed in Latin, although a thirteenth-century Old French translation of the Latin text of the entire Coutumier survives, as do fragments of a Norman dialect version.[3]

3 The most recent edition of the Latin text of the Coutumier was published by Ernest-Joseph Tardif in 1881.[4] The same editor published an edition of the Old French version in 1903, including fragments of the Norman dialect version, and this remains the most recent edition of the French version of the text.[5] Tardif’s critical editions have benefitted scholars for many years. Nevertheless, the edition of the Latin text of the first part of the Coutumier, as found in Tardif’s 1881 publication, has several deficiencies, and a new edition of this work has been long overdue.

4 Such an edition of the Latin text of the first part of the Coutumier will shortly be published by the Jersey and Guernsey Law Review, with a parallel English translation.[6] The purpose of this article is to provide an overview of this treatise and the forthcoming publication. It first outlines the manuscript tradition of the Très ancien coutumier and how this collection has previously been treated by editors. It then discusses in more detail the first treatise found in the Coutumier. The new edition of this text is then discussed. By its nature, this article takes the form of an abridged discussion of several matters treated at length in the introduction to this new edition. Readers are therefore directed to this introduction if they desire further elaboration on any point discussed below.

The printed history and manuscript tradition of the Très ancien coutumier

5 Somewhat counter-intuitively, before examining the manuscript tradition of the Coutumier, it is necessary to begin with a discussion of two nineteenth-century printed editions of the text. This is because much of our historical understanding of the work was originally derived from these publications.

6 First, in 1839, Ange-Ignace Marnier published what he called Les anciens Établissements et anciennes Coutumes du duché de Normandie.[7] This was a printing of the Old French translation of the Coutumier contained in Bibliothèque Sainte-Geneviève ms 1743, a manuscript dating to the end of the thirteenth century.[8] In this manuscript, as reflected in Marnier’s edition, both treatises are combined as one and there is no indication that the entire text is anything other than a single work.

7 Shortly afterwards, in 1848, Léopold Auguste Warnkoenig published a broadly equivalent Latin version of this entire text, again as if it were a single work.[9] Warnkoenig gave his text the title ‘Statuta et Consuetudines’ (‘Statutes and Customs’). The precise textual source of this edition is not, in fact, known, and it has been suggested that Warnkoenig may have been relying on an early-modern edition of the work.[10]

8 The impression given by these publications was therefore that the work now known as the Très ancien coutumier was a single treatise, existing in both Latin and Old French versions. However, in 1869 the German scholar Heinrich Brunner argued, based on his reading of Marnier and Warnkoenig’s printed editions, that the work in fact contains two separate treatises.

9 Brunner’s argument was compelling. In short, he explained that some subjects appear twice, once near the beginning, and once more towards the end of the collection. It might be supposed, if this were a single work, that the later treatment of topics which had already been considered was intended to expand upon the earlier discussion. Instead, we see a repetition of earlier topics, presented as if this is the first time they are being discussed. Furthermore, there are no references in the later text to corresponding provisions in the earlier material.[11] Brunner also noted some variations between certain rules discussed in the earlier and later parts of the treatise, reflecting developments over time, which suggest that the later material was written independently, at a later date.[12]

10 Brunner, using the Latin text of Warnkoenig’s edition, suggested that the first treatise ended with a chapter titled ‘De iuramentis’ (‘Concerning Oaths’). This is the final chapter of a section of the text concerning legal complaints heard by the seneschal. Immediately following this chapter is an account of an inquest from the reign of Henry II of England, followed by the text of a 1135 Constitutio (enactment) of King Henry I concerning the breach of the truce and peace of the Church. This is followed by another Constitutio, this time of King Richard, concerning the privileges of the clergy. This material was, Brunner thought, an appendix to the first treatise.[13] The second treatise then began with the words ‘Prius tractandum est de possessione’ (‘First, possession ought to be considered’) which introduce a series of chapters on disputes concerning land.[14]

11 Brunner’s view that the Coutumier comprises two separate works is supported by an examination of the manuscript evidence. There is only one manuscript which transmits the entire Coutumier as if it were a single work. This is Bibliothèque Sainte-Geneviève ms 1743, the manuscript containing the Old French version of the text used by Marnier to produce his edition. We should note, however, that in this manuscript the first few chapters are missing, although this is only a small amount of material.[15] This is the only Old French version of the Coutumier that we have, although, as mentioned, some fragments of a Norman dialect version do also survive.

12 As for the Latin version, almost the entire Latin text of the Coutumier is transmitted within the manuscript BnF (Paris) Latin 11032. The text is not, however, presented as a single, continuous work. Rather, it has been split into numerous smaller pieces, which have been intercalated throughout a copy of the later thirteenth-century treatise known as Le Grand Coutumier de Normandie.

13 We do not know of any other manuscript which transmits the entire Latin text of the Coutumier. However, two manuscripts transmit the complete Latin text of the later material in the Coutumier. These are BnF (Paris) Latin 18368 and BnF (Paris) Latin 4653. Notably, in both these manuscripts this text begins with the chapter which Brunner thought was the opening of the material appended to the first part of the Coutumier, that is, the chapter concerning the inquest from the reign of Henry II. The text then includes the remainder of this ‘appendix’ material, before reaching the comment ‘Prius tractandum est de possessione’ and then moving on to the chapters concerning disputes over land.

14 Although the precise textual source is unknown, Warnkoenig’s Statuta et Consuetudines seems to have relied to some extent on all the above manuscripts. Sainte-Geneviève Ms 1743 and BnF 11032 seem to have been used as a justification for treating the Coutumier as a single work.[16] The edition also contains references to BnF Latin 18368 and BnF Latin 4653.

15 Tardif examined these manuscripts when he produced his editions of the Latin and Old French versions of the Coutumier. Although Tardif named the whole work Le Très ancien coutumier de Normandie, he agreed with Brunner that the text did indeed contain two distinct treatises. Guided by the manuscript evidence, Tardif thought that the chapter concerning the inquest made in the time of Henry II formed the beginning of the second treatise, rather than the beginning of an appendix to the first work as Brunner had thought.[17]

16 Producing an edition of the Latin text of this second treatise was relatively straightforward. Tardif had two manuscripts, BnF Latin 18368 and BnF Latin 4653, each beginning the text with the same chapter, and each containing the complete subsequent work. He also had access to the same material, albeit intercalated throughout another text, in BnF Latin 11032.

17 Tardif was faced with a greater challenge concerning the first treatise in the Coutumier. Lacking any manuscript which contained the continuous Latin text, he reconstructed the work by searching for fragments of text intercalated throughout BnF Latin 11032 which corresponded to provisions in the Old French translation as contained in Bibliothèque Sainte-Geneviève Ms 1743, having concluded, quite reasonably, that the text preceding the chapter on the inquest of Henry II in this manuscript was a translation of the first treatise. Once he had found the material in BnF Latin 11032, he arranged it in the same order as it appeared in Bibliothèque Sainte-Geneviève Ms 1743. Tardif found that BnF Latin 11032 contains all the chapters which are present in the French manuscript apart from four, which in the French text have the titles: ‘De partie de frères’, ‘De vente de bois’, ‘De terre donée’, and ‘De mehaing’. Tardif also found five additional fragments of text in BnF Latin 11032 which also seemed to belong in this treatise. He kept these five fragments in the order in which they appeared in BnF Latin 11032 and placed them at the start of the treatise, reasoning that they represented material which had been lost from the beginning of the French version.[18]

18 Tardif dated the two treatises comprising the Très ancien coutumier from evidence contained in the text; the first to c1199–1200, the second to c1220. In dating the first treatise, Tardif noted that it implies that Richard I of England is now dead (Richard died in April 1199). He also suggested that the treatise implies that the seneschal William fitzRalph is still alive (William died in June 1200).[19] Tardif dated the second treatise to between 1218 and 1223. He reasoned that it is later than spring 1218 because it contains a reference to an act of the Norman Exchequer passed in Easter of that year. This does not appear to be an interpolation as it connects logically with the preceding text. Tardif explained that the treatise is earlier than mid-1223 because it refers to Philip II of France as a living king (Philip died in July 1223).[20]

19 Seemingly unknown to Tardif, while he was working on his 1881 edition, another manuscript containing a complete Latin text of the first treatise of the Très ancien coutumier exists. This manuscript, held in the Vatican Library, first seems to have come to the attention of French scholars through Alexandre Tuetey’s report on his visit to Rome, contained in the 1880 issue of Archives des missions scientifiques et littéraires. However, the manuscript is mentioned only in a footnote as something meriting further investigation, with reference to it containing, amongst other things, ‘jura et instituta Normannica’ (‘Norman laws and institutes’). No explicit mention is made to any text of the Très ancien coutumier.[21] The contents of the manuscript were subsequently described more fully by Lucien Auvray in 1888.[22] By the time he was working on his 1896 edition of the later Normandy treatise, the Summa de Legibus (the Latin version of Le Grand Coutumier de Normandie), Tardif certainly knew about this manuscript and the fact that it contained the Latin text of the first treatise of the Coutumier, because the manuscript also contains a copy of the Summa de Legibus which he made use of for his edition of this later treatise.[23] Regardless of the precise timing of Tardif’s discovery of this manuscript and¾ most importantly¾its contents, this revelation of course came too late for his edition of the Très ancien coutumier.

20 This Vatican manuscript has something of a storied history. Produced in either the later years of the thirteenth century, or the first half of the fourteenth, it contains various texts relating to the law and administration of medieval Normandy.[24] We know from a sheet of parchment glued to one of the front folios that it became part of the famous collector Paul Pétau’s library. Following Pétau’s death in 1614 it was subsequently obtained by Queen Christina of Sweden (1626–1689; r 1644–1654). The manuscript is referenced in the catalogue of her library produced by Isaac Vossius following her abdication in 1654.[25] When Christina died in Rome in 1689, most of her books passed to the Vatican Library but this manuscript, along with 71 others, was instead deposited in the Vatican Archives. A few years later, 53 of these 72 volumes, including the present manuscript, found their way into the hands of Baron Philip von Stosch, a Prussian art and antiquities collector (and possible forger) who for a time also operated as a spy for the British against the ‘Old Pretender’ James Stuart, who had taken up residence in Rome.[26] It is likely that von Stosch obtained these volumes through dishonest means, probably with assistance from some employees of the Vatican.[27] When von Stosch died in Florence in 1757, the Vatican was invited to examine his collection of books and the Vatican librarian, Cardinal Domenico Passionei, thus travelled to Florence to purchase 52 of these 53 volumes. No questions seem to have been asked about the provenance of any of these works.[28] The present manuscript was among these 52 volumes recovered by the Vatican, and it has now acquired the designation Ott Lat 2964 of the Vatican Library’s Ottobonianus collection.[29]

21 In contrast to BnF Latin 11032 (the only other manuscript which transmits the Latin text of the first treatise in the Coutumier), the text in Ott Lat 2964 runs in uninterrupted sequence. It also contains material corresponding to the four chapters which can be found in Sainte-Geneviève Ms 1743, but which cannot be found in BnF Latin 11032. As noted above, in the French text these are given the titles: ‘De partie de frères’, ‘De vente de bois’, ‘De terre donée’, and ‘De mehaing’. They are found in Ott Lat 2964 under the headings ‘De partitione inter fratres et non de sororibus’ (‘Concerning distribution between brothers, and not sisters’), ‘De venditione nemorum’ (‘Concerning the sale of woods’), ‘De terra data’ (Concerning land that has been sold’), and ‘De duellis’ (‘Concerning trials by battle’). Furthermore, the beginning of the text in Ott Lat 2964 also contains the material found in BnF Latin 11032 that Tardif placed, without any guidance available from the French version, at the start of his edition, and in the same order. Likewise, the text concludes at the end of the section concerning legal complaints made to the seneschal with the chapter titled ‘De iuramentis’. This further supports Tardif’s view that this is the end of the first treatise, and that a second, separate work begins with the account of the inquest from the time of Henry II.

The content of the first treatise

22 The content of the first treatise within the Coutumier may be summarised as follows. The text begins with a list of duties each new duke of Normandy swears to uphold. It then briefly turns to issues concerning excommunication, before moving to matters concerning family property such as dower, marriage portions, wardship, and inheritance. Following this, the text discusses matters of ducal jurisdiction, such as justice on the highways and the duke’s peace. It then discusses recognitions (which also fall under the duke’s jurisdiction). Recognitions were a form of legal procedure in which a panel of, where possible, twelve ‘recognitors’ would provide a sworn verdict on a specific question put before them. Different types of recognition existed to address different types of wrong. Following this discussion on recognitions, the treatise considers procedural issues relating to the hearing of pleas in the duke’s court, and ducal jurisdiction over certain types of disputes. The treatise then briefly returns to matters of family and inheritance, before discussing a prohibition on the sale of woods in certain border regions. It then turns to certain legal aspects of homage. Following this, homicide is discussed, before the text turns to procedure in criminal matters more generally, and then moves to discuss procedure in certain civil cases. This is followed by a discussion of the rights and duties arising from a relationship of lordship between two parties. There is then a chapter on usury, which is followed by a discussion of the crime of rape, and then the provision of sureties in relation to accusations of homicide and other criminal offences. Following this, we return once more to the pleas that fall under the duke’s jurisdiction and the operation of ducal courts. There is then an abrupt change in topic, as the treatise discusses issues arising from religious institutions claiming or holding land in alms. The following chapters then return to matters of jurisdiction in relation to both ducal and seigniorial courts, before considering the exercise of customary rights in forests and other locations, and subsequently turning to issues surrounding the construction of a mill or mill-pool on someone’s land. The final part of the treatise discusses reforms implemented by the seneschal, who can be identified as William fitzRalph, in response to various injustices that had arisen throughout the duchy.

23 We do not know who was responsible for the work, and the question of authorship is further complicated by the possibility that the text which survives has been subject to various modifications over time. Modifications to the text are partly suggested by the disorder of he treatise, apparent from the above outline. However, we must not rush to conclusions on this fact alone, as there are other medieval treatises that are just as much, if not more, disorderly but still seem to be the work of a single author.[30] Nevertheless, other evidence from the text also suggests that the treatise has undergone considerable modification and expansion.[31]

The potentially composite nature of the text

24 One suggestion about the potentially composite nature of the text was made by Jean Yver in 1971.[32] Building upon comments first made by Brunner, Yver argued that a significant part of the treatise comprises of a record of ducal ordinances, which can be identified in the text by the words ‘statutum est’—‘it has been established’ (concerning a legislative act)—or a construction such as ‘ne quis audeat’/’ne quis presumat’—‘no one should dare’/‘no one should presume’ (concerning a ducal prohibition).[33] Other chapters, such as those concerning recognitions, may also derive from ducal ordinances but not be presented in this way.[34] Yver then suggested that the compiler of the work, or a later redactor, added a second layer of text to this ‘core’ material.[35] Yver did not pursue the suggestion that the treatise is a composite work in great detail. Nevertheless, a considerable amount of other evidence exists to support the view that the work contains not just one, but several layers of text.

25 Additional evidence for the composite nature of the work, which may or may not be connected to Yver’s suggestion, is provided by the references to a ‘scriptum generale’ which appear in certain places in the treatise, seemingly referring to a ‘general text’ of some sort.[36] Paul Viollet suggested that these references to a ‘scriptum generale’ refer to a separate introductory work or ‘préface générale’ which was partially integrated into the text which now survives.[37] It is also possible that references to the ‘scriptum generale’ refer to a text that formed the ‘core’ template of the work as a whole, rather than merely existing as a prefatory work. If this latter suggestion is correct, the parts of the text which refer to this ‘scriptum generale’ would be later additions to the work.[38] For example, the chapter beginning ‘In scripto generali prenotatur . . .’ (‘It has been mentioned in what has been written before on this matter . . .’), which discusses the inheritance of sisters,[39] and perhaps the following chapter, which discusses marriage portions,[40] may have been intended to supplement other parts of the treatise concerning the inheritance and marriage portions of sisters.[41] Likewise, the chapter which discusses issues which could arise from land held in alms, which begins ‘In scripto generali predictum est . . .’ (‘It has been said in what has been written before on this matter . . .’),[42] may be an addition that was intended to provide more information on the earlier discussion about recognitions concerning land held either as a lay fief or in alms.[43] It is possible, on this analysis, that this ‘scriptum generale’ refers to the ‘original’ work that is based upon a collection of ducal ordinances, as discussed by Yver.[44]

26 Evidence that other chapters may also be later additions to the text are discussed at length in the introduction to the new edition, including the possibility that the chapters missing from BnF Latin 11032 but found in Ott Lat 2964 (and Sainte-Geneviève Ms 1743) are subsequent additions to the treatise, and that the scribe who produced BnF Latin 11032 worked from a manuscript which lacked this additional material. Likewise, the final part of the treatise is given particular attention. As Yver suggests, the earlier parts of the treatise may be based around a collection of ducal ordinances, but the final chapters of the treatise focus exclusively on the legal reforms implemented by the seneschal.[45] These chapters are written in a distinctive, narrative style which differs from much of the preceding material. A past practice that had caused injustice is described, and a case concerning real people is often used to illustrate this point. We are then told what the seneschal has done to rectify the problem. Other stylistic differences also exist. For example, this part of the treatise often reports on what the seneschal has said, whereas this reporting of speech is not found elsewhere in the work. Likewise, new vocabulary is used, such as ‘placitator’ (‘pleader’), which is not encountered elsewhere in the treatise. It is therefore possible that this final section of the treatise has a separate provenance to the earlier material in the work, a fact which¾as will be seen¾is significant for our dating of the text.

27 In addition to the possibility that various chapters have over time been added to a ‘core’ treatise, there is also evidence that a commentary, perhaps originally beginning as a gloss, has become incorporated into the text which now survives. This commentary has a moralising tone and is sometimes introduced by a rhetorical question. For example, a chapter concerning the crime of rape includes a discussion about allegations of rape which are made by the accuser to force the accused to marry her.[46] The chapter explains that if these allegations are not supported by certain forms of evidence, they will not be entertained in court. We then find the comment:

‘Quare? Quia multe sunt mulieres male maligno spiritu perturbate, que vellent vitam suam in casum ponere, ut amasium suum, quem odio habent, possent interficere innocentem.’ (Why? Because there are many women, wickedly stirred up by malicious spirit, who would be willing to put their life in moral danger in order to be able to destroy their innocent lover, whom they hate.)

28 This statement has the appearance of a marginal comment that has found its way into the main text. As it stands, it breaks the flow of the chapter and, despite appearing abruptly, fails to explain coherently the rule that it addresses. Contemporaries may have been concerned that a woman might allege rape in order to force the accused to marry her, but it is unclear why hatred for the man would be a motivation for this.

29 There are other examples of this style of commentary appearing somewhat awkwardly in various chapters, quite probably the result of it being inserted into the main text from the margin where it originated.[47] Furthermore, in one chapter this style of commentary engulfs the text to such an extent that is unclear whether the chapter as a whole is derived from this putative gloss. This is the chapter concerning wardship of fatherless heirs.[48] Brunner noted the chapter’s ‘rhetorische Erguss’, and a reviewer of Marnier’s Établissements et coutumes also commented that it was written ‘avec le verbiage d’un glossateur’.[49] Again, rhetorical questions are used by the writer, although in this instance much of the chapter is structured around them and the overall result is one of coherence, rather than confusion. The chapter begins with the question: ‘Orphanus heres, quem oportet esse in alicuius custodia, quis custodiet eum?’ (‘It is necessary for a fatherless heir to be in someone’s custody—who will have custody over him?’). The following discussion is then structured around a succession of further questions: ‘Mater? Non [. . .] Quis ergo custodiet eum? Consanguinei? Non. Quare?’ (‘The mother? No [. . .] Who, therefore, will have custody of him? His relatives? No. Why?’). Eventually, it is explained that lords, rather than certain relatives, should have wardship over heirs. A rhetorical question is used to explain the rule: ‘Domini autem quo modo possunt odio habere quos nutrierunt?’ (‘For their part, how can lords hate those they have raised?’).

30 The treatise is therefore most likely a patchwork of legal material. The form in which it now survives is probably the result of two processes: (i) the combination of at least two texts, and (ii) the incorporation of a gloss. The relationship between these two processes is, however, unclear. A gloss may have been added to the treatise once the ‘main’ texts which comprise the work had already been combined. Alternatively, some of the constituent parts of the treatise may have been glossed before they were joined with other material. It is also possible that one of the ‘main’ texts was itself a substantial gloss on the original core material. It is, however, impossible to establish with any certainty the exact process which led to the formation of the text as it survives today.

The date of the treatise

31 The potentially composite nature of the work also complicates the question of dating. As we have already noted, Tardif dated the treatise to between April 1199 and June 1200, noting that the treatise implies that Richard I of England is dead, and that William fitzRalph is still alive. However, Tardif’s dating of the work demands some reconsideration. Although Tardif suggested that the text implies that William is still alive, there is evidence to contradict this. Furthermore, the potentially composite nature of the text, which Tardif did not consider in detail, further complicates matters, as evidence in various parts of the text suggesting a terminus post quem or terminus ante quem cannot be applied to the work as a whole.

32 The source of Tardif’s assertion that William fitzRalph was living when the text was written is found in the part of the treatise concerning the reforms of the seneschal. Throughout this part of the treatise, the seneschal whose reforms are being discussed is never explicitly identified by name. Tardif argued that this unnamed seneschal must be William, and the fact that he is not explicitly identified is because, during his lifetime, it would be obvious that he was ‘the seneschal’ whose reforms were being discussed. If William were dead, and his two successors, Guérin de Glapion and Ralph Tesson, were in office, Tardif assumed that the author would wish to identify by name which of these seneschals was responsible for the reform in question.[50] It is true that William FitzRalph is probably the seneschal in question. William enjoyed a long period in office and had a significant influence on Norman law.[51] However, Tardif’s argument as to why William is not named directly is not entirely convincing. The author of this part of the work may have been content simply to record that these reforms were introduced through the office of the seneschal. Likewise, the author may have thought that William’s long period in office and influence on the administration of justice would make it obvious that he was the seneschal in question, even after his death. Furthermore, Tardif’s suggestion that William is still alive is directly contradicted by a reference within this part of the treatise to events happening ‘in the time of William the seneschal’. It is clear from the context of the chapter that this ‘William the seneschal’ is the same seneschal whose reforms are being discussed in this part of the work, almost certainly William fitzRalph.[52] As reference to ‘the time of’ an individual suggests that this time has now passed, this comment suggests that the author of this part of the treatise was writing after William’s death in June 1200, not before it.[53]

33 A different clue towards a date for this part of the treatise is found only in the Vatican manuscript Ott Lat 2964. As such, it was not discussed by Tardif in his 1881 edition of the text. It was, however, noted by Paul Viollet in 1906.[54] This is found in the chapter concerning the activities of the duke’s serjeants, where it is explained that the seneschal ordered that, if any of these serjeants accuse people unfairly, they shall be sent to prison ‘until he who is duke sets them free’.[55] The text in Ott Lat 2964 then adds, ‘that is to say, the king of England or France’ (‘scilicet rex Anglie vel Gallie’). The Latin manuscript that Tardif had access to, BnF Latin 11032 instead contains the comment ‘that is, the king of France’ (‘scilicet rex Gallie’). The Old French translation found in Sainte-Geneviève Ms 1743 also comments that this would be the king of France (‘ce est li rois de France’).[56] Tardif thought that this comment was clearly an interpolation added at a date after the Capetian conquest of Normandy, when the rule of the French Crown over the province was certain. However, the comment as it appears in Ott Lat 2964 connects the text to a period when it was much less certain who would have control over Normandy. Although there were subsequent attempts by English kings to regain Normandy, this period is likely to be the later years of the 1200–1204 conflict between John and Philip II. If the remark ‘scilicet rex Anglie vel Gallie’ in Ott Lat 2964 formed part of the original text, were can therefore date the text to around 1203–1204.[57] If the comment was instead a marginal note that has now become incorporated into the main work, the original text most likely pre-dates this period, although, for reasons explained above, it post-dates the death of William fitzRalph in June 1200.[58]

34 The above analysis relates to the part of the treatise concerning the reforms of the seneschal, which may have been composed separately from other parts of the work. However, evidence of a broadly similar date of composition is found throughout the other parts of the text.

35 We have already encountered Tardif’s terminus post quem of 1199. This is derived from a reference in a chapter located near the beginning of the work to a time when Richard I was in possession of Normandy, and therefore alive.[59] This may mean that all the other text found in the treatise has a broadly similar terminus post quem to the seneschal material, in that it post-dates 1199. However, it is also possible that this comment is a later addition to the chapter in which it is found. The composite nature of the treatise also means that we cannot use this comment as a definite terminus post quem for all the remaining ‘non-seneschal’ parts of the work. Some material in the treatise may therefore have been composed before 1199.

36 As for the terminus ante quem of the other ‘non-seneschal’ material in the treatise, there is evidence that at least some parts were written no later than the early thirteenth century.[60] There are several references to the unilateral ordeal of trial by hot iron or water, a procedure which fell out of use following the prohibition of clerical involvement in 1215. These references therefore date this part of the text to the years before this prohibition.[61] Likewise, the part of the treatise which discusses the Norman equivalent of the English action of darrein presentment refers to the pre-1207 procedure that was used for these actions, rather than the procedure introduced by Philip II in October of that year, which suggests a pre-1207 date of composition for this part of the text.[62] Other parts of the treatise suggest a pre-1204 date of composition. The chapter concerning the crime of rape explains that an accuser will be fined if defeated in judicial combat, whereas William le Breton writes that shortly after the Capetian conquest of Normandy in 1204 Philip ordered that a losing accuser in a criminal matter was to suffer corporal punishment.[63] Likewise, the part of the treatise concerning dispossession discusses an older form of the procedure used for the recognition of novel disseisin, rather than the updated procedure which came to be used in both England and Normandy.[64] Both Brunner and Tardif suggested that the procedure in Normandy would have been updated to follow the new English procedure while the Anglo-Norman realm remained intact, so the use of the old procedure dates the text to before 1204.[65] A pre-1204 date for much of the material is also suggested by the fact that the ruler of Normandy is referred to throughout as the ‘duke’. There are some exceptions to this, in which reference is made to ‘the king’, seemingly referring to the king of France, but these may be later additions to the text.

37 As discussed in the introduction to the new edition, none of these arguments is entirely conclusive. There is evidence, for example, that the later procedure for the Norman version of darrein presentment did not become widely used until the 1220s.[66] Likewise, there is evidence that procedure in some recognitions was updated to follow English practice even after the Capetian conquest of Normandy, which means that the Norman action of novel disseisin may not necessarily need to have been updated before 1204.[67] We must also treat references to ‘the duke’ with caution. The later thirteenth-century Summa de Legibus, composed after the Capetian conquest, also contains many references to the duke of Normandy, but explains that ‘the duke or the ruler of Normandy is said to be he who reigns over all the duchy, which dignity the lord king of France retains for himself’.[68] Nevertheless, the weight of the above evidence does suggest that, in addition to the ‘seneschal material’, much of the remainder of the treatise was composed before c. 1215, perhaps before 1204.

The significance of later thirteenth-century developments

38 The above arguments about the date of the work are derived from evidence within the text. However, Nicholas Vincent has recently cautioned against relying on this evidence by suggesting, tentatively, that the treatise may in fact be a later thirteenth-century ‘reimagining’ of the earlier laws and customs of the Normandy’s legal past.^[69] There is not space here to describe Professor Vincent’s argument in detail. In short, however, he suggests that the text was influenced by the content of England’s Magna Carta, in particular the 1225 reissue of the Charter which was circulating in revised form in Normandy in the thirteenth century. The ‘reimagining’ derives, he suggests, from the efforts of Norman lawyers in the later thirteenth-century to protect Normandy from the legislative and fiscal encroachments of the Capetian kings. Nonetheless, despite this suggestion, a late twelfth or early thirteenth century date for the various parts of text comprising the treatise remains tenable, for reasons given in the introduction to this new edition.[70] However, Vincent’s suggestion remains significant. As outlined above, the treatise is most likely a composite work, and survives only in later thirteenth- or early fourteenth-century manuscripts. It is therefore possible that various parts of the text were brought together later in the thirteenth century for precisely the reasons Vincent suggests, and that this process gave the work the form in which it now survives.

The new edition and translation of the text

39 The existence of two manuscripts, BnF Latin 11032 and Ott Lat 2964, each containing the Latin text of this work, raises questions about how the readings in each manuscript differ from one another. As may be expected, the process of copying a treatise by hand invites the possibility of scribal error, for example by oversight or misreading, and changes may also have been made deliberately by the scribe. It is therefore unlikely that any two manuscripts will transmit precisely the same text. There are indeed some significant variations between the text in BnF Latin 11032 (and therefore the contents of Tardif’s edition) and the text in Ott Lat 2964. We have already encountered one such variation in the appearance of the comment ‘scilicet rex Anglie vel Gallie’ in Ott Lat 2964, rather than ‘scilicet rex Gallie’ as found in BnF Latin 11032 in our discussion of the date of the work.

40 There is not space in this article to discuss all the other variant readings in depth. This discussion is set out in full in the introduction to the new edition. However, a few examples will provide and a good indication of the extent of some of these variations. For example, the very first chapter of the work concerns the oath that is sworn by a new duke of Normandy. Ott Lat 2964 explains that, amongst other things, this oath binds the duke to protect (or maintain peace on) the highways (‘kemina’), whereas the reading of BnF Latin 11032 has uses the word ‘bonam’ rather than ‘kemina’, substantially changing the reading of the passage to explain that the duke is bound to maintain good peace (‘bonam pacem’).

41 Other significant variations arise throughout the treatise. For example, concerning homicides, Ott Lat 2964 provides a fuller discussion of what is to happen to a mother who murders her son (she is to be burned).[71] This manuscript also provides a better reading than BnF Latin 11032 of the length of time an heir might allow his inheritance to be possessed by another before losing the right to challenge this possession in court (‘twelve harvests’, rather than what Tardif thought was ‘twelve months’ in BnF Latin 11032).[72] Also significant is the fact that Ott Lat 2964 and BnF Latin 11032 frequently provide different readings of numeric values. Amongst other examples, we find variations in chapters which deal with the amount payable as a fine following defeat in judicial combat. BnF Latin 11032 consistently stipulates that ‘xl’ (40) s. should be paid, whereas Ott Lat 2964 consistently stipulates that payment should be ‘lx’ (60) s.[73] Likewise, the sum of money owed by knights who have fallen into the mercy of the duke is given as at least ‘x’ (ten) s. in Ott Lat 2964, and at least ‘xx’ (twenty) s. in BnF Latin 11032.[74]

42 On occasion, Ott Lat 2964 provides additional content which is lacking in BnF Latin 11032. We have already noted that it contains the four chapters which are found in Sainte-Geneviève Ms 1743, but which cannot be found in BnF Latin 11032. In addition to these substantial parts of text, we also find shorter passages in Ott Lat 2964 which are not present in BnF Latin 11032. To provide but one example, in the chapter concerning justice on the highways, both manuscripts describe the procedure used to prove that someone has been wounded on the highway.[75] Ott Lat 2964 then contains a comment, not found in BnF Latin 11032, describing a special procedure to be used if someone is wounded ‘within the banlieu’ (‘infra banleucam’) in which the offender is to defend himself through his oath ‘forty-eight-handed, according to the law of our land’ (‘per iurationem suam xlviii secundum legem patrie’).

43 Despite the differences touched upon above, BnF Latin 11032 and Ott Lat 2964 also display some shared errors. For example, in both manuscripts the text omits the crucial word ‘non’ from a passage which seeks to explain that minors will not be compelled to answer in court concerning their right to the inheritance they have just entered until they have reached the age of majority.[76] This omission therefore reverses this statement of a general principle of law, well-established at the time. Likewise, to provide just one further example, a chapter concerning juries begins in both manuscripts with the word ‘Videlicet’ (‘Namely . . .’) rather than ‘Licet’ (‘It is permitted . . .’), a nonsensical reading in the context of the chapter, which seeks to explain that it is permitted (‘licet’) for a jury to be held in anyone’s court concerning movable property or inheritance.[77]

44 Shared errors such as these, together with the shared presence of some interpolated material, discussed further in the introduction to the new edition, suggest that Ott Lat 2964 and BnF Latin 11032 derive, at some point in the history of the transmission of the text, from a now-lost common source, somewhat removed from the ‘original’ version.[78] Nevertheless, the significant textual variations discussed above place some distance between these witnesses in the manuscript tradition. Indeed, the presence in Ott Lat 2964 of material likely to have been omitted from BnF Latin 11032 through homeoteleutic error provides strong evidence that the text in the former is not a direct copy of the latter.[79] Likewise, the presence of material in BnF Latin 11032 that is omitted through the same type of error in Ott Lat 2964 shows that the former is not a direct copy of the latter.[80]

45 The above summary concerns the relationship between the Latin manuscripts. What of the relationship between the Latin manuscripts and the Old French version of the treatise contained in Sainte-Geneviève Ms 1743? In his 1903 edition of this Old French version of the treatise, Tardif demonstrated that the text in Sainte-Geneviève Ms 1743 is closer to Ott Lat 2964 than it is to BnF Latin 11032.[81] The French version follows Ott Lat 2964 more closely than BnF Latin 11032 in the way it divides the text into chapters and the headings it provides for these chapters. The French text also tends to follow the reading of Ott Lat 2964 when the readings of the Latin manuscripts diverge.[82] Text missing from BnF Latin 11032 but present in Ott Lat 2964 is also found translated in the French version.[83] It is therefore clear that the French translation is not based on the text in BnF Latin 11032. Neither, however, is it based directly on Ott Lat 2964. Although the arrangement of chapter headings and the division of chapters in Sainte-Geneviève Ms 1743 often follows that in Ott Lat 2964, there are also some significant differences between the two manuscripts.[84] Furthermore, some passages of text omitted from Ott Lat 2964 by homeoteleutic error are translated in the French version of the text.[85] This leads to the conclusion that the French version of the treatise relied upon a now-lost Latin witness which was closer to Ott Lat 2964 than to BnF Latin 11032. The author of the Norman dialect version of the Coutumier, from which various fragments also survive, also seems to have consulted a manuscript similar, but not identical to Ott Lat 2964 for the text of the first treatise, with some use also made of the French text in Sainte-Geneviève Ms 1743.[86]

46 The new edition of the first treatise within the Très ancien coutumier, soon to be published by the Jersey and Guernsey Law Review, is based on the text in Ott Lat 2964, although the text in BnF Latin 11032 assists with deficient readings, as does the Old French version where appropriate. This new edition of the treatise thus provides several benefits over Tardif’s 1881 edition. Most obviously, it has been able to make use of one more manuscript witness to the Latin text than was available to Tardif. As this new edition naturally takes into account the variation in readings between the texts, it thus corrects unsatisfactory readings in Tardif’s edition. Furthermore, the new edition benefits from the fact that the text in Ott Lat 2964 exists as an uninterrupted whole, rather than it being fragmented and intercalated within another work. As such, this text confirms the correct sequence of chapters, and also supplies the Latin text of those chapters which do not appear in BnF Latin 11032. An additional benefit of an edition based on the reading of Ott Lat 2964, rather than BnF Latin 11032 is that the text contained in Ott Lat 2964 is a closer representation to the text that influenced two vernacular translations, that is, the Old French and Norman dialect versions of the work. It therefore possibly represents a more widely known form of the treatise than does the text in BnF Latin 11032, even if we cannot be sure that it represents the ‘original’ work more accurately than the text in the latter manuscript.

47 This new edition and translation also seeks to make the work more accessible. The parallel English translation of the Latin text is designed to assist students and those unfamiliar with the Latin of medieval legal treatises. Furthermore, unedited transcriptions of both Latin texts as contained in BnF Latin 11032 and Ott Lat 2964 are also provided in an appendix, set out in parallel columns, so readers who prefers to work with the Latin can directly compare the text contained in each manuscript for themselves.

48 A further advantage of this new publication relates to our more general perceptions of the treatise. By publishing this text separately from the second treatise also contained in the so-called Très ancien coutumier, the new edition emphasises the fact that the first treatise should be treated as a distinct text in its own right. Despite the fact that Tardif recognised that the Coutumier contained two separate texts, his decision to combine them and publish the whole collection under the title of Le Très ancien coutumier de Normandie allows the unwarranted impression that this material is all part of the same work to persist. This is exacerbated by the fact that Tardif named the first treatise ‘Pars prima’ (‘The first part’), and the second treatise ‘Pars altera’ (‘The second part’) of the Coutumier. Even more regrettably, he continued his numbering of the chapters from the beginning of the first treatise to the end of the second, so that the first chapter of the second treatise numbered chapter 66, following the final chapter (chapter 65) of the first treatise.

49 A new, separate edition of the first treatise will therefore help to oust this artificial connection between the texts, a connection which initially seems to have arisen solely from the translation in Sainte-Geneviève Ms 1743 of material supposedly comprising a single coutumier into Old French, and which is unsupported by any other manuscript evidence. Furthermore, in order to reinforce the entirely separate nature of the two works within the Coutumier, we have rejected the use of ‘Pars prima’ for the title of the treatise in this new edition, preferring instead the title given to the text in Ott Lat 2964: ‘Antiqua consuetudo Normannie’ (‘The Ancient Custom of Normandy’).

50 The Antiqua consuetudo Normannie is clearly a text which gives rise to many questions, and there is still much to be done on the nature of the work and the manner in which it was produced. It is hoped that, in addition to addressing certain deficiencies in the way the work has been published over the years, and enhancing our understanding of the laws and customs of ducal Normandy, this new edition and translation will facilitate further study of the text itself. The tools provided by this new publication, such as the parallel English translation of the edited Latin text, and the inclusion of transcriptions of the text found in BnF Latin 11032 and Ott Lat 2964, should assist with these endeavours. More broadly, it is hoped that this new edition and translation will increase the accessibility of the work for anyone who has an interest in the legal history of the duchy of Normandy and, by extension, the legal history of the Channel Islands.

Dr William Eves is an Assistant Professor in Law at the University of Nottingham, with a background in law and history. He was previously a postdoctoral research fellow in the University of St Andrews’ Department of Medieval History, and a Scouloudi doctoral research fellow at the Institute of Historical Research (School of Advanced Study, University of London).