14 Pierrepont at a crossroads of literatures

An instructive parallel between the first branch of the Karlamagnús Saga, the Dutch Renout and the Dutch Flovent

Abstract: In the French original of the first branch of the Karlamagnús Saga [= fKMSI], in the Dutch Renout and in the Dutch Flovent – three early 13th century texts from present-day Bel- gium – a toponym Pierrepont plays a conspicous part (absent, however, from the French models of Renout and Flovent); fKMSI and Renout even have in common a triangle ‘Aimon, vassal of – Aie, his wife – Pierrepont, their residence’. The toponym is shown to mean Pierrepont () near in all three texts. In fKMSI, it is due almost certainly to the intervention of one of two Bishops of Liège (1200−1238) from the Pierrepont family, and in the other two texts to a similar cause. Consequently, for fKMSI a date ‘before 1240’ is proposed.

According to van den Berg,1 the Middle Dutch Flovent, of which only two frag- ments are preserved,2 was probably written by a Fleming (through copied by a Brabantian) and can very roughly be dated ‘around 1200’ on the basis of its verse technique and syntax. In this text, Pierrepont plays a conspicuous part without appearing in the French original.3 In the first fragment, we learn that King Clovis is being besieged in Laon by a huge pagan army (vv. 190 ss.). To protect their rear, the pagans build a castle at a distance of four [presumably French] miles [~18 km] from Laon. Its name will be Pierlepont (vv. 208 ss.). At the beginning of the second fragment, Christian relief forces coming from the east under the command of Clovis’s son Flovent have already conquered the castle and temporarily entrenched themselves in it. But soon, leaving only two squires in it (vv. 444 s.), they hurry on to the relief of Laon. ‘Those from Pierle- pont’ now refers to these forces (vv. 432 s.; cf. vv. 338, 354, 370). Thanks to them, the pagans suffer a decisive defeat. Towards the end of his story, the author must have narrated how King Clo- vis (or Flovent as his successor) invested one of his warriors with the new castle. In sum, then, the Flovent informs us how Pierlepont originated in Merovingian times. A look at the map suffices to make it clear beyond any doubt that the

1 Van den Berg, 1987, pp. 13 s., 32. – For full bibliographical data, see the References below. 2 Edited by Kalff 1886 [1967], pp. 180–203. 3 This fact was briefly mentioned but not analyzed by Loke, 1906, p. 113, n. 3. For the French original see Andolf’s edition, 1941; the part we are interested in comprises at most vv. 2271– 2534.

Note: First published in: Neophilologus 89 (2005), 587–603.

Open Access. © 2019 Gustav Adolf Beckmann, publiziert von De Gruyter. Dieses Werk ist lizenziert unter der Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 Lizenz. https://doi.org/10.1515/9783110615692-014 356 Renaut de author means Pierrepont (Aisne), 15 km north-east of Laon, so the form Pierle- pont with its -l- is simply a playful literary variant. Situated in the middle of the marsh of Saint-Boëtien, Pierrepont commanded the passage through that marsh and was therefore fortified by the tenth century.4 Still today, some villages up to 8 km away add ‘-lès-Pierrepont’ to their names. But all this means compara- tively little, since other roads from Laon to the north-east by-pass the marsh on its north and south. So we may conclude precisely from its at best mediocre importance that Pierrepont must have been introduced into the story by (or on the initiative of) someone with an intense personal interest in that place, in other words, a native speaker of French; this interest, however, becomes visible to us only when the story reaches present-day . *** The same curious constellation – Pierrepontian enthusiasm in a literary text from Belgium – recurs in two more texts from the same time. At the outset, let us examine the first branch (henceforth: KMSI) of the Karlamagnús Saga (henceforth: KMS). The importance of the 13th century KMS for the study of the French genre of chansons de geste is undisputed.5 Strictly speaking, what students of literature are interested in, is not the branches of a KMS themselves, but the French-language originals underlying them, and a KMS branch may be particularly interesting if the French text is no longer extant. This holds true for the first branch which we may define as an epic biography of Charlemagne,6 breaking off at the moment of the Spanish campaign when the Chanson de itself takes over.7 As a plethora of details suggests, the French original (henceforth: fKMSI) of KMSI hails from French-speaking Belgium,8 but nothing in detail is so far known about its milieu of origin nor its more exact date within the first 60 years of the 13th century.9

4 Gysseling, 1960, s. v.; Matton, 1871, s. v.: castrum Petrae Pontis a. 938. 5 Editio citanda for the branches covered in it: Loth, 1980. I have not seen the facsimile edition of ms. a by Halvorsen, 1989. Translations include: branches I, III, VII and IX into French by Annette Patron-Godefroit in Loth, 1980; branch I into French by Aebischer, 1972, pp. 93–139; the whole KMS into English by Constance B. Hieatt, 1975–80 (the translations so far mentioned have called forth some criticism of details); the whole KMS into French: Lacroix, 2000. 6 Aebischer (1972, p. 13 s.) called it a Vie romancée de Charlemagne, more precisely a Vie et chronique guerrière, mondaine et scandaleuse de Charlemagne et de sa cour. 7 We are not concerned here with the fact that certain later parts of KMS, such as the end of the Roland and the Moniage Guillaume, seem to come again from that biography; see Skårup, 1990, passim. 8 See e.g. Aebischer, 1972, pp. 6 ss., 18, 33, 39. 9 See e.g. Skårup, 1990, passim, in whose opinion this Vie de Charlemagne [= fKMSI] n’a guère pu être antérieure au second tiers du XIIIe siècle. For practical purposes, a terminus ante quem 14 Pierrepont at a crossroads of literatures 357

The events of the first 25 chapters of KMSI take place in a rather narrow strip of land between Bitburg in the Eifel and Tongres to the north of Liège.10 On the order of an angel, Charlemagne had to try his hand at burglary in the company of the master-thief Basin and thereby overheard a conversation be- tween Count Rainfroi of Tongres and his wife about a conspiracy against Charle- magne’s life. Rainfroi named no less than a dozen conspirators: they fall into four groups having their fiefs respectively at Tongres (Belgian Limburg), Waes- Aardenburg (westernmost Flemish-Dutch borderland), Orléanais-Breteuil-Pierre- pont- (central and northern ) and finally Trier-Salm (or Saarburg or Saarbrücken) plus an ambiguous Homb(o)urg (all on the border of or inside the German language area).11 The possessions of the traitors thus surround what implicitly stands out as the land of the faithful, to wit, Wallonia plus the county of Flanders. In full knowledge of the conspirators’ plans, Charlemagne then convoked his vassals from throughout the empire to a diet at Aix-la-Chapelle. About 90 vassals are named, the geographical density of the list being least in Italy and southern France, greatest in the fiefs immediately west of the Rhine and also in the and Ardenne regions, where even smaller fiefs like , Chiny, Durbuy, La Roche and Esch-sur-Sûre are duly registered.12 Two vassals are portrayed with marked deference: the archbishop of (in the Middle Ages the spiritual overlord of the bishop of Liège) and the count of Flanders13 (at the time of the compilation of fKMSI mostly on excellent terms with Liège against their com- mon adversary Brabant). The author also seizes the opportunity to describe Aix-

is 1263, the year of the death of King Hakon IV Hakonarson of who was almost cer- tainly the promoter of the whole untertaking. 10 Twenty-six chapters in Unger’s old edition of 1860. For a full discussion of the geographical aspects see Beckmann, 1973, passim. 11 For the Waes, Old French Waise, the correct Veisa appears in Chs. 4 and 25. Aardenburg is the modern name of medieval Rodenburg. In Hirson, medieval Iriçon and similar spellings, the Norse translator mistook ç for k, writing Irikun.InSalim-, Salen(am)borg (and the clearly mis- taken hapax Salernisborg) the last syllable may, but need not be a clarifying addition of the Norse translator; a dissimilation r > l is attested for Saarburg and/or Saarbrücken (and a vacil- lation between ‘bridge’ and ‘castle’ for Saarbrücken) in (mostly French) sources from the 11th to the 13th century. Hoenborg or similar is one of the Homb(o)urgs near Verviers, , Forbach and Saarbrücken, respectively. Generally speaking, geographical names in KMSI are in a bad state, but by comparing the different Norse manuscripts and using the tools of historical geog- raphy, we can identify most of them – a task whose scope, however, would clearly exceed the space of the present article. 12 Kretest is a mistake for Retest, the old name of Rethel; Chims/Chimz in b (misread as Thuns in B) is Chiny, Dyrbo Durbuy, Eysu Esch, Fjalli translates La Roche. 13 He still resides at (which, in reality, Flanders lost to Philip Augustus in 1184). 358 Renaut de Montauban la-Chapelle (which till 1802 belonged to the diocese of Liège) in conspicuous detail. The conspiration theme – and with it the first half of KMSI – is brought to a close by the seizure, trial and execution of the traitors. So far, the subject matter of fKMSI definitely gives the impression of having been collected in Romance territory to be sure, but close to the German language boundary, espe- cially close to Aix-la-Chapelle. An ideal collecting point would be Liège itself, the eastern bulwark of Wallonia.14 Integrated into the first half on KMSI towards its end is a narrative of five or six chapters15 only loosely connected with the conspiracy and relating mainly how a certain Aimon of Galicia (af Galizu, Galiza, Galiz) finally becomes an Aimon of Pierrepont16 through his marriage. The narrative runs as follows. When they hear of Charlemagne’s accession to the throne, both Aimon of Galicia and Raimbalt of Friesland17 in their respective homelands decide to go to Aix-la-Chapelle to join Charlemagne’s retinue. Since, at the accession of the historical Charles in 768, Galicia was the south-western and Friesland the north- eastern borderland of Christianity, the basic idea of the story seems to be that a few well-inspired warriors even from the ends of Christianity already placed their hopes in Charles.18

14 One might object that precisely a bishop of Liège is absent from the subject matter of KMSI. Is he? In ch. 6 a bishop – evidently a suffragan of the archbishop of Cologne – is called herra Valtir, with the honorary title herra ‘lord, sire, His Eminence’ given only to him among the roughly 200 persons appearing in KMSI! He is called bishop of Intreitt and almost on the same manuscript page bishop of Nasten. Intreitt alone might rather be Utrecht than Maastricht, but both misreadings taken together point to Maastricht, and indeed in the later diocese of Liège, Maastricht was the seat of the bishop until the transfer to Liège in the early 8th century. The author of KMSI may have ignored the exact date of the transfer, or the older title may have struck him as more official. 15 In Loth’s edition, it comprises: version A Chs. 18, 25 (beginning and end), 26–30; version B, Chs. 16, 25, 28–30. It was studied by Aebischer, 1957, pp. 23–54 [= 1967, pp. 35–55]. 16 In the B version of KMSI this place-name is consistently spelt Pirapunt/Pirapont.InAitis first spelt Pirafunt in Chs. 26 and 28, due to a momentary confusion with the better known Pierrefonds (), but then three times Pirapunt in Ch. 30. From a comparison of these read- ings (and from the geographical proximity of Hirson in both versions, A Ch. 26, B Ch. 28, AB Ch. 30), it becomes evident that Pierrepont (Aisne) was originally meant in A, too. 17 In A, Raimbalt consistently appears as Reinbal(l)dr friski, in B Ch. 16 as Reimballdur friske, but from Ch. 24 on as Reinalldr friski. It is evident even for B that we have to do with the Raimbalt de Frise of so many chansons de geste, Rabeu lo Fris of the , who continues the name (albeit in a somewhat irregularly Frenchified form) and the nationality, if nothing else, of Friesland’s foremost hero, Duke Radbod (in Frisian Redbad), contemporary of Charles Martel. 18 Galicia as a geographical concept was quite familiar to Frenchmen by 800, among other things because Alfonso II called himself a vassal of Charlemagne, see Einhart’s Vita Karoli Ch. 16; it became still more familiar with the pilgrimages to Compostela. For the basic geo- 14 Pierrepont at a crossroads of literatures 359

On their way, the two happen to meet,19 and, both being somewhat over- bearing in a question of precedence, they engage in a duel. Fortunately, they soon recognize that their strength is equal, swear each other brotherhood-in- arms and continue together to Aix-la-Chapelle. When together with the other traitors, Folkvard of Pierrepont is executed, his relative Varner starts an open rebellion at Pierrepont. Aimon duly informs Charlemagne of this, and Raimbalt, who has become Charlemagne’s brother-in-law, proves worthy of his new dig- nity by killing Varner in duel. Charlemagne then gives Varner’s widow in mar- riage to Aimon and enfiefs him with Hirson, Pierrepont, “and his own land, Galiza”.20 We hear nothing about the offspring of the couple, but since marria- ges in epics usually do not remain barren, KMSI leaves the reader with the vague idea that Aimon will have ended his days as the founder of a new local dynasty at Pierrepont. To be sure, the question of the genesis of this story merits a few remarks. As early as (the part of) the Chanson de Roland (v. 3073) Rembalt e Hamon de Galice appear together as commanders of Charlemagne’s eighth corps, the Flemings and Frisians. Without any more specific arguments, Aebi-

graphical idea underlying our story, compare Charlemagne’s dream at the beginning of the Pseudo-Turpin: he sees the galaxy extending a mari Frisiae [...] ad Gallaeciam. – However, a different identification of our ‘Galicia’ cannot altogether be ruled out. In her article “Le grand duc Autcarius [...]”, 1991, p. 297, Suzanne Martinet speaks about “Laon avec tout son quartier de la Galise, la Valise actuelle”, unfortunately without giving the slightest reference. In view of the interest of this remark for the present discussion, I contacted the Archives départementa- les of the Aisne department. Their director, Mme Frédérique Pilleboue, kindly informed me (on October 3, 2002) that no forms with G- can be found for la Valise and that this toponym is rather a derivative of Lat. vallis ‘valley’: surviving today only in the names of several alleys, it denoted in the Napoleonic land register a zone between Laon proper and its present-day sub- urb of Vaux, the latter continuing the medieval Vallis subtus Laudunum. Though improbable, Martinet’s proposal will be taken into account in the notes (see n. 19, 20, 22) ‘for safety’s sake’. 19 In reality, two persons following the shortest routes from Galicia (or from Laon, see n. 18) and from Friesland respectively to Aix-la-Chapelle would only meet at their point of destina- tion where, under the eyes of Charlemagne, they could no longer afford to start a private duel. The narrator foresaw this and pretended that Aimon had to go far northward out of his way for lack of a bridge across the Meuse. This would indeed produce the desired effect of meeting Raimbalt since the shortest route from the county of Holland (which still was a part of Fries- land) to Aix-la-Chapelle crossed the Meuse twice. 20 Galiza is mentioned here only by A who probably considers Charlemagne even as overlord of Galicia since in Ch. 22 he is called ʻthe legitimate emperor of the whole worldʼ; but we cannot quite exclude the possibility that A takes Galiza for a region in France (see n. 18, Marti- net’s claim). B replaces ‘and his own land, Galiza’ by ‘and many other possessions’, most probably because he identifies Galiza with the Spanish Galicia (as we do). but feels sure that it lay outside Charlemagne’s empire. 360 Renaut de Montauban scher21 concluded from this that the author of the Roland already knew the whole story as we read it in KMSI. In partial support of his view, one might point to the fact that many of the commanders of Charlemagne’s ten corps do have some geographical affinity to their corps; so possibly in the opinion of the author of the Roland, Aimon de Galice, too, had some specific qualification for leading ‘northerners’. But even for that purpose only his brotherhood-in-arms with Raimbalt of Friesland, not his marriage and his settling down precisely at Pierrepont would be necessary.22 We therefore have no evidence whatsoever that the motifs ‘marriage’ and ‘Pierrepont’ were connected with Aimon de Galice before fKMSI. On the other hand, in this latter text, Pierrepont is unambiguously again the small Pierrepont (Aisne): it is clearly in the vicinity of Hirson,23 and the fact that the wife of Varner of Pierrepont is the daughter of the Count of Laon suggests that Pierrepont is also in the vicinity of Laon. Now, whatever the genesis of the story, the very space allotted to Pierrepont in fKMSI is out of all proportion with its slight importance in a normal all- French (or, for that matter, in a regional Belgian) perspective. And the allotment of space is clearly a factor in the responsability of the compiler of fKMSI, not of any predecessor. The situation thus dovetails well with that in the Dutch Flo- vent: there we had unexpectedly learnt about Pierrepont in Merovingian times, here we unexpectedly learn about Pierrepont in Carolingian times – but both texts come from present-day Belgium, not from the vicinity of Pierrepont. Why then did the compiler give Pierrepont such undue prominence? This time, an answer is possible and, though never given till now, even simple. We have already stated that the first half of fKMSI is likely to have been collected near or at Liège. Who then was Prince-bishop of Liège at the time in question? From 1200 to 1229 it was Hugues de Pierrepont, from 1229 to 1238 it was his nephew Jean d’Eppes, Pierrepont being again the place in the Aisne department and Eppes lying 12 km to the south-south-west of it. A coincidence is practially out of the question; one of the two must have been involved in the formation of fKMSI and evidently could not resist giving the compiler some hints about a tradition in his family. In this context, it is important to note that Hugues’s mother was the daugh- ter of Witer Count of Rethel and the granddaughter of Godefroid Count of Namur. Hugues’s uncle Albert de Rethel rose to the position of archdeacon of St. Lambert in Liège, and it is certainly through Albert’s influence that Hugues

21 1957/1967, passim. 22 Or, alternatively, for the author of the Roland, the Galice would be a region in northern France; cf. n. 18. 23 See n. 16. 14 Pierrepont at a crossroads of literatures 361 made his career in Liège, i.e. in the imperium, and not in his native diocese of Laon in the regnum Franciae.24 In other words, the Pierrepont element in the history of Liège is due to an individual combination of circumstances, it remains an intermezzo of less than 40 years. Let us mention at least two facts that may shed some light on Huguesʼ per- sonality: his resounding military victory over Brabant at Steppes in 1213, and the fact that Gislebert of Mons called him a clericus satis litteratus et discretus.25 Such a man may well have also taken a vivid interest in the battles of old and in the legendary history of his own family.26 As to the date of fKMSI, we may now place the text – allowing a few years for the compiler to complete his work – between the first years of the 13th cen- tury and approximately 1240. I confess that, until there is proof to the contrary, I even tend to exclude the last part of that span of time because I cannot find any influence of the Pseudo-Turpin in fKMSI, though French translations of that work had been available in the North, roughly speaking, from the turn of the century onward and though Hélinand de Froidmont (writing before 1216) and Aubri de Trois-Fontaines (writing till 1241) in their Latin chronicles as well as Philippe Mousket (writing till 1242) in his French chronicle drew on the Pseudo- Turpin most heavily.27 To avoid a misunderstanding, let us have a look at the second half of KMSI, too. It can in most scenes be interpreted as a careful, but unobtrusive prepara- tion of the Chanson de Roland, so that it is almost impossible to say where the information comes from. One exception, again unnoticed till now, is of some interest to us. The compiler has much to say about Charlemagne’s sister Gisla, in KMSI Gilem, Gelem (from the Old French oblique case Gilain). After her brother forced her to commit incest with him, she is married first to Milon d’Angliers, who thus becomes Roland’s ostensible father (Ch. 33), then, after Milon’s death, to (Ch. 51). So far, the events are also known from other legendary or epic sources. But KMSI goes on: ‘learned men’ soon found out that Ganelon and Gisla were relatives to a prohibited degree, so they were separated, and Gisla was married to Duke Efrad and gave birth to an Adalrad and an

24 See de Moreau, 1945, 130–139, the article ‘Hugues de Pierrepont’ in the Biographie nationale [...] de Belgique (7½ columns long) and e.g. the genealogical table at the end of Gade, 1951. See also Poncelet, 1946, introduction. The key-word Namur may also explain the interest shown in KMSI Ch. 24 for the legendary foundation of Namur by the epic Duke Naimes, confidant of Charlemagne. 25 Poncelet, 1946, p. VII. 26 Due to the shortness of his pontificate, I do not dispose of similar information on Jean d’Eppes. 27 See e.g. Moisan, 1987, passim, Walpole, 1947, passim. 362 Renaut de Montauban

Efrad.28 These elements are not known from any nor do they have any narrative function within KMSI, so that one may suspect them of being based on learned local reminiscences. And in fact, for anyone somewhat famil- iar with 9th century genealogies, no doubt is possible. Gisla, sister of ‘King Charles’ [the Bald!], was married to a Duke Eberhard, and among their children are both an Eberhard who died young and an Adalhard who became advocate – for that epoch, one is tempted rather to say: holder − of the abbey of Cysoing in French Flanders a few kilometers south-east of . Eberhard and Gisla had founded that abbey as their ‘home abbey’, the spiritual centre and prospective burial place of the entire family. However, as the main branch of the family soon became rooted in distant north-eastern Italy,29 the abbey went through two difficult centuries, but by the 12th century it flourished again, then partly due to donations by the Counts of Flanders. As most of what is known today about Eberhard, Gisla and their children comes from the early charters of Cy- soing30 (and as by 1200 tombs of members of the family will still have been visible), we may be confident that the information in KMSI also comes from Cysoing. We need not decide whether the monks sincerely believed their Gisla to have been the sister of Charlemagne instead of Charles the Bald or whether they knowingly transferred her to a more glorious age. In sum, then, Liège was probably the most fertile, but certainly not the only collecting-point for fKMSI. Rather, if the materials came from both Liège near the eastern border of present-day French-speaking Belgium and Cysoing slightly beyond its western border, in other words, if the catchment area of the collection was pan-Franco-Belgian instead of just liégeois, then the same is likely for its area of circulation. This means that the Pierreponts’ epical preten- sions may have become known early in the French-speaking regions and circles of Flanders as well.

***

In KMSI, the name of Aimon’s wife, the widow of Varner of Pierrepont, appears twice: it is Aein, in the accusative. Here -ein clearly reflects the Old French end- ing -/-ien of the oblique case of monothematic feminine names. The question

28 Part of the Norse manuscript tradition still has the etymologically correct Efrard (without the dissimilatory loss of the second -r-); see ed. Loth Ch. 51 l. 9, ed. Unger Ch. 54 l. 11. 29 The literature on that family (whose history culminated in the person of the Italian emperor Berengar) is ample. See the articles ‘Cysoing’ and ‘Eberhard 3’ in vol. 3, ‘Unruochinger’ in vol. 8 of the Lexikon des Mittelalters. For a stemma including all persons mentioned above, see Werner, 1967, pp. 447, 452 and the Table at the end of the volume. 30 See de Coussemaker, 1886, nrs. 1–6, pp. 1–11. 14 Pierrepont at a crossroads of literatures 363 then is: which Old French name can hide behind a nominative *A-e or similar? Aebischer31 opted for Aie, in the oblique case Aiien or, in a more usual spelling, Aien. As the reference works on Old French epic names32 show, this is indeed the only possible identification. We may therefore say that the Aimon story in KMSI culminates in a triangle ‘Aimon, vassal of Charlemagne – Aie, his wife – Pierrepont, their residence’. And here, a most curious parallelism seems to have gone unobserved to this day: the same triangle is found in the Dutch verse epic Renout van Montalbaen (which, together with its Dutch and German derivatives, we shall call the Dutch Renout tradition, ‘dR’ for short). Here again, we must go into some detail. According once more to van den Berg,33 the Dutch Renout, preserved only in fragments, was written by a Flem- ing, just like the Flovent, and can be dated by its verse-technique and syntax very roughly ‘before 1225’ or even ‘around 1200’, again much like the Flovent.It ultimately derives from the 12th century French Renaut de Montauban (the rhymed tradition of which we shall refer to as ‘fR’), but differs from it in a great many traits.34 One of them concerns the home of the protagonist and his broth- ers: in fR, their parents Aimon and Aie live at Dordon(e), an imaginary castle not too far from the ; in dR, Aimon and Aie mainly live at Pierle- pont35 – again the Spielform we already know from the Flovent! –, though Dor-

31 Aebischer, 1972, p. 59. 32 Moisan, 1986; still useful due to its handiness: Langlois, 1904. 33 The oldest extant fragments of the Renout seem to have been written in the last quarter of the 13th century. For the text itself, van den Berg first advocated a date ‘before 1225’, later (with the same arguments, it seems) ‘around 1200’; see van den Berg, 1983, p. 222; 1985, pp. 13 and 23; 1987, pp. 13 s. and 34. In fact, I think that in working with his method, deviations of two or three decades will on principle be unavoidable. 34 The standard work on the subject is Spijker, 1990. For a concise presentation of basic problems in French and English respectively see also Spijker 1993 and 1994, for reflections of dR in Dutch folklore Spijker, 2000. 35 It is true that the roughly 15 % still extant of the Dutch Renout do not include a passage in which we might expect this toponym to occur. (Editio citanda for all fragments known before 1939 is Diermanse, 1939 [on which the popular edition by van Maelsaeke, 1966, is based], to be completed by Duinhoven, 1973. For a convenient list of all finds and their contents, see Hogenhout-Mulder, 1984, p. 35, or Spijker, 1990, pp. 271 ss.) But Pierlepont (with minimal vari- ants) does appear in the three traditions deriving from the Dutch verse epic: (a) in the Ripuarian Histôrie van Seint Reinolt, ed. Reifferscheid, 1874, pp. 275 ss., 279 s., 283; (b) in the German verse epic Reinolt van Montelban, ed. Pfaff, 1885, vv. 218, 312, 476, 753, 2149, 3263, 3460 (this text is little more than a faithful, often slavishly close translation of the Dutch verse epic); c) in the Dutch prosification De Historie van den vier Heemskinderen, ed. Overdiep, 1931, pp. 15, 17 s., 23, 29, 33, 53 (2x), 69 (2x), 70 (2x); likewise in the German prosifica- 364 Renaut de Montauban doene, too, still plays a considerable part. Pierlepont is not restricted to formu- las, but is well intergrated into the narrative: Charlemagne’s messengers go there twice, Aimon returns there from the court, Aimon bequeaths the castle to Renout, Aimon’s four sons return there after Louis’ death. In contrast to this, no mention of Pierrepont or similar has ever been found in any manuscript of fR. Consequently, the situation is comparable to the one in Flovent and KMSI: though not situated in Belgium, Pierrepont ‘surfaces’ only in Belgium. It goes without saying that the identity of the two triangles Aimon – Aie – Pierrepont in fKMSI and dR cannot be attributed to chance. Let us first consider its consequences for fKMSI. Remembering that in fKMSI the presence not only of Pierrepont, but also of Aie must be explained, we propose the following sce- nario. In the late 12th century, when the story of the Quatre fils Aimon rapidly became more and more popular throughout northern France, the Pierreponts somehow persuaded themselves – or were persuaded by a flatterer – that they were descended from the protagonists of that story. The particular reason for this is unknown, but also immaterial: in those days, similar pretensions in the nobility were not rare and, once invented, tended to be perpetuated at least by the families concerned. When a Pierrepont became Bishop of Liège, he automat- ically had much better means of disseminating the tradition of his family. A particular opportunity presented itself with the compilation of fKMSI. The com- piler received the bishop’s story with due respect, but reproduced it only up to Aimon’s and Aie’s marriage, excluding the following conflict of their sons with Charlemagne: after all, he had to compile a biography in honour and not in dishonour of Charlemagne. Moreover, it may already have been a Pierrepont, or at the latest, it was now the compiler who remembered that an Aimon already appeared in the Roland and who had the idea (“which Aimon if not ours?”) of hooking the Pierrepont story onto that Aimon.36 tion (mainly based on the Dutch one) Das deutsche Volksbuch von den Heymonskindern, ed. Pfaff, 1887, pp. 9, 11, 15 s., 21, 44, 61. And, finally, the Cologne printing of 1493, preserved in one copy only and not yet reedited, corresponds almost word to word to the Dutch prosifica- tion and, as Dr. Irene Spijker most kindly informed me (on January 6, 2004), also contains the toponym, e.g. in the heading of chapter 9. (For some general information on that printing see Weifenbach, 1999, particularly p. 180 s.) In view of these attestations, no specialist ever doubted – and I think none will ever doubt – that Pierlepont already appeared in the Dutch verse epic quite a few times; see e.g. Loke, 1906, pp. 25, 112 s., Diermanse, 1939, p. 88, Hogen- hout-Mulder, 1984, p. 157 n. 1, Spijker, 1990, pp. 45 s. In terms of the stemma: Pierlepont already appeared in the common archetype of all known texts (including fragments), called ʻvʼ in Hogenhout-Mulder’s stemma (1984, p. 89). 36 Inevitably, in such matters, one’s judgement will depend to a certain degree on one’s opin- ions on the genesis of fR in particular and of the genre of the chansons de geste in general. The above statement is in accordance with the belief, almost universal in our days, that fR 14 Pierrepont at a crossroads of literatures 365

And what does the identity of the two triangles mean for dR? Of course, students of dR have duly registered Pierlepont.37 As early as around 1870, Mat- thes identified it with Pierrepont (Aisne).38 In 1906, Marie Loke knew that there is also a Pierrepont (Meurthe-et-) east of the Meuse, 10 km south-east of Longuyon, but she opted for Pierrepont (Aisne).39 Conversely, in her 1984 Groningen dissertation, Maaike Hogenhout-Mulder40 did not need to come to a decision between the two Pierreponts for her main line of argument, but expressed her view in a footnote that Pierrepont was prob- ably the place in Meurthe-et-Moselle. She points out that in the Dutch Renout (as attested by the Ripuarian Histôrie, the German Reinolt and the Dutch prose Heemskinderen) Aimon bequeaths to Renaut all his possessions, namely “Pierle- pont, Montagut and Falkenstein”.41 According to her, Montagut is the medieval county of Montaigu extending around Marcourt in the Belgian Ardennes, and Falkenstein would probably have to be identified with the castle of that name 5 km north of the Luxemburgish town of Vianden, just across the German bor- der, both Montaigu and Falkenstein being closer to Pierrepont (Meurthe-et- Moselle) than to Pierrepont (Aisne). First, however, around 1100, Bishop Hugues’s great-grandfather Roger, lord of Pierrepont (Aisne), had acquired through marriage Montaigu (Aisne), 13 km farther south, with a 10th century castle ʻperched like an eagle’s nest on the top of a steep hillʼ. He gave it to his younger son Robert so that it was passed on in a collateral line of the family until it reverted before 1200 to the main line in

came into being at some time in the 12th century. However, a traditionalist hardliner, still believing, like Longnon, in an oral continuity of fR from the 8th century onward, would possi- bly consider Aimon de Galice as the father of the Quatre Fils Aimon all along. Even if the latter is nowhere called ‘from Galicia’, hadn’t he at least fought in his youth, according to dR, seven years for a heathen king of ? (See Duinhoven, 1973, verses corresponding to Pfaff, 1885, vv. 2208 s., 2216 ss., Overdiep, 1931, p. 54, Pfaff, 1885, vv. cit., Pfaff, 1887, p. 45.) 37 See end of n. 35. 38 Quoted by Loke, 1906, p. 113. 39 Ibidem – but with a rather weak argument, it must be admitted: in the Dutch prose Heem- skinderen, the name of the Oise does appear (ed. Overdiep, 1931, p. 184), but at such a distance from any mention of Pierrepont (see above n. 35c) that it proves nothing. 40 Hogenhout-Mulder, 1984, p. 157 n. 1. – I am indebted to Dr. Irene Spijker for pointing out this note to me. 41 Ed. Reifferscheid, 1874, p. 277 (-stein); ed. Pfaff, 1885, vv. 753 s. (allein: -stein); ed. Overdiep, 1931, p. 29 (allene, -stene). – Several times, Renaud is also called ‘Count of Merewoud’, with Merewoud appearing only in that title and only in rhyme. Hogenhout-Mulder, 1984, pp. 157– 164, interprets it as Mirwart in the Belgian Ardennes, but shows that it is probably just due to a misreading plus a generalization. 366 Renaut de Montauban the person of Bishop Hughes’ elder brother Robert.42 Under these circumstan- ces, we may assume that Falkenstein, a typical name of castles (‘the falcon’s stronghold’), is used just for the rhyme with the preceding allein because it occurs only once in the story. And, second, let us look at the events of the story. In all of dR, Charlemagne has a permanent residence, ; it is superfluous to give references. He leaves it only for clearcut purposes: for a pilgrimage to Santiago, for expeditions against Aimon’s sons – and, in one more case, rather early in the story. When he is willing to conclude peace with Aimon and to make public amends for his involvement in the murder of Aimon’s nephew Hugues, he invites Aimon to come from Pierrepont to Senlis. There Charlemagne does penance before him, barefoot and clad in wool. After the meeting, he at once returns to Paris and Aimon to Pierrepont. So Charlemagne’s intention was clearly to meet Aimon, if not exactly halfways, then at least at some intermediate place, as is befitting a council of peace.43 And when Charlemagne later invites Aimon to his son’s coronation, the messengers, with Aimon and his sons, return from Pierrepont via Senlis to Paris.44 Now, as the map shows, Senlis is a reasonable intermediate stop between Paris and Pierrepont (Aisne), but not between Paris and Pierre- pont (Meurthe-et-Moselle). We may therefore be confident that in dR, just as in the Dutch Flovent and fKMSI, Pierrepont (Aisne) is meant, at least originally. But since even Pierrepont (Aisne) is not a fascinating place in itself and is distant from the area even more than from the Franco-Belgian area, its appearance in dR once again cannot be ascribed to an author’s disinter- ested imagination. The only alternative in sight is again the ambition of the bishops’ family: it must be ‘somehow’ at work here, too. However, to go beyond this ‘somehow’ is very difficult. That Bishop Hugues or his nephew should have interfered with the formation of the Dutch text itself is improbable for three reasons. First, on general cultural grounds: a Francophone prelate of the early 13th century is unlikely to have been interested in Dutch literature. To be sure, this argument is weakened in our case by the fact (a) that a large part of the

42 Melleville, 1857, pp. 303–306, 310, 317. For pointing out Melleville’s still valuable study to me, I am again indebted to Mme Frédérique Pilleboue, head of the Archives départementales de l’Aisne (letter of March 1, 2004). 43 Ed. Reifferscheid, 1874, p. 275; ed. Pfaff, 1885, vv. 261, 266, 312 s.; ed. Overdiep, 1931, p. 16– 18. 44 This time, the Ripuarian Histôrie, which drastically shortens the whole secular part of the story, suppresses the mention, but we find it in ed. Pfaff, 1885, vv. 908 s., and ed. Overdiep, 1931, p. 32. 14 Pierrepont at a crossroads of literatures 367 bishop’s flock – from Maastricht and Tongres down to the coast – consisted of native speakers of Dutch, (b) that the Renout may be read as the Vita of a mar- tyr, and (c) that by virtue of the Pierrepont motif, the bishop would present himself to his flock as the descendant of a martyr. Second, however, the dialec- tal traits in the Dutch Renout point to Flanders,45 not to Brabant or Limburg. This alone would not be decisive either, because it was not rare in those centu- ries for poet and patron to be from different regions.46 Relations between Liège and Flanders as a rule were friendly at the time in question; Flanders had been playing a prominent cultural, economic and political part on a truly European scale at the latest since the time of Philip of Alsace; and though the family of the counts preferred French, Flanders was productive precisely in Dutch karele- piek from around 1200 onwards.47 So the bishop might conceivably have com- missioned a Fleming. But, third, for the most part, the Renout gives the impres- sion of being based on oral transmission of a French model.48 In contrast to this, if a Francophone patron − and a bishop at that − should really have com- missioned a Dutch author to translate a French text into Dutch, he would cer- tainly have provided him with a manuscript. Taken together, these three arguments suffice to make a direct interference of a Pierrepont bishop in the formation of the Dutch text improbable. We are left with two possibilities: 1) That dR should have the toponym from fKMSI is logically the simplest solution. But, though Pierrepont occupies a conspicuous position in fKMSI, this solution appears psychologically less probable because the four sons of our couple are not even mentioned in that text. 2) So finally: since in fKMSI a Pierrepont interfered, naturally enough, with a French-language text – why should things not be the same in the present case? The whole of fR is represented by 13 manuscripts. If we abide by the dates assigned to them by Thomas,49 only four of them are from the 13th and the turn of the 14th century. In other words, we are far from reliably knowing what the

45 See above n. 33. 46 Names come to mind even from the 12th century such as Chrétien de (writing also for the count of Flanders), Gautier d’Arras (with Arras belonging till 1184 to the counts of Flanders), the later Veldeke. There can hardly be a doubt that the phenomenon was more widespread and that often we simply do not have the twofold information necessary to trace it. 47 Van den Berg, 1987, p. 13. 48 Spijker, 1990, pp. 203–227, 261 s. and other passages listed in the index under ‘orale over- levering’, ‘orale verhalen’. 49 Thomas, 1962, vol. I, pp. 19–136, particularly 136: ms. DZPN. Other scholars assign ms. L, too, to the 13th century. 368 Renaut de Montauban

Renout tradition near the Dutch language boundary was like around or shortly after 1200. Consequently, a Pierrepont may well have succeeded, if only in geo- graphically and chronologically narrow limits, in imprinting his mark on the French tradition. In contrast to fKMSI, we have no clue pointing to the bishops in person, but given their means and their closeness to the language boundary, they should remain at least in consideration.50 One difference between the two texts still deserves a discussion. In KMSI, Aie, widow of Varner of Pierrepont, is the daughter of a count of neighboring Laon – a detail which simply seems to reflect the compiler’s constant preoccu- pation with geographical preciseness. In dR, however, Aie is Charlemagne’s sis- ter, that is, King Pepin’s daughter. In France, this relationship is mentioned briefly by Aubri de Trois-Fontaines51 and, within fR, quite incidentally by the manuscripts D and P:

Et mult en fu loez del boen reis Karlemaine, Mult par aime Renaut filz sa seror germaine.52

According to Thomas’s stemma for the épisode ardennais,53 the mention must belong in the archetype and was lost early. As it stands, brief and functionless, there is no evidence to connect it with the Pierreponts. In dR, however, we see it excellently integrated into the initial part of the plot: after their first feud, Charlemagne gives Aimon his sister Aie in marriage as a pledge of their recon- ciliation. In other words, not the motif itself but its meaningful employment seems to be coextensive with the mention of Pierrepont and so was probably intended to flatter the Pierrepont family: they came to have Carolingian blood in their veins.54

50 None of the possibilities discussed above raises serious chronological problems; see n. 33. 51 Monumenta Germaniae Historica, SS. 23, p. 723. 52 Ed. Thomas, 1962, vol. I, ms. D v. 95; vol. 2, ms. P v. 94; ed. Thomas, 1989, v. 935. On the other hand, in vv. 9423 s. of the latter edition (see Thomas’s note ad loc.), DPNL have it that Charlemagne is related to Aimon’s sons through Aimon himself. The contradiction illustrates how little importance is still attached to this relationship even in those French manuscripts that mention it at all. 53 Thomas, 1962, vol. I, p. 190. 54 Why did fKMSI not know of Aie’s royal descent? One is tempted to answer: because at that time the Pierreponts themselves did not yet know of it. That would make dR slightly younger than fKMSI. 14 Pierrepont at a crossroads of literatures 369

Postskriptum 2018

Der ‘doppelte’ (H)Aimon. Einerseits ist also der Hamon de Galice des Rolandslie- des identisch mit dem Eim (< afrz. [H]Aime, Obl. [H]Aimon ‘Haimon’) von Galiza der Ersten Branche der Karlamagnús Saga (und ihrer afrz. Grundlage, einer ver- lorenen Vie de Charlemagne), welcher ebendort zum Herrn von Pierrepont wird, nämlich durch Heirat mit der Aein (< afrz. Obl. Aien, zum Rektus Aie ‘Aja’) von Pierrepont (Aisne, knapp 20 km nordöstlich Laon). Am Ende der KMS I haben wir also ein noch kinderloses Paar (H)Aimon und Aja ‘von Pierrepont’. Andererseits heißen im afrz. Renaut de Montauban auch die Eltern der Hai- monskinder Ayme/Aymes (Obl. Aymon) und Aie, also (H)Aimon und Aja, hier ‘von Dordone/Dordon’, einem in der Realität unauffindbaren Ort. Im niederländischen Renout van Montalbaen schließlich sind beide Paare identisch: (H)Aimon und Aja von Pierlepont (eine auch anderweitig belegte - iante von Pierrepont), aber ebenso noch von Dordoene, sind die Eltern der Hai- monskinder. Ist die Verschmelzung erst im Renout eingetreten, so gab es also vorher zwei zufällig gleichbenannte Paare namens (H)Aimon und Aja. Nun lernen wir aber das eine Paar nur als noch kinderlos kennen, das andere nur als Eltern von vier majorennen Brüdern. Man kann sich somit fragen, ob nicht – trotz unterschied- licher Lehensangabe – die beiden Paare schon vorher letztlich identisch waren. Und da die eine dieser beiden Erzählhandlungen gute Chancen hat, älter zu sein als das erhaltene Rolandslied, hieße das: der Hamon de Galice des - liedes wäre … der Vater der Haimonskinder, der ja gegen Söhne immer treu zu Karl hielt.

References

Aebischer, Paul, ‘Raimbaud et Hamon. Une source perdue de la Chanson de Roland’, Le Moyen Âge 63, 1957, pp. 23–54, reprinted in P.Ae. Rolandiana et Oliveriana, Geneva: Droz, 1967, pp. 35–55. Aebischer, Paul, Textes norrois et littérature française du Moyen Âge, vol. II: La première branche de la Karlamagnús Saga, Geneva: Droz 1972. Andolf, Sven, Ed., Floovant, chanson de geste du XIIe siècle. Uppsala: Almquist and Wiksell, 1941. Beckmann, Gustav Adolf, ‘Zwischen Trier und – Der geographische Rahmen der altfranzösischen Basinerzählung (Karlamagnús Saga I 1–26)’, in Verführung zur Geschichte, Festschrift zum 500. Jahrestag der Eröffnung einer Universität in Trier, 1473–1973. Eds. Georg Drœge et al., Trier: NCO-Verlag 1973, pp. 60–70. Berg, Evert van den, Middelnederlandse versbouw en syntaxis, Utrecht: HES 1983 (Dissertation Utrecht). 370 Renaut de Montauban

Berg, Evert van den, ‘De Karelepiek. Van voorgedragen naar individueel gelezen literatuur’, in Tussentijds: bundel studies aangeboden aan W. P. Gerritsen ter gelegenheid van zijn vijftigste verjaardag. Eds. A. M. J. van Buren et al., Utrecht: HES, 1985, pp. 9–24. Berg, Evert van den, ‘Genre en gewest, De geografische spreiding van de ridderepiek’, Tijdschrift voor Nederlandse taal- en letterkunde 103, 1987, pp. 1–36. Biographie nationale […] de Belgique, vol. 9, Brussels: Jamar 1886–87. Coussemaker, Ignace de, Cartulaire de l’abbaye de Cysoing et de ses dépendances. Lille: Desclée et de Brouwer 1886. Diermanse, P. J. J. (Ed.), Renout van Montalbaen. Leiden: Brill 1939. Duinhoven, A. M., ‘De Haagse fragmenten van de Renout van Montalbaen’, De Nieuwe Taalgids 66, 1973, pp. 177–201. Gade, John A., Luxemburg in the Middle Ages. Leiden: Brill 1951. Gysseling, Maurits, Toponymisch woordenboek van België, Nederland, Luxemburg, Noord- Frankrijk en West-Duitsland (vóór 1226), 2 vols. Brussels: Belgisch Interuniversitair Centrum voor Nederlandistiek, 1960. Halvorsen, Eyvind Fjeld, Ed., Karlamagnús Saga and Some Religious Texts. [Facsimile edition of ms. A.] Copenhagen: Rosenkilde and Bagger, 1989. Hieatt, Constance, Transl., Karlamagnús Saga. 3 vols. Toronto: The Pontifical Institute of Medieval Studies, 1975–1980. Hogenhout-Mulder, Maaike, Proeven van tekstkritiek. Groningen 1984. (Dissertation Groningen.) Kalff, Gerrit, Middelnederlandsche epische fragmenten, Arnhem: Gysbers en van Loon, 1886, reprint Arnhem, 1967. Lacroix, Daniel W., Transl., La Saga de Charlemagne. Paris: Le Livre de Poche, 2000. Langlois, Ernest, Table des noms propres de toute nature compris dans les chansons de geste imprimées. Paris: Bouillon, 1904. Lexikon des Mittelalters, Munich: Artemis/Lexma, vol. 8, 1986, and 9, 1997. Loke, Marie, Les versions néerlandaises de . : Privat, 1906. Loth, Agnete, Ed., Karlamagnús Saga, Branches I, III, VII et IX. Édition bilingue projetée par Knud Togeby et Pierre Halleux, texte norrois édité par Agnete Loth, traduction française par Annette Patron-Godefroit, avec une étude par Povl Skårup. Copenhagen: Reitzel, 1980. Maelsaeke, D. van, Ed., Renout van Montalbaen. Antwerpen: De Nederlandsche Boekhandel, 1966. Martinet, Suzanne, ‘Le grand duc Autcarius préfigure d’Ogier le Danois’, in Histoire et Littérature au Moyen Âge, Actes du Colloque d’Études Médiévales de l’Université de Picardie ( 20–24 mars 1985). Ed. Danielle Buschinger, Göppingen: Kümmerle 1991, pp. 291–300. Matton, Auguste, Dictionnaire topographique du Département de l’Aisne. Paris: Imprimerie nationale, 1871. Melleville, Maximilien, ‘Notice sur Pierrepont’, Bulletin de la Société académique de Laon 6, 1857, pp. 295–322. Moisan, André, Répertoire des noms propres de personnes et de lieux cités dans les chansons de geste franςaise et les æuvres étrangères dérivées, 2 parts in 5 vols., Geneva: Droz 1986. Moisan, André, ‘Clercs et légendes épiques: Hélinand de Froidmont, Aubri de Trois- Fontaines, Vincent de et la Chronique du Pseudo-Turpin’, Au carrefour des 14 Pierrepont at a crossroads of literatures 371

routes d’Europe: la chanson de geste. Xe Congrès International de la Société Rencesvals ( 1985). Aix: Cuer Ma, 1987, vol. II, pp. 913–925. Moreau, Édouard de, Histoire de l’Église en Belgique, vol. III, Brussels: Édition Universelle, 1945. Overdiep, Gerrit S., Ed., De Historie van den vier Heemskinderen. Groningen: Wolters, 1931. Pfaff, Fridrich, Ed., Reinolt von Montelban. Tübingen: Litterarischer Verein 1885, reprint: Amsterdam: Rodopi, 1969. Pfaff, Fridrich, Ed., Das deutsche Volksbuch von den Heymonskindern. Freiburg i. Br.: Herder, 1887. Poncelet, Édouard, Actes des princes-évèques de Liège – Hugues de Pierrepont 1200–1229, Brussels: Palais des Académies, 1946. Reifferscheid, Al., Ed., ‘Histôrie van Sent Reinolt’, Zeitschrift für deutsche Philologie 5, 1874, pp. 271–293. Skårup, Povl, ‘La fin de la traduction norroise de la chanson de Roland’, Revue des Langues Romanes 94, 1990, pp. 27–37. Spijker, Irene, Aymijns kinderen hoog te paard, Hilversum: Verloren 1990 (Dissertation Utrecht). Spijker, Irene, ‘Source écrite ou source orale? Le cas du ‘Renout van Montalbaen’’ in Charlemagne in the North, Proceedings of the Twelfth International Conference of the Societé Rencesvals (Edinburgh 4th–11th August 1991). Eds. Philip E. Bennett et al., Edinburgh, 1993, pp. 95–102. Spijker, Irene, ‘Renout van Montalbaen’ [in English], 19, 1994, pp. 155–166. Spijker, Irene, ‘Les Quatre Fils Aymon aux Pays-Bas’, in Entre épopée et légende: ‘Les Quatre Fils Aymon’ ou ‘Renaut de Montauban’. Ed. Danielle Quéruel, 2 vols., Langres: Gúeniot, 2000, vol. II, pp. 149–159. Thomas, Jacques, Ed., L’épisode ardennais de ‘Renaut de Montauban’, édition synoptique des version rimées, 3 vols. Bruges: De Tempel, 1962. Thomas, Jacques, Ed., Renaut de Montauban, édition critique du manuscrit Douce. Geneva: Droz, 1989. Unger, Carl Richard, Ed., Karlamagnús Saga ok kappa hans. Christiania 1860. Walpole, Ronald N., Philippe Mousket and the ‘Pseudo-Turpin Chronicle’, Berkeley (California), 1947 (University of California Publications in Modern Philology 4). Weifenbach, Beate, ‘Johann Koelhoff der Jüngere: ‘Die vier Heymschen Kinderen’. Zur Bedeutung der Kölner Inkunabel aus dem Jahre 1493 für die Drucktradition von Haimonskinder-Texten in Deutschland’, Amsterdamer Beiträge zur älteren Germanistik 51, 1999, pp. 169–193. Werner, Karl Ferdinand, ‘Die Nachkommen Karls des Großen. 1.–8. Generation‘, in Karl der Grosse, Lebenswerk und Nachleben. Ed. Wolfgang Braunfels, vol. IV, Das Nachleben. Düsseldorf: Schwann, 1967, p. 403–482 and Table at end of volume.