¹ Ward, H. L. D., Catalogue of Romances in the Department of Manuscripts in the British Museum I (London 1883) 630–31.Google Scholar

² Michel, Francisque, ed., La Chanson de Roland ou de Roncevaux (Paris 1837; repr. Geneva 1974) 228–42. The existence in Cotton Titus A. xix of ‘Versus elegiaci de bello in Runcevalle’ had been noted earlier in the Catalogue of the Manuscripts in the Cottonian Library Deposited in the British Museum (London 1802; repr. Hildesheim 1974) 513–14.Google Scholar

³ Grimm, Wilhelm, ed., Ruolandes Liet (Göttingen 1838) xcix.Google Scholar

⁴ von Orelli, Johann Caspar, ed., ‘Carmen de bello in Runcivalle,’ Index lectionum in academia Turicensi (Zürich 1839) 1–13.Google Scholar

⁵ Paris, Gaston, ‘Le Carmen de prodicione Guenonis et la légende de Roncevaux,’ Romania 11 (1882), 465–518.CrossRefGoogle Scholar

⁶ Tavernier, Wilhelm, ‘Beiträge zur Rolandsforschung: II. Carmen de prodicione Guenonis und Rolandsepos,’ Zeitschrift für französische Sprache und Literatur 37 (1911) 83–103. An overview of the earlier debate had been offered by Brückner, Gustav, Das Verhältnis des französischen Rolandsliedes zur Turpinschen Chronik und zum Carmen de prodicione Guenonis (Rostock 1905). There are reviews of prior scholarship in Tavernier and in Curtius (see n. 10 below).Google Scholar

⁷ ‘Les combinaisons logiques des critiques restent impuissantes à reconstituer des poèmes perdus, précisément parce qu'elles sont des combinaisons logiques’: Bédier, Joseph, Les Légendes épiques: Recherches sur la formation des chansons de geste III (Paris 1912) 279–80; see also 394–98.Google Scholar

⁸ Barroux, Robert, ed., ‘Carmen de prodicione Guenonis,’ in Les Textes de la Chanson de Roland (ed. Mortier, Raoul) III: La Chronique de Turpin et les Grandes Chroniques de France, Carmen de prodicione Guenonis, Ronsasvals (Paris 1941) v, xiii, 105–17.Google Scholar

⁹ Mortier, Raoul, ed., Les Textes de la Chanson de Roland I: La Version d'Oxford (Paris 1940)Google Scholar

¹⁰ Curtius, Ernst Robert, ‘Das Carmen de prodicione Guenonis,’ Zeitschrift für romanische Philologie 62 (1942) 492–509.Google Scholar

¹¹ Schumann, Otto, ‘Zum Carmen de proditione Guenonis,’ Zeitschrift für romanische Philologie 62 (1942) 510–27.CrossRefGoogle Scholar

¹² Horrent, Jules, La Chanson de Roland dans les littératures française et espagnole au moyen âge (Paris 1951) 96–99, 373–76.Google Scholar

¹³ Chanteux, Henry, ‘Remarques sur le collège des pairs et la toponymie dans le Carmen de proditione Guenonis,’ Études de civilisation médiévale (ix ^e –xii ^e siècles): Mélanges offerts à Edmond-René Labande (Poitiers 1974) 117–27.Google Scholar

¹⁴ Pézard, André, ‘En Roland naît la peur …,’ Mélanges Labande 595–602. See also the commentary on Pézard's article by Stäblein, Patricia Harris, Olifant 4 (1977) 224–28.Google Scholar

¹⁵ Moisan, André, ‘La Mort de Roland selon les différentes versions de l'épopée,’ Cahiers de Civilisation Médiévale 28 (1985) 101–32 at 113.CrossRefGoogle Scholar

¹⁶ Patricia Stäblein consulted the manuscript in the British Library in 1978 and 1980, and William Paden did so in 1985. We have also used photographic enlargements of folios 153^r–155^r and a microfilm of the manuscript.Google Scholar

¹⁷ ‘En plus d'un passage j'ai dû me borner à confesser que je ne comprenais pas le texte, sans être en état de l'améliorer’: Paris, art. cit. (above, n. 5) 465. Curtius, , art. cit. (above, n. 10) 505, agree with Ruggero M. Ruggieri that ‘la lettura del poemetto latino … è qualcosa di estremmamente faticoso e noioso,’ Il processo di Gano nella ‘Chanson de Roland’ (Florence 1936) 117. Ruggieri further declared that ‘varrebbe la pena ripublicarlo e ristudiarlo a fondo … ma io non posso ora propormi un tal compito.’ Schumann, , art. cit. (above, n. 11) 519, concludes his textual remarks with the claim that all but a few passages are now ‘in Ordnung und verständlich.’ Google Scholar

¹⁸ See also satis in vv. 95f., magis in v. 160, vix in vv. 342 and 446; cf. Curtius 503–504.Google Scholar

¹⁹ Faral, Edmond, Les Arts poétiques du xii ^e et du xiii ^e siècle (Paris 1924) 348, vv. 353–56. On the date of this work, see Faral 39. Murphy, James J. treats the Laborintus as written 'shortly after mid-century': Rhetoric in the Middle Ages (Berkeley 1974) 162.Google Scholar

²⁰ Poetria Nova, vv. 1709–60: Faral, 249–51; trans. Nims, Margaret F., Poetria Nova of Geoffrey of Vinsauf (Toronto 1967) 77–79. On the date, see Nims 11–12, pace Faral 27–33.Google Scholar

²¹ Ch. Thurot, , Extraits de divers manuscrits latins pour servir à l'histoire des doctrines grammaticales au moyen âge (Paris 1869; repr. Frankfurt am Main 1964) 158–59.Google Scholar

²² Faral, Edmond, ‘Le Manuscrit 511 du “Hunterian Museum” de Glasgow,’ Studi Medievali, 2d ser., 9 (1936) 18–121 at 45–47. For the text, see du Méril, Edélestand, Poésies populaires latines antérieures au xii ^e siècle (Paris 1843) 400–405, v. 14.Google Scholar

²³ See Cornog, William H., ‘The Poems of Robert Partes,’ Speculum 12 (1937) 215–50; absque fere pp. 228, 235. For Henry of Settimella, see Marigo, Aristides, ed., Henrici Septimellensis Elegia sive De Miseria (Paravii 1926); the date, p. 23; absque fere, v. 666.Google Scholar

²⁴ Curtius, Ernst Robert, European Literature and the Latin Middle Ages (trans. Trask, Williard R.; New York 1953).Google Scholar

²⁵ Schumann, art. cit. (above, n. 11) 519–20. Text of the Altercatio in Schumann, Otto, ed., Carmina Burana 1.2: Die Liebeslieder (Heidelberg 1941), no. 92; see stanzas 50–52. The poem ‘belongs, perhaps, to the first half of the twelfth century,’ according to Raby, F. J. E., A History of Secular Latin Poetry in the Middle Ages (Oxford 1957) II 291.Google Scholar

²⁶ Faral, Edmond, Recherches sur les sources latines des contes et romans courtois du moyen âge (Paris 1913) 198–201. The passage is in Isidori Hispalensis Episcopi Etymologiarum sive Originum libri XX , ed. Lindsay, W. M. (Oxford 1911) 12.1 (45–46); on the horse in general, see 12.1 (41–56).Google Scholar

²⁷ Gédéon Huet, who first noticed this resemblance between the Altercatio and De Tradicione Guenonis, supposed that the latter was the original: ‘Sur l'origine du poème De Phyllide et Flora,’ Romania 22 (1893) 536–41 at 538–39.Google Scholar

²⁸ Nickel, Helmut, ‘Arms and Armor,’ Dictionary of the Middle Ages 1 (New York 1982) 521–36, at 525, 529. On horse armor, see further Demay, G., Le Costume au moyen âge d'après les sceaux (Paris 1880) 179–85, and Lacking, Guy Francis, A Record of European Armour and Arms Through Seven Centuries III (London 1920) 147–208.Google Scholar

²⁹ Holden, A. J., ed., Le Roman de Rou de Wace (Paris 1970–1973), v. 7490 and note; on the date of this work, III 13–14. Horse armor figures neither in The Carmen de Hastingae Proelio of Guy Bishop of Amiens (edd. Morton, Catherine and Muntz, Hope; Oxford 1972), written no later than 1068, nor in the verses and figures of the Bayeux tapestry (ed. Stenton, Frank; 2d ed., London 1965). Twelfth-century references to horse armor in French include de Troyes, Chrétien, Le Chevalier de la Charrette (ed. Roques, Mario; 1963) v. 3539, and Partonopeu de Blois (ed. Gildea, Joseph; Villanova, Penn. 1967) vv. 2977–80. Epic references are listed in Bangert, Friedrich, Die Tiere im altfranzösischen Epos (Marburg 1885) 64, beginning with Aliscans and Aspremont; de Bar-sur-Aube, Bertrand, Girart de Vienne (ed. van Emden, Wolfgang; 1977) v. 3144. In Occitan, de Born, Bertran, ‘Be.m platz lo gais temps de pascor,’ v. 10: see The Poems of the Troubadour Bertran de Born (edd. Paden, William D., Tilde Sankovitch, Patricia Stäblein, H.; Berkeley 1986), no. 30. In German the Lanzelot of Ulrich von Zatzikhoven (ca. 1195), cited by Schultz, Alwin, Das höfische Leben zur Zeit der Minnesinger (2nd ed.; Leipzig 1889) II 101 n. 4.Google Scholar

³⁰ Horrent, , La Chanson de Roland 97–98.Google Scholar

³¹ In the History of the Franks 2.12, Childeric seeks refuge from his rebellious subjects with King Bisinus and his wife Basina until he can regain his throne. ‘Now that Bisinus and Childeric were both kings, Queen Basina … deserted her husband and joined Childeric. He questioned her closely as to why she had come from far away to be with him, and she is said to have answered: “I know that you are a strong man and I recognize ability when I see it. I have therefore come to live with you. You can be sure that if I knew anyone else, even far across the sea, who was more capable than you, I should have sought him out and gone to live with him instead.”’ Gregory of Tours, The History of the Franks (trans. Thorpe, Lewis; Harmondsworth 1974) 128–29; see the Latin text, ed. Oldoni, Massimo, Gregorio di Tours: La storia dei Franchi (n.p. 1981) 1.138. For earlier discussion of the argument, see Paris, Gaston, Romania 14 (1885) 595, and Brückner, , Das Verhältnis des französischen Rolandsliedes (above, n. 6) 81–85.Google Scholar

³² Ward, , Catalogue of Romances I 290.Google Scholar

³³ See Parkes, M. B., English Cursive Book Hands 1250–1500 (Oxford 1969), esp. xx–xxi and Plates 11 (i) and (ii). The hand is compressed but clear, upright, with occasional Anglicana graphs of a and short s; g has the diamond-shaped lobe rather than horns; the stem of r is broken. The shaft of t protrudes above the headstroke and the ascenders are large loops, as in Plate 11 (ii).Google Scholar

³⁴ Zumthor, Paul, ‘La Chanson de geste: état de la question,’ in Mélanges de langue et de littérature du moyen âge offerts à Teruo Sato (Nagoya 1973) 1.97–112.Google Scholar

³⁵ Gröber, G., ‘Zum Haager Bruchstück,’ Archiv für das Studium der neueren Sprachen und Literaturen 84 (1890) 297 n. 2; Horrent, , La Chanson de Roland (above, n. 12) 98–99. On the difficulties of the method see Löfstedt, Einar, ‘Local Variations in Latin,’ in his Late Latin (Oslo 1959) 39–58; trans. as ‘Regionale Unterschiede im Lateinischen,’ in Mittellateinische Philologie (ed. Önnerfors, Alf; Darmstadt 1975) 385–410.Google Scholar

³⁶ Baist, , ‘Mitteilungen zu Roland–Turpin,’ Verein Deutscher Philologen und Schulmänner, Verhandlungen 43 (1896) 96–97; Foerster, , ‘Das Carmen Rotholandi und sein Verfasser,’ Archiv 135 (1916) 121–38.Google Scholar

³⁷ Tavernier, (above, n. 6) 98; cf. Horrent 98.Google Scholar

³⁸ Horrent, 99.Google Scholar

³⁹ Horrent cites the opinion that the poet was French expressed by Alexander Haggerty Krappe in his Alliteration in the Chanson de Roland and in the Carmen de Prodicione Guenonis (Iowa City 1921) 37. Krappe provides no evidence for his assumption, which seems simply to follow from the French subject of the poem.Google Scholar

⁴⁰ Wesle, Carl, ed., Das Rolandslied des Pfaffen Konrad (3rd ed.; Halle 1966) vv. 9077–83.Google Scholar

⁴¹ Interpretation of the last two lines raises two questions. Bedwngin was glossed ‘bezwingen, zwingen zu’ by Bartsch, Karl, Das Rolandslied (Leipzig: 1874), but simply as ‘translate, render, put’ by other scholars (‘übersetzt’ by Kartschoke, Dieter, ed., Das Rolandslied des Pfaffen Konrads [Munich 1971] 395; English ‘render’ by Keller, Hans-Erich in Olifant 8 [1980] 185–90 at 190; ‘comme le livre était écrit en langue française, je l'ai mis en la latine,’ is the translation by Graff, J., Le Texte de Conrad, in Les Textes de la Chanson de Roland , ed. Mortier, Raoul, X [Paris 1944] 182). If the verb suggests more than an ordinary effort, it might connote Konrad's effort in dealing with the Latin language, or an effort to create artistic effects, or an effort to abbreviate the story, and so on. Danne has been construed by Bartsch and Kartschoke as implying that Konrad translated from his own Latin into German, but by Keller in the sense that he worked from the original French into Latin, and then from French into German.Google Scholar

⁴² Ashcroft, Jeffrey, ‘Miles Dei–Gotes Ritter: Konrad's Rolandslied and the Evolution of the Concept of Christian Chivalry,’ Forum for Modern Language Studies 17 (1981) 146–66 at 156; Keller, Hans-Erich, ‘Der Pfaffe Konrad am Hofe von Braunschweig,’ in Wege der Worte: Festschrift für Wolfgang Fleischhauer (ed. Riechel, D.; Cologne 1978) 143–66.Google Scholar

⁴³ Konrad goes on to say that he has neither added nor omitted anything, which, if it applies to his Latin version (as well as the German), would scarcely apply to De Tradicione Guenonis. Google Scholar

⁴⁴ Martène, Edmundus and Durand, Ursinus, edd., Thesaurus Novus Anecdotorum III (Paris 1717; repr. Farnborough 1968), cols. 307–76. On the date, see Die deutsche Literatur des Mittelalters: Verfasserlexikon (2d ed.; 1981) 3.1187.Google Scholar

⁴⁵ Text in Lachmann, Karl, ed., Wolfram von Eschenbach (6th ed; Berlin 1926) xliii–xliv and more recently in Kleinschmidt, Erich, ‘Die lateinische Fassung von Wolframs “Willehalm”-Prolog und ihr Überlieferungswert,’ Zeitschrift für deutsches Altertum und deutsche Literatur 103 (1974) 95–114. Wolfram composed the original in 1215–18: Verfasserlexikon 4 (1953) 1080.Google Scholar

⁴⁶ Text ed. Schmeller, J. A., ‘Gregorius,’ Zeitschrift für deutsches Alterthum 2 (1842) 486–500. Commentary by Seelisch, Adolf in Zeitschrift für deutsche Philologie 19 (1887) 126–28, 405 n. Hartmann's literary activity has been dated from ca. 1180 to after 1200 in the Verfasserlexikon (2d ed.; 1981) 3.502; on the date of the Latin version, which is contained in a fourteenth-century ms, see Kleinschmidt, 99.Google Scholar

⁴⁷ See Kratz, Dennis M., ed. and trans., ‘Waltharius’ and ‘Ruodlieb’ (New York 1984).Google Scholar

⁴⁸ So Keller, Hans-Erich, ‘Der Pfaffe’ (above, n. 42) 153, 166.Google Scholar

⁴⁹ Other Latin translations of vernacular originals listed in a note by Curtius are rendered in prose, rhythmical verse, or other quantitative measures (European Literature and the Latin Middle Ages [above, n. 24] 26 n.). Latin prose versions include the Historia destructionis Troiae by delle Colonne, Guido (ed. Nathaniel Edward Griffin [Cambridge, Mass. 1936]; trans. Meek, Mary Elizabeth [Bloomington, Indiana 1974]), from the Roman de Troie of Benoit de Sainte-Maure; Gesta Ernesti Ducis (ed. Lehmann, Paul, Abh. Akad … München 32.5 [1927]), from Herzog Ernst; the fourteenth-century Historia septem sapientum Romae, on which see Verfasserlexikon (1953) 3.341–42; a translation of the Contes moralisés of Nicole Bozon (edd. Smith, Lucy Toulmin and Meyer, Paul [Paris 1889] 195–227); and the Latin version of the Fuerre de Gadres discovered by Ross, D. J. A., ‘A New Manuscript of the Latin Fuerre de Gadres and the Text of the Roman d'Alexandre Branch II,’ Journal of the Warburg and Courtauld Institutes 22 (1959) 211–53. There are two Latin versions — one in prose, the other in rhymed rhythmic stanzas — of The Anglo-Norman Voyage of St. Brendan by Benedeit: see discussion in the edition by Waters, E. G. R. (Oxford 1928) cv–cxxv. Another rhythmic version is the Aurea Fabrica of Franco of Meschede (ca. 1330) (ed. Leyser, H., ‘Incipit Avrea Fabrica de laudibus virginis gloriosae,’ Zeitschrift für deutsches Alterthum 2 [1842] 168–76) from Conrad of Würzburg's Goldene Schmiede. A varied quantitative meter was used in an earlier translation of Hartmann's Gregorius (ed. von Buchwald, Gustav, Arnoldi Lubecensis Gregorius peccator [Kiel 1886]).Google Scholar

⁵⁰ Klopsch, Paul, Einführung in die mittellateinische Verslehre (Darmstadt 1972) 63–92; cf. the review by Dieter Schaller in Gnomon 49 (1977) 80–83. There is a recent study of Carolingian texts by Ranieri, Isabella, ‘La tecnica versificatoria nel “Carmen in honorem Hludowici Caesaris” di Ermoldo Nigello e la tradizione dattilica latina,’ Studi Medievali 25 (1984) 93–114. For an analysis of classicizing technique in a text perhaps written in the second half of the twelfth century, see Munari, Franco, ed., M. Valerio: Bucoliche (2d ed.; Florence 1970) lxiv–lxxviii, and Castagna, Luigi, ‘L'esametro di Marco Valerio,’ Studi Medievali 22 (1981) 805–19. For a more general view see Martin, Janet, ‘Classicism and Style in Latin Literature,’ in Renaissance and Renewal in the Twelfth Century (edd. Benson, Robert L. et al.; Cambridge, Mass. 1982) 537–68.Google Scholar

⁵¹ As Janet Martin explains, ‘In the classical Latin hexameter a word-ending was not allowed to fall at the end of the first long syllable of the fifth foot: that is, the line-ending could not consist of a word or word-group of four syllables (though a quadrisyllabic ending was not quite so strictly forbidden if a monosyllable preceded, as in a line ending di genuerunt)’ (562).Google Scholar

⁵² On productio and elision, see Klopsch, (above, n. 50). On the pentameter, see Norberg, Dag, Introduction à l'étude de la versification latine médiévale (Stockholm 1958) 69; Strecker, Karl, ed., MGH PLAC 4 (Berlin 1923) 1163, s.v. producitur; Vollmer, Friedrich, ‘Zur Geschichte des lateinischen Hexameters: Kurze Endsilben in arsi,’ Sb. Akad … München (1917) 3.41 n. 2, 48 n. 4, 50 n. 5.Google Scholar

⁵³ See Klopsch, (above, n. 50) 73–74. On Paul the Camaldulian, see Thurot, Ch., Extraits de divers manuscrits latins pour servir à l'histoire des doctrines grammaticales au moyen âge (Paris 1869; repr. Frankfurt am Main 1964); for his date, see p. 24. Das Doctrinale des Alexander de Villa-Dei (ed. Reichling, Dietrich; Berlin 1893); for his date, see xxxvi–xxxvii. Gervais von Melkley: Ars Poetica (ed. Hans-Jürgen, Gräbener; 1965); for the date, see xxviii. The ‘Parisiana Poetria’ of John of Garland (ed. Lawler, Traugott; New Haven 1974); for the date, see xv.Google Scholar

⁵⁴ See Klopsch, 70–72, and the table on p. 86.Google Scholar

⁵⁵ The count of 34 lines includes 5 in which the caesura is followed by enclitic -que; this is treated as a separate word, following Klopsch's argument, 91–92 n.Google Scholar

⁵⁶ See Klopsch, 68–70.Google Scholar

⁵⁷ That is, 26 times in the third longum, twice in the second (vv. 243 and 481, the former corrected by Schumann), and once in the fourth (v. 227); total: 29 of 242.Google Scholar

⁵⁸ After the first foot 3 times (vv. 120, 268, 378, the last corrected by Paris), and at the mid-point 9 times; total: 12 of 239.Google Scholar

⁵⁹ In prosody, too, the poem is characteristically medieval. Variation of quantity occurs in Gueno, Guenonis, where the first syllable is usually long, but short at vv. 1, 75, 163, 215, 378, and in Gero, which is long at vv. 241 and 267 but short at 320. H is treated as a consonant in Patricius hic 253, as is j in reprimit juga 213 (unless these cases should be regarded as productio in the second and fifth longum respectively). Mute and liquid make position in reprobus 111, Patricius 207, 253, 449, 473, Patris 391, utrique 462. Final -o is short in the ablative of the gerund, and may be short in adverbs (ergo 33, 270, 434, as rarely in Ovid and the post-Augustan poets; primo 363; crebro 369), nouns (sanctio 4; Gueno, Gero), and verbs (adnuo 36). Ubique has a short -e at vv. 223 and 282 by confusion with the enclitic. In semiferus 137 with a long i, and in duodecimus 254 with a long o, the compounds are perhaps treated as separable with productio ob caesuram in the fourth and second longum. Google Scholar

⁶⁰ The table is based on the present edition of DTG (481 lines, of which 242 are hexameters); E = Ernestus , ed. Martène–Durand, , Book I (477 metrical hexameters, disregarding three imperfect ones); W = Willehalm , ed. Lachmann, (superior to Kleinschmidt in several metrical details — 60 hexameters); and G = Gregorius, ed. Schmeller (453 hexameters). See above, nn. 44–46.Google Scholar

⁶¹ Paul the Camaldulian in Thurot (above, n. 53) 444; Gervase of Melkley 208.Google Scholar

⁶² The glossator in Thurot (above, n. 53) 447–48 (ms compiled in 1284, p. 33); Alexander of Villedieu (above, n. 53) v. 2434.Google Scholar

⁶³ Thurot, 448.Google Scholar

⁶⁴ Alexander of Villedieu, vv. 2413–22. Anonymous in Thurot 448–49.Google Scholar

⁶⁵ Gervase of Melkley 208–209.Google Scholar

⁶⁶ John of Garland (above, n. 53) 96. John conservatively permitted both synaloepha and apheresis on occasion (92), and counseled students to avoid long words at the end of the hexameter (96).Google Scholar

⁶⁷ Gervase of Melkley 209. As for prosody, Paul the Camaldulian treated final -o as indifferent except in monosyllabic verbs and second-declension datives or ablatives (Thurot 438). Alexander of Villedieu recognized free variation in quantity in the root of proper nouns, and countenanced the variable treatment of mute and liquid (v. 1558 and Thurot, 426; vv. 1611–14 and Thurot, 420).Google Scholar

⁶⁸ Thurot 460, quoting the thirteenth-century gloss Admirantes to Alexander of Villedieu; cf. Gervase of Melkley 6; Eberhard, vv. 191–92, in Faral, , Les Arts poétiques (above, n. 19) 344.Google Scholar

⁶⁹ Alexander of Villedieu, vv. 2412, 2442–43; cf. Thurot 477, citing the definitions of systole and diastole from Evrard of Béthune (died 1212).Google Scholar

⁷⁰ Ernestus, vv. 38, 167, 230, 237, 386. Gregorius, vv. 4, 24, 79, 228. Paris allowed the first instance in our poem to stand, but corrected the following three.Google Scholar

⁷¹ The following hexameter begins with a word in -que (46–47, 78–79, 100–101, 128–29, 200–201, 208–209, 302–303, 440–41, 448–49), or a relative pronoun (cujus 92–93, -qui 124–25 later in the line), or a subordinating conjunction (nam 140–41, dum 454–55); or it extends parallel structures (quinque … quinque 268–69, the subject 398–99, tu … te 466–67); or it modifies an element of the preceding pentameter (12–13), or is modified by such an element (hic 34–35).Google Scholar

⁷² Eberhard (above, n. 19), vv. 819–24; on final monosyllables, cf. Klopsch's table (above, n. 50) 69–70; on the bisyllable ending the pentameter see also John of Garland (above, n. 53) 96.Google Scholar

⁷³ John, 96. Eberhard recomends the present tense for an eyewitness effect at vv. 435–38. On the use of present-tense forms in past narrative in Old and Late Latin as well as Old French, Spanish, and Occitan, see Paden, , ‘L'Emploi vicaire du présent verbal dans les plus anciens textes narratifs romans,’ in XIV Congresso internazionale di linguistica e filologia romanza: Atti 4 (Naples 1977) 545–57; most recently, on the device in French, see Fleischmann, Suzanne, ‘Evaluation in Narrative: The Present Tense in Medieval “Performed Stories,”’ Yale French Studies 70 (1986) 199–251.Google Scholar

⁷⁴ V. 26; see also vv. 41, 107. Eberhard, vv. 333–36; Geoffrey, , Poetria Nova, vv. 668–86, trans. Nims, 39–40.Google Scholar

⁷⁵ V. 165; see also vv. 465–67, 469–70. The figures listed here from exclamatio through dubitatio are discussed by Eberhard the German, vv. 451–509, and by John of Garland, pp. 114–26. We have provided several definitions from pseudo-Cicero, , Ad C. Herennium de Ratione Dicendi. Google Scholar

⁷⁶ V. 24; see also vv. 166, 285–86, 288, 299–300, 311–12, 357–58.Google Scholar

⁷⁷ V. 288; see also vv. 468, 477–78.Google Scholar

⁷⁸ V. 353–54; see also vv. 371–72.Google Scholar

⁷⁹ V. 276.Google Scholar

⁸⁰ V. 425–26; see also vv. 131, 133.Google Scholar

⁸¹ Vv. 36–37; see also vv. 95–96, 195–96.Google Scholar

⁸² V. 82.Google Scholar

⁸³ Vv. 125–26; see also vv. 8, 28, 55, 117, 155–56, 175–76, 191–92, 213–14, 356, 367–68, 427–28, 453–54. Cf. Eberhard, vv. 699–704. ‘Diese Verse sind seit etwa 1100 ungemein beliebt’: Curtius (above, n. 10) 504.Google Scholar

⁸⁴ Eberhard, , vv. 477–78. John of Garland, pp. 116–20; traductio, ibid., p. 114, called polyptoton by Alexander of Villedieu (above, n. 53) vv. 2486–88. On annominatio in our poem, see Curtius 500–502, 504. On the figure, cf. Ad Herennium (above, n. 19); Faral, , Les Arts poétiques 93–97; Curtius, , European Literature and the Latin Middle Ages (above, n. 24) 273–301; Regalado, Nancy, Poetic Patterns in Rutebeuf (New Haven 1970) 205–54.Google Scholar

⁸⁵ V. 86; see also vv. 63, 97–98, 111–12, 120–21, 399, 418.Google Scholar

⁸⁶ V. 179; see also vv. 75–76, 184, 210–12, 377, 412, 419–20. John of Garland, p. 112.Google Scholar

⁸⁷ V. 240; see also vv. 22, 327. John of Garland, p. 112.Google Scholar

⁸⁸ V. 445; see also vv. 12, 77. Alexander's epanalempsis, vv. 2470–71, is Curtius' epanadiplosis (above, n. 10) 502.Google Scholar

⁸⁹ V. 49–50; see also vv. 115, 161, 280, 313–14. Alexander of Villedieu, vv. 2493–94.Google Scholar

⁹⁰ Vv. 367–68: see also v. 100. Alexander of Villedieu, vv. 2538–40. Curtius did not note the figure of parenthesis in the poem.Google Scholar

⁹¹ Vv. 35–36.Google Scholar

⁹² Vv. 5–6; see also vv. 14,323.Google Scholar

⁹³ Vv. 281–82; see also vv. 313–14.Google Scholar

⁹⁴ It is tempting to see in certain features of the medieval hexameter — the interruption of flowing continuity, caesuras and rhyme, alternation of tenses, even the freedom of form in proper nouns — specific influences of the French epic. For the opposite hypothesis see Ferrante, Joan M., ‘Was Vernacular Poetic Practice a Response to Latin Language Theory?,’ Romance Philology 35 (1982) 586–600.Google Scholar

⁹⁵ Even a late estimate of the composition of the Roland makes it earlier than De Tradicione Guenonis. The date of ms Oxford, Bodleian Digby 23, long considered established in 1130–1150, has been pushed down to ca. 1170 by Horrent, , La Chanson de Roland (above, n. 12) 32–42, and by Short, Ian, ‘The Oxford Manuscript of the Chanson de Roland: A Palaeographic Note,’ Romania 94 (1973) 221–31. The traditional date of the ms has been defended by Samaran, Charles, ‘Sur la date approximative du Roland d'Oxford,’ Romania 94 (1973) 523–27, and more persuasively by Parkes, M. B., ‘The Date of the Oxford Manuscript of La Chanson de Roland (Oxford, Bodleian Library, ms. Digby 23),’ Medioevo Romanzo 10 (1985) 161–75. The composition of the text has been advanced by Hans-Erich Keller from the usual date ca. 1100 to 1148–1150 in ‘La Version dionysienne de la Chanson de Roland,’ in Philologica Romanica: Erhard Lommatzsch gewidmet (edd. Bambeck, M. and Christman, H. H.; Munich 1975) 257–87. We refer to the text of the Roland as edited by Brault, Gerard J., The Song of Roland: An Analytical Edition (University Park, Penn. 1978).Google Scholar

⁹⁶ See Stäblein, Patricia Harris, ‘The Structure of the Hero in The Song of Roland: Heroic Being and Becoming,’ Olifant 5.2 (1977) 105–19.Google Scholar

⁹⁷ Laisse 113 in the ms is renumbered as laisse 125 by Brault, with commentary in II 201–202. For our present purposes, we prefer to follow the order of laisses in the ms.Google Scholar

⁹⁸ We wish to express our gratitude to Paul Sainger, Hans-Erich Keller, the late Eliott, Alison G., and an anonymous reader for Traditio for their help with various aspects of this study. Rubric of Prologue: prologus] descender of p destroyed by hole in L; bello] b obscured by hole in L. 2 Indented as pentameter MPB. 5 presignis] first s obscured, first i destroyed by hole in L. 11 succescit M. 17 exicidium B. 21 eum] enim MO; pervenit MO. 23 nihil] nobis MO. 29 lege MO. 31 Vel] Ut MO. 34 hoc MO; quia] quoque MO. 50 ‘On pourrait corriger ira fuit; mais cf. v. 86’ P. 51 ferre breve O. 53 Talia] T partly obscured by hole in L; repeated above line in L. 56 hunc urit] huc urit B. 68 discendens M. 70 aget agit L, aget crossed out. 71 simul MOP, hole in L, omitted B; horror] h and ro obscured by holes in L. 74 sunt] unt obscured by hole in L. 75 Eminus P. 76 paret B. 77 paret LMO, parum PBS. 83 audacem] d expunctuated after ce in L. 85 abire] ab ira (?) S. 87 intratque P. 88 reto L, retro MOPB. 89 Demum LMOB, Deinde P. 92 Bravimunda MO. 94 Magficat L, Magnificat MOPB. 96 tamen] tum MO. 100 mage MO. 101 Et que B, Et quia MO, Bisque P. 104 metro LMOB, metus P. 113 Do LB, Da MOP. 114 morte MOPB. 116 terat PB. 120 Unde] Ut dici (?) S. 121 First Hinc] Hinc or Huic L, Hinc MSOS, Huic PB; second hinc LMOS, huic PB. 123 subito subito B. 126 urbanos ho MOPB, patch over hole in L. 127 parere pares] patch over hole in L, parere pares OP, parare pares M, pares B; tibi MOP, no indication of lacuna in L, omitted B. 129 regna] re nearly destroyed by hole in L; abreviare LMPB, abbreviare O. 131 lectum PB. 133 Inspiat B. 134 miratus MO. 137 semi-] semet S. 144 Quod] Que P. 145 -que] quia MO. 151 Rollandus L, Rollandus (sic) MB, Rollandum OP. 153 spondit LMB, spondet OP. 156 pulchra MO, plucra B; citi PB, titi L, sibi MO. 157 et] at O, at or qui S. 159 Nunc MOP; magis auget] mage auget MO. 160 Munere O; mota magis] mota mage MO. 161 Sive S, seu LMOPB; timor] livor P; donum] dolor B. 164 sua] s obscured by hole in L. 165 cupido] i destroyed by hole in L, cupido MOPBS. 166 qui] que P, que or quia S. 171 suadet] a destroyed by hole in L, suadet MOPB. 173 Carolo M. 175–76 Inverted with 177–78 by P. 176 adit P. 181 Hec B. 190 quecunque MOPB. 195 videre MO. 199 rex omitted B. 200 ille] ire PBS. 203 indicat MO. 204 Indicii MO; sinet MO. 209 ducum PS, ducit LMOB. 210 equos LB, eques MOPS. 211 viarum] maris (viaris?) L, maris MOB, viarum P. 213 reprimunt MO. 221 graditur] t destroyed by hole in L, graditur MOPB. 224 Haut] Hanc M, Hac O. 225 inspirare] re obscured by hole in L. 226 subveniendo] ub half destoyed by hole in L. 227 Hoc] Hic P; Numquid quod] Numquis quod MO. 229 terreretur LM, terreret OPB. 236 First Nunc] Non in text but Nunc in note P, Nunc or Non S. 237 octima B. 238 Vincere] vinci L, vincere MOPB. 240 First mori] mori in text but virum in note P, viro S; fogiendo B. 241 Huic] Hinc MO. 242 nemoro O. 243 Dans animos animus S; hos tres] nostros in text but ‘hos tres pourrait se défendre’ P. 245 clipeis] clipeus B; latun B. 246 inermis MOP. 249 regentes] tenentes MO. 252 No lacuna in L. 253 Patricios PS; hic] hiis M, his O. 255 -que omitted O. 256 Sed regum] Se regnum B. 261 parare or perare L, perire P, parare M, parere O, perare B. 262 condicioni O. 263 fert] ferit P; turma PB. 264 prelia] urelia B. 267 Sampson] Samson MOPBS. 269 cetera] corpora PBS. 273 mox P, omitted LMOB, et S; et MOPS, omitted LB. 275 furans] funeris P. 277 quia] -que MO. 285 robore B. 288 haec O, hec P. 288–89 B inserts repetition of v. 292. 290 -que crossed out between sua and proque in L. 291 dimediat MO. 292 Permittitque MO, Prenittique B; retinetque decem B, hole in L, lacuna MO, retinetque decem P; marginal gloss scilicet reges L. 295 tota] o destroyed by hole in L, tota MOPB. 301 taligus B. 303 viris] viri P. 305 et ferit] at feri B. 310 eques] equus MO; precicitatur B. 311 morte B. 321 prius] post MO; corcora B. 322 humus] (h)onus S. 328 si] hole in L, lacuna M, si OPBS. 329 fremit] finit M, furit O. 330 fit] sit M; mente] -e obscured in L. 331 leniter MO. 333 invisis] immissis M, invisus OP. 335 premunivit MO. 336 tutus equus B; mage tutus] mage tu tus B; eques] equus LM, eques PB. 337 speciem] spernet MO; laudat] illegible superscript in L; laudet (?) S. 338 lauda B. 341 cavus] canus MO, capus B. 343 laxa LB, laxat MOP. 344 celeres MOP, sceleres with initial s crossed out in L, sceleres B. 345 it] fit O. 346 dat] lower half of word covered by patch on hole in L; pene] fine P. 347 Proximus huic Rollandus] patch over hole in L, Proximum … Rollandus MO, Proximus huic Rollandus PB. 349 non curat] -n c- obscured by hole in L; parcere] perdere M. 351 similis] sic corrected to sis in L. 352 inimice OPS or deus ipse S, inimisse L; probat LMPB, probas OS. 353 Hec] Hic MOP. 356 prava] parva O. 357 Callia B. 358 defleres with s expunctuated in L. 361 Rollandus B. 366 Tene pudet] ne te pudet LMOB, tene pudet? S, te pudeat P. 367 Num] Non O; intolerabile OP, intollerabile LMB. 369 crebo B. 370 monet] coget B; cogit] rogit M, rogitat O. 371 hec] hunc P. 374 utroque LMOB, utraque P. 375 inter PS, iter LMOT. 378 Vinctus] Vinclis P. 382 submonet] t partly destroyed by hole in L. 384 mori] m obscured by hole in L. 385 vires omitted B. 386 Exercetque OP; in stimulando LMB. 387 peractus M. 389 visus] hole in L, vis. MO, visus PB; instatque B. 390 Prebet] Intret B; intret] terret P. 392 ante] autem MO. 397 Oliverus] Hivernus B; augit M. 398 Agalisus MO. 402 in moriendo LMOPB, immoriendo (?) S. 403 jam] jaff B. 405 Rollandum (the same abbreviation represents Rollandum at vv. 45 and 205) LOP, Rollandus (sic) M, Rollandus B; illi P, omitted LMOB. 411 adictus O. 412 Id] initial H crossed out in L. 415 submissus OPB. 416 Proh] t before h in both occurrences in L, both times expunctuated; perit] parit LMB, perit OP. 418 est] plus est B; -que] quam MO. 421 huic] habet B; una] inde MOP. 423 -que pedes] pedites P, -que pede S. 428 viros] piros B. 431 certamina S. 435 Catchword in bottom margin of L: Omnibus hoc solum (= v. 436). 436 hoc: oc obscured by hole in L, hoc MOPBS. 438 itaque B; utrique OP. 439 ex hinc MO. 440 vi] i nearly destroyed by hole in L, vi MOPB. 441 moritur] o nearly destroyed by hole in L, moritur MOPB. 443 exanguis] ex sanguis MOP. 444 munda B. 445 dum] cum O. 446 hic sine vi MO. 449 Circumquecque B. 451 hunc MO, hoc B; ab inde MO. 455 vincere MO. 459 adeunt] credunt P. 462 secum O. 464 latet expunctuated between nemo and jacet in L. 471 cum] quid P. 474 obiter P; vidit MO. 477 quos twice MO; summo] summos B; mediocres] mediocre B; mediocri] mediocre M. 478 quos MO. 482 finata B. Title Supplied from explicit; see n. 2 The prologue comprises two Leonine hexameters (aa/bb), balanced by the rhymed elegiac couplet in vv. 481–82 (caudati Leonini, aa/aa). 36–38 There is similar word-play on the Scholastic theme of nonbeing in non-gens 287 and non-hostem 409. Cf. the Geta of Vitalis of Blois, v. 403, and the commentary by Hunt, Tony, ‘Aristotle, Dialectic, and Courtly Literature,’ Viator 10 (1979), 95–129 at 123–25. 41 ‘Amoris amore signifie ici simplement “par amitié,” amor étant pris dans le sens de causa’ (P). For a use of the Augustinian phrase, amore amoris, in Guillaume de St-Thierry, see Blaise, A., Lexicon Latinitatis medii aevi (Turnhout 1975) 42. 44 For a similar ironic use of praemia, cf. Aeneid 2.537.Google Scholar

64 Parare is used here for parere, ‘to obey,’ as a result of confusion which elsewhere produced parere for parere, ‘to adorn’ (Du Cange 6.171; Blaise, Lexicon 655). The confusion stems from the normal evolution of parare into French parer. Cf. parit 416 (MS), emended to perit. 88 retro] Cf. retro 192. 174 palliat] Palliare, ‘to cover, hide, conceal’; see Niermeyer, F., Mediae Latinitatis lexicon minus (Leiden 1976) 755. 199 Allusion to the etymology of securus, se + cura, ‘without care’ (P). 204 Sinit in the sense of desinit, ‘cease,’ was unknown to P and to S (p. 513); but see Lewis, C. T. and Short, C., A Latin Dictionary (Oxford 1879) 1708: ‘rarely … to give up, cease’; Blaise, A., Lexicon 849: ‘laisser aller, cesser.’ 252 The missing passage may have been a single pentameter or some greater odd number of lines; possibly it ran through a list of pagan peers from Marsile's nephew down to the eleventh peer in v. 253, like the sequence of laisses on the twelve pagan peers in the Oxford manuscript (ed. Brault, laisses 69–78). In the Oxford version Marsile's nephew Aelroth asks to strike the first blow against Roland (cf. v. 251), does so, and is slain. The parallel suggests that Hic 257 may refer to Marsilii nepos 251, and not to patricius duodecimus 253–54.Google Scholar

275 furans horror] Corrected to funeris horror by P, but the versus spondiacus appears occasionally in medieval verse as well as in classical texts; see Munari, F., M. Valerio: Bucoliche (above, n. 50) lxxvi, and Louis Nougaret, Traité de métrique latine classique (3rd ed.; 1963) 45–46. 281 ‘Festinare dans ce sens [i.e. transitif] n'est pas latin’ (P); but cf. the transitive sense, ‘to move quickly,’ in Glare, P. W., Oxford Latin Dictionary (Oxford 1982) 694. 291 P mistook exercitus for a genitive with false quantity (short), but S points out that it is nominative. 311–12 Num for nonne (S). 375 inter agendum PS, iter agendum L. Translated ‘während seines Tuns’ by S .Google Scholar

399 relisus] ‘mindestens ungewöhnlich für laesus’ (S 513); relido [re + laedo]: ‘to dash (in quot., a wave) in the direction from which it comes,’ Oxford Latin Dictionary 1605. 404 debilitat] The intransitive use is idiosyncratic (P; S 513). Explicit: tradicione] The post-classical usage was replaced in the title by Paris, G. with prodicione, following a rubric in the Chronique de Turpin. Prodicio does not occur in the poem, although fraus, sedicio, scelus, insidiae, livor, cupido, and nephas are applied to Ganelon's deed. Pertinent meanings of Medieval Latin tradicio include ‘dolus, fraus, perfidia’ (Du Cange 8.147); ‘trahison, perfidie’ (Blaise, A., Lexicon 921); ‘acte de haute trahison; dessein malicieux, attentat, machination; meurtre prémédité; attaque au dépourvu, coup de main’ (F. Niermeyer, Lexicon 1036f.). For Paris' source see the rubric ‘De proditione Ganaloni et bello Runcievallis et passione pugnatorum Karoli,’ in Les Textes de la Chanson de Roland (above, n.8) III 58.Google Scholar