From the Roman de Troie to the Historia destructionis Troiae: notes and a proposal on the source of Guido delle Colonne
p. 27-42
Texte intégral
1Since the very first critical approaches to the Historia destructionis Troiae it has been recognized that its model has to be sought in the French Roman de Troie. Even considering Guido delle Colonne’s “misdirection” – as he wishes to present himself as inspired by more trustworthy models1 – the version of the Trojan legend handed down by the Historia is without a doubt “telle qu’elle a été imaginée par Benoît de Sainte-Maure”.2 However, providing a precise identification of the actual version of the poem that Guido could have used for his operation of mise en prose has proved to be much harder. The Roman de Troie has a rather complex textual tradition, not to mention the number of derivative works circulating at the same time as the poem3. Guido, for its part, skillfully selects, reworks and inlays his text with contributions from different sources. The aim of this essay is to shed new light on the complex and sometimes problematic relationship between the Historia and the Roman de Troie and to identify the features of the version of the French poem that underlies the Latin text.
2The first question that needs to be addressed is whether Guido had a direct access to the poem or to a transposition of the Roman de Troie, possibly already in prose form. As a matter of fact, this question has appeared quite early in the critical discussion of the Historia. Egidio Gorra posed it in very precise terms in his investigation of the texts of the Trojan legend in Italy.4 However, he suspended his judgement on the issue due to the lack of sufficiently probative elements. The second modern editor of the Roman de Troie, Léopold Constans, dedicated a long chapter of his work to the examination of known texts derived from it – mainly the Roman de Troie en prose and the Historia destructionis Troiae.5 His own critical edition allowed Léopold Constans to meticulously examine the relationship between the Latin prose and the Roman de Troie, and therefore to highlight a few discrepancies between the two texts.6 In this regard, the central question seems to be whether such discrepancies are due to the fact that Guido used, instead of the poem, “une rédaction en prose, français ou italienne, du poème” as his source.7 As far as the involvement of an Italian intermediate source is concerned, Léopold Constans noted a fundamental lack of textual and critical material for a significant study of the matter. On the other hand, with regards to a French prose version, he proposed a comparison with the only Roman de Troie en prose known to him, the one today known as Prose 1.8 Because of irreconcilable macro-structural differences, he could not reach a positive conclusion.9
3In 1942, Kathleen Chesney insisted on the existence of a French intertext.10 Having more information at her disposal, she thought she recognised the text used by Guido in a newly identified mise en prose distinct from the one previously known, the “southern” (that is, Italian) prose version of the Roman de Troie en prose or Prose 2, to use the current updated nomenclature. This text is transmitted in the manuscripts Grenoble, Bibliothèque municipale, 861 (263 Rés.), Paris, Bibliothèque nationale de France, NAF 9603 and Oxford, Bodleian Library, Douce 196 – the latter was reported for the first time in Kathleen Chesney’s essay. Her identification was based on the evaluations of some structural loci in which both Prose 2 and the Historia show similar differences compared to the critical text of the Roman de Troie.11 Chesney’s hypothesis has sporadically resurfaced in the subsequent critical literature,12 but it does not seem to have generally gained much popularity.13 While her arguments are undoubtedly pertinent, her attempt to trace a relationship of filiation between the two mises en prose, the Latin and the French one, proved inconclusive. An in-depth study of Prose 2 seems in fact to reveal more differences than affinities between the two prose versions, especially regarding some characterizing elements of Prose 2 which are not found in the Historia.14 Ultimately, one has to agree with Marc-René Jung’s definitive dismissal of Kathleen Chesney’s suggestion.15
4If Prose 1 and Prose 2 are then to be rejected from the list of possible intertexts of Guido’s, the other known mises en prose should also be excluded: Prose 5 for chronological reasons,16 Prose 3 and Prose 417 for macrostructural discrepancies and, in the case of the latter, for its too limited and exclusively extra-Italian circulation.18 This being the case, the simplest and most logical conclusion that can be drawn from the survey is that Guido had to have directly known the poem by Benoît de Sainte-Maure.19 Therefore, the model of the Historia should be found within the tradition of the Roman de Troie.
5Before one can successfully propose such an identification, however, it is necessary to take into consideration two preliminary issues. The first is that the text of the Historia, as Nathaniel E. Griffin provided it, is not really reliable, and neither are the critical premises of his edition.20 The second is the inherent complexity in the classification of the witnesses of the Roman de Troie. Let me examine these problems in order.
6As Ezio Franceschini famously noted in his review, Nathaniel Griffin’s edition of the Historia presents a number of significant methodological issues.21 Griffin stated that he was aware of some 136 manuscripts and that he directly consulted 94 of them. However, he did not provide the shelf-marks for all these codices, referring instead to a future essay that has never seen the light of day.22 He declared that it was impossible to classify such a vast number of witnesses in a meaningful way (“Obviously from so many manuscripts no authoritative stemma could be constructed”23), and decided to edit the text based only on the eight oldest dated manuscripts at his disposal.24 Aside from depriving scholars of a fundamental preliminary work in organizing such a large textual tradition, this approach is especially problematic when trying to establish comparisons between texts, since it implies a general unreliability not only of the edition itself but also of the apparatus, which registers a set of quantitatively and qualitatively defective variants.
7In short, the only modern edition of Guido’s work makes it difficult to determine to what extent the discrepancies between the Guido’s and Benoît’s texts are dependent on the internal variability in the tradition of the Historia destructionis Troiae. Similarly, the classification of the Roman de Troie’s witnesses also presents quite a few problems in identifying a more or less circumscribed model for the Latin mise en prose, albeit for completely different reasons. More specifically, as far as the sorting of the witnesses of the French poem is concerned, the scheme developed by Léopold Constans25 is still in use. It distinguishes two families, α and β, which in turn are each divided into two groups:26
| Family α | Family β |
Group v M2 A R S A2 V1 V2 Fragments: B1 L3 N3 R1 |
Group y D M1 E H J P Fragments: B2 B4 C2 O S3 |
Group x G L F N L1 Fragments: B3 C4 L4 N1 N2 P1 P2 V |
Group z A1 L2 C W C1 B I S1 K M Fragments: M3 S2 Vi |
Not precisely classified: Fragments: T P3 |
Not precisely classified: F1 N4 Fragments: B5 C3 G1 G2 M4 P4 P5 |
8However, this grouping is characterized by a declared permeability between branches, mainly due to the high rate of contamination, active at both the level of the single codex and of the groups, as y and z both appear to be varyingly indebted to x.27 Of the thirty-nine manuscripts put on the stemma by Léopold Constans, twenty-eight are openly recognized as contaminated.28 At the beginning of his monumental recensio, Marc-René Jung – who uses terms which clearly show the fickleness of this classification to describe the groups, such as “assez instable” (v), “très solidaire” (x), “parfois solidaire pour D E H M1” (y)29 – clearly warns that “lorsque nous dirons, dans la présentation des manuscrits, que le texte appartient à la première ou à la deuxième famille, il ne s’agira que d’indications générales”.30 This circumstance does not prevent the search for Guido’s source, but some additional methodological caution is necessary. One needs to be wary, in particular, of the mere concordance of single isolated loci between one or another witness of the Roman de Troie and the Historia, which can be non-significant at best and misleading at worst. Thus, it is best to proceed with a systemic comparison on several significant textual articulations that may, on the whole, suggest a tendential proximity with a family or a more or less defined group of witnesses.
9A basis for this comparison is provided, once again, by Marc-René Jung’s work on the tradition of the Roman de Troie, in particular by his study of significant portions of the text where the distinction between the second and first family and, within each family, between groups of single codices, is mainly produced.31 A meaningful comparison is not always easy, due to the radically different forms of expression of the two texts – verse and prose –, and to Guido’s well-known tendency to play with the narrative material at his disposal, an element that often compromises an effective evaluation. This implies that the lack of an episode (or part of it) between the Latin text and a certain group of witnesses of the Roman has generally no real analytical value, as it seems difficult to demonstrate with sufficient certainty whether a gap is caused by some genetic issue or by the author’s willingness to summarise, especially if the omission involves single lines of the poem or easily isolated narrative plots. Furthermore, due to Guido’s storytelling some narrative passages which normally constitute a privileged tool for the classification of both the witnesses of the poem and its derivative works are effectively unrecognizable.32
10An example that can clarify this statement is the encounter of Achilles and Hector during a moment of truce. Marc-René Jung divides the passage into four different thematic portions, each of which can be found in the manuscripts in a long or in a short version.33 The ten different combinations that are concretely realized in the witnesses provide a very effective tool for classification. However, in the Historia, the whole episode is transfigured in such a way that an adequate placement in this scheme is practically impossible.34 Additionally, two descriptions such as those of the Chamber of Beauty and the tomb of Achilles present in the witness of the Roman distinctive forms of amplification and reduction,35 but neither have any place in Guido’s narration, as he considerably shortens the first, which he considered too fanciful to correspond to the historical truth,36 and abruptly omits the second.37
11Among the loci that could successfully be compared, one is quite well known. Based on the text of Dares,38 Benoit lists the names of the eight children that Priam fathered with Hecuba: five sons (Hector, Paris, Deiphobus, Helenus and Troilus) and three daughters (Andromache, Cassandra and Polyxena).39 The presence of Andromache, Hector’s wife, as the king’s eldest daughter and therefore sister of the Trojan hero, clearly appears to be incorrect. That is why Léopold Constans tries to resolve the question by splitting the character of Andromache into two distinct figures with the same name, one, mentioned exclusively in this context, who is the daughter of Priam, and the other who is his daughter-in-law.40 If today we can be persuaded into believing that the misunderstanding stemmed from an overly literal interpretation of the term liberis (which could have been found in Dares’ text by Benoît de Sainte-Maure41), the inclusion of the lady’s name must have sounded erroneous already to many of the poem’s copyists. Six codices (in fact seven, if one adds the acephalous M1), all belonging to the second family, correct the number of children to seven, and completely remove the reference to Andromache.42
12The behaviour of two isolated codices of the second family, however, is even more interesting for the search of Guido’s source. Both codices, A1 and L2, belong to group z.43 In them, the copyists bring the number of Priam’s children back to the original count of eight, but they replaced Andromache’s name with that of Creusa, Aeneas’ wife, whom the myth describes, indeed, as Priam’s daughter. The reading, Creüsa, qui ert mariee, seems to be very closely recalled by the Historia, which translates Filiarum autem prima vocabatur Creusa. Hec asseritur Henee fuisse coniugem […]44 and adds the usual amplification. The mention of Creusa of all the possible daughters Priam had according to the myth, accompanied by the indication of her marriage, seems a bit too specific to be suspected of polygenesis: Guido must have found in his source a reading similar to that of A1 and L2. Both those manuscripts were copied much earlier than the composition of the Historia, A1 in 1237,45 according to its colophon, and L2 around 1200–121046. The original integration of Creusa is therefore present in actual witnesses of the Roman de Troie at a high chronological point and, as mentioned, entirely within group z.
13Further in the plot of the poem, after their arrival in Asia and on their way to Troy, the Greeks devastate a few enemy castles. Two of these castles are named, at line 6066, as “Tenedon” and “Laurientel”.47 This pairing is not common in the tradition. In fact, it is characteristic of the first family – except for A (Paris, BnF, fr. 60), which reports the singular reading Tenedon sous la tentel. In the second family D, M1, J and P, which constitute the majority of group y, as they are usually positioned in it, as well as W go with the first family, while H, the not precisely classified codices F1, G1 and N4, and above all group z in its almost entirety reproduce, with variations, Qui si estoient fort et bel (reading of K).48 Here the Historia widely reworks the episode, which remains, however, sufficiently recognisable in book XI.49 In that book, the two castles are in fact named as “Thenedon” and “Sarronabo”. This is not the place to further explore the origin of the name Sarronabo, but I should mention at least that such a name does not seem in any way to descend from a corruption of “Laurientel”. The presence of Tenedos is not problematic at all, as the name is also mentioned in the episode of the taking of the castle just about sixty lines before the one in question, as well as in a considerable number of other occurrences during the course of the narration. However, it is the absence of Laurientel, a name that appears exclusively in this precise passage, which leads to believe that it most likely did not appear in the codex of the poem used by Guido.
14An interesting case of alternating readings in the Roman de Troie’s tradition occurs at the beginning of the duel that leads to the death of Patroclus, when the second battle starts. In the first version, Patroclus is the one who strikes the first blow at Hector,50 while in the second Hector is the first to strike.51 In accordance to the canonical scheme of the battles in the Roman de Troie, where the first one who attacks is ultimately defeated at the end of the battle, the genuine reading is certainly the first one,52 duly followed by Guido in book XV of the Historia.53 The oscillation between these two versions in the tradition does not concern either family in a distinctive way, so much so that Marc-René Jung presents it as an exemplary case of the fact “qu’un simple renvoi à une des deux familles ne suffit pas pour savoir ce que le lecteur lit dans le manuscrit qu’il a devant lui”.54 Nevertheless, observing how the two readings are distributed at a lower level of the stemma, within the individual groups, it is apparent that the reading shared by Guido is found in a preferential way in all the codices of group x, which confirms itself once again as “très solidaire”, and in a large number of witnesses of z (A1, L2, I and S1), joined by N4, one of the “unclassified” manuscripts of the second family.55 By contrast, this version appears to be decidedly minoritarian in v, where it can be found only in R and A2, and in group y, represented by E alone.56
15The last element of comparison presented here is provided by the description of the sumptuous tomb of Hector, a two-storied tabernacle built near the high altar of the temple of Apollo, raised by four columns supported by caryatids.57 All of the witnesses agree on this part of the description, as well as the fact that Hector’s embalmed body is located on the second story of the funeral monument. However, the tradition later splits over two main elements of the description, namely the visibility of the body of Hector from the outside and whether a golden statue with his features is placed at the top.58 With regard to the first point, the reading found in most of the manuscripts shows the embalmed body of Hector seated on a throne under the vault of the tabernacle, visible from the outside.59 Opposed to that, the codices L, L1, N, G and F replace “voute” with “tombe”, and the body, placed on the inside of the tomb, disappears from the eyes of the public.60 In any event, for the purposes of comparison with the Historia, the most interesting version is that of R (first family, group v) and I (second family, group z):61 here, Hector is placed, hidden from sight, Ens la chambre d’ebenus. The origin of this singular reading is not exactly clear, but in book XXII of the Historia the body of Hector is enclosed by an ebony wooden wall:
Et sic, completo opere tabernaculi ipsius, circa ipsius ambitum statuerunt quandam clausuram ex lignis ebani, que clausa poterat esse quandoque et quandoque poterat aperiri ut corpus Hectoris sic locatum aspicere uolentibus apertius pateret aspectus.62
16The tradition divides itself once again a few lines later: the Constans’ edition, which follows the common reading of group x, reports the presence of a golden statue of Hector represented in the act of wielding a sword as a sign of threat to the Greeks,63 and this version is also found, outside of x, in A2 and P3; in the second family and in the majority of the witnesses of group v, it is the embalmed body of Hector, not the statue, that holds the sword.64 I and R are again the exception: they take after the group x version but present a passage developed on sixteen lines in which the statue is hoisted on a wonderful spire on top of the monument.65 Guido’s text easily demonstrates his dependence on a codex which bears the first version:
In excelso uero testudinis ipsius tabernaculi supra culmen predicti magistri constituerunt quandam auream statuam, similitudinem Hectoris describentem, habentem ensem nudum in manu, cuius aspectus et facies erat ex ea parte in qua Grecorum exercitus in eorum tentoriis morabatur, qui cum eius ense Grecis minas inprimere uidebatur.66
17The stress on the top position of the statue could again bring the Historia closer to a witness next to I and R, and therefore to group z.
18Taking into account the cautions put forward in the preliminary phase, some relevant, albeit provisional, conclusions on the features of the Roman de Troie at Guido’s disposal can be drawn from the analysis of the loci explored earlier. First of all, some key textual choices typical of the second family can be found with a certain frequency, as highlighted sporadically in the critical literature.67 However, the Historia does not appear to be equally close to all the subclassifications of this family. In fact, the codices generally indicated as being part of group y have rarely emerged in the analysis of the loci: setting aside the presence of Creusa in place of Andromache, a reading peculiar just to A1 and L2 of group z, only H in the case of Laurientel and E in the case of the death of Patroclus show readings approachable to Guido’s text. On the contrary, witnesses of group z frequently appear, and in all the cases considered they present a text similar to that of the Historia. It is certainly not a matter of an exclusive and univocal relationship: regarding Hector’s tomb, for example, we have seen how Guido’s description overlaps the version transmitted in primis by the “très solidaire” group x. In this regard, the use of several models, especially for a work produced in a context such as the Historia’s, cannot be excluded a priori. However, it seems in this case to be a less convenient and ultimately less likely hypothesis compared to postulating a single model, since a textual form superimposable to the one traced so far is realized in different ways in codices of group z. In addition, as mentioned above, group z is subjected to contamination from group x at a high level of the stemma. The proposal made at the end of this survey is therefore that Guido’s model should be identified as a lost, contaminated codex of the second family, which tends to belong to group z. It should have had some significant shared elements in particular with A1, I and L2, but cannot be surely identified with any of them or with other known manuscripts.
19To conclude, a thorough investigation of Guido’s model is destined to shed light on our knowledge of the circulation of the Roman de Troie in Italy. Considering only the thirty complete manuscripts surveyed by Marc-René Jung, ten of them have been more or less accurately traced back to the Italian area.68 Six were probably copied in the North,69 and the Veneto region appears to be one of the irradiating centres of the Roman’s reception.70 One manuscript is from Florence, while the localization of two others oscillates, according to the studies, between central Italy and, again, the North of the peninsula.71 Marc-René Jung places the production of only one manuscript in southern Italy: Venice, Biblioteca Marciana, fr. XVII (= V1), a fourteenth-century illuminated codex of the first family, presumably produced in Naples in 1330–1340 but soon included in the Gonzaga’s collection.72 The Neapolitan localization, however, is not universally accepted: the southern Italian origin of the manuscript is acknowledged by Keith Busby and Ilaria Molteni,73 but rejected by Günther Holtus and Peter Wunderli (who favour a northern origin), by Gabriele Giannini (who believes the manuscript to have been copied in western Veneto or in eastern Lombardy) and by Sebastiano Bisson (who accepts the Venetian hypothesis).74 At present, therefore, the circulation of codices of the Roman de Troie in southern Italy is directly proven by either one or, possibly, no manuscript, in clear dissonance with the material reality of a geographical area which, in addition to the Historia destructionis Troiae, has seen the production and circulation of a considerable amount of literary material around the Trojan legend. “La postérité d’un texte à travers ses adaptations et ses traductions offre, on le sait, un bon terrain d’observation de sa tradition manuscrite.”75 This is partly what I have tried to do in this essay: to restore a piece of the history of the circulation of the Roman de Troie in Italy more deeply than what the codicological evidences alone allow.
Notes de bas de page
1See Francesco Bruni, “Tra Darete-Ditti e Virgilio: ‘fabula’ e ‘storia’, ordo artificialis e ordo naturalis”, Studi medievali, III/37, 1996, p. 753–810, reproduced in Rosa Casapullo et al. (dir.), Tra popolo e patrizi. L’italiano nel presente e nella storia, Firenze, Cesati, 2017, p. 65–113, from which I quote, here at p. 89–100.
2Marc-René Jung, La légende de Troie en France au Moyen Âge. Analyse des versions françaises et bibliographie raisonnée des manuscrits, Basel/Tubingen, Francke, 1996, p. 565.
3See Jung, La légende de Troie…, op. cit. For the five mises en prose see also Luca Barbieri, “Roman de Troie, Prose 1”, “Prose 2”, “Prose 3”, “Prose 4”, “Prose 5”, in Maria Colombo Timelli et al. (dir.), Nouveau Répertoire de mises en prose (xive-xvie siècle), Paris, Classiques Garnier, 2014, p. 773–848.
4Egidio Gorra, Testi inediti di storia trojana preceduti da uno studio sulla leggenda trojana in Italia, Torino, Loescher, 1887, p. 118–119.
5Benoît de Sainte-Maure, Le Roman de Troie, publié d’après tous les manuscrits connus, ed. by Léopold Constans, Paris, Firmin Didot, 1904–1912, vol. VI, p. 264–345.
6Ibid., p. 328–331. It should be noted that, in the absence of an edition, Léopold Constans reads the Historia destructionis Troiae from the incunabulum, Aix-en-Provence, Biblioteca Méjanes, n. 118.
7Ibid., p. 331.
8The nomenclature in use for the five different mises en prose was set by Marc-René Jung, see Jung, La légende de Troie…, op. cit., p. 440–562. Léopold Constans himself had begun to work on an edition of Prose 1, which remained incomplete due to his death in 1916 and was only partly continued by Edmond Faral (Le Roman de Troie en prose, ed. by Léopold Constans, Edmond Faral, Paris, Honoré Champion, 1922).
9Benoît de Sainte-Maure, Le Roman de Troie…, op. cit., vol. VI, p. 331–334.
10Kathleen Chesney, “A neglected prose version of the Roman de Troie”, Medium Aevum, 11, 1942, p. 46–67. A critical text of Prose 2 is missing to date: it is my aim to provide an edition based on the entirety of the recensio (in the context of my doctoral project at Università Ca’ Foscari Venezia and École Pratique des Hautes Études, under the direction of Marco Infurna and Fabio Zinelli).
11Ibid., p. 55–58.
12See for example Le Roman de Troie en prose (Version du Cod. Bodmer 147), ed. by Françoise Vielliard, Cologny/Genève, Fondation Martin Bodmer, 1979, p. 8 (« […] l’Historia Destructionis Trojae de Guido delle Colonne terminée en 1287, vraisemblablement issue elle-même de la version “méridionale” du Roman de Troie en prose »); Louis-Ferdinand Flutre, Francine Mora, « Benoît de Sainte-Maure », in Geneviève Hasenohr, Michel Zink (dir.), Dictionnaire des lettres françaises. Le Moyen Âge, Paris, Fayard, 1992, p. 139–141, p. 140 (« Même l’ouvrage en prose latine de l’Italien Guido delle Colonne, l’Historia destructionis Trojae (1287), qui prétend suivre Darès, démarque en fait le roman de Benoît [et sans doute aussi la version méridionale de sa mise en prose] »).
13Arianna Punzi, “Le metamorfosi di Darete Frigio: la materia troiana in Italia (con un’appendice sul ms. Vat. Barb. Lat. 3953)”, Critica del testo, VII/1, 2004, p. 163–211, p. 184.
14Among these there are first of all structural issues, such as the absence of the description of Hector’s tomb (see infra). Also relevant is the duel between Penthesilea and a Greek prince during the twenty-first battle: in the majority of the witnesses of the Roman de Troie and in Guido it is Menelaus, whilst in Prose 2 and in the codices E and R of the poem it is Ulysses. No less important are at last some onomastic peculiarities found in the French mise en prose – which are precise marks that links the text to specific branches of the poem’s tradition – that are absent in the Historia.
15Jung, La légende de Troie…, op. cit., p. 485: « K. Chesney croyait que Guido avait utilisé Prose 2, ce qui est évidemment une erreur. »
16Prose 5 has likely been composed between the end of the 13th and the beginning of the 14th century: see Barbieri, “Roman de Troie, Prose 5”, art. cited, p. 823. A critical edition of this text has recently been provided by Anne Rochebouet: see Le Roman de Troie en prose. Prose 5, ed. by Anne Rochebouet, Paris, Classiques Garnier, 2021.
17A critical edition of the entirety of Prose 3 is still missing. Regarding Prose 4 see Le Roman de Troie en prose (Version du Cod. Bodmer 147), op. cit.
18Alice Ducati, La prosa latino-francese di argomento troiano del codice Barb. Lat. 3953 e la fortuna medievale della materia troiana in Italia, doctoral dissertation, dir. by Luca Crescenzi, Università degli Studi di Trento, 2018, p. 255.
19In any case, this opinion appears to be prevalent among the scholars who raise the question of Guido’s source in these terms; see, for example, Arianna Punzi: “Guido si serve largamente di Benoît de Sainte-Maure, o, ma sembra ipotesi meno probabile, di una sua prosificazione, come ha sostenuto K. Chesney” (“Omero Sire?”, in Franco Montanari, Stefano Pittaluga (dir.), Posthomerica I. Tradizioni omeriche dall’Antichità al Rinascimento, Genova, Dipartimento di archeologia, filologia classica e loro tradizioni, 1997, p. 85–98, p. 97); see also Stefania Cerrito, who assumes that Guido used a manuscript of the Roman de Troie (“Dialogando intorno alla leggenda troiana: dal Roman de Troie alle Historiae destructionis Troiae”, in Giovanni Borriero et al. (dir.), Amb. Dialoghi e scritti per Anna Maria Babbi, Verona, Fiorini, 2016, p. 185–196, p. 192).
20Guido delle Colonne, Historia destructionis Troiae, ed. by Nathaniel Edward Griffin, Cambridge (Mass.), Mediaeval Academy of America, 1936.
21Ezio Franceschini, review of “Guido de Columnis, Historia destructionis Troiae, ed. by Nathaniel Edward Griffin”, Bollettino di Filologia Classica, 11–12, 1937, p. 288–296.
22Guido delle Colonne, Historia destructionis Troiae, op. cit., p. xi. It should be noted that at present there is still no complete list of all the witnesses of the Historia: Marc-René Jung states that he knows some 240 manuscripts but does not provide any further detail (Jung, La légende de Troie…, op. cit., p. 565). A list of 56 shelf-marks is given by Nicola De Blasi in his edition of the Neapolitan vernacular version (Libro de la destructione de Troya. Volgarizzamento napoletano trecentesco da Guido delle Colonne, ed. by Nicola De Blasi, Roma, Bonacci, 1986, p. 34). A decisive step forward was recently made in Alice Ducati’s doctoral dissertation, which includes in the appendix a long list of just under 300 manuscripts of the Historia (Ducati, La prosa latino-francese…, op. cit., p. 337–355). This summary is highly useful, but it is nevertheless to be regarded as a « mere working tool » (ibid. p. 337–338), and does not entirely satisfy the need for a complete and detailed recensio.
23Guido delle Colonne, Historia destructionis Troiae, op. cit., p. xi. This approach has been effectively criticized by Ezio Franceschini, who in his review points out that just from the varia lectio of the incipit (“Licet/Etsi/Si et”) it is possible to set up some form of preliminary organization of the vast set of witnesses. Further conjunctive loci are reported in Libro de la destructione de Troya…, op. cit. and id., “Il rifacimento napoletano quattrocentesco della Historia destructionis Troiae. I. Rapporti con la tradizione latina e con i volgarizzamenti conosciuti”, Medioevo romanzo, VI, 1979, p. 93–134, here p. 110–113 and in Ducati, La prosa latino-francese…, op. cit., p. 256–258.
24The manuscripts in question, which are dated between 1334 and the middle of the 14th century, are the only ones for which a precise reference is provided by Griffin: London, British Library, Add. 36671 (= A), Cambridge, Houghton Library, lat. 35 (= C), London, British Library, Harley 4123 (= H), Paris, BnF, lat. 5694 (= P1), Paris, BnF, lat. 5695 (= P2), London, British Library, Add. 15477 (= A1), London, British Library, Add. 22155 (= A2) and London, British Library, Royal 15.C.XVI (= R); I have updated the shelf-mark where needed.
25Benoît de Sainte-Maure, Le Roman de Troie…, op. cit., vol. VI, p. 67–105, stemma on p. 105. The partitioning of the witness in two families, later taken up and developed by Léopold Constans, was previously proposed by Paul Meyer (“Fragments du Roman de Troie”, Romania, 18, 1889, p. 70–106, here p. 99).
26The data here is taken from Jung, La légende de Troie…, op. cit., p. 24–25; description of the manuscripts p. 16–330. Marc-René Jung amplifies Léopold Constans’ recensio bringing the list to thirty complete manuscripts and twenty-eight fragments, but he does not change the general classification hypothesis. I have added a few fragments to the table, to begin with Brussels, Bibliothèque Royale, II 5606, vol. II, fragment 3 (= B5) and New Haven, Yale University Library, Beinecke Rare Book & Manuscript Library, Beinecke MS 1132, olim Torino, Collezione Erminio Morselli (= T). Both were known to Marc-René Jung who, however, omits the first one from his classification, recognizing it in its description as the bearer of a text of the second family (Jung, La légende de Troie…, op. cit. p. 311), and does not give information on the second, referring to a future edition of Giuliano Gasca Queirazza which never materialized (Ducati, La prosa latino-francese…, op. cit., p. 32–33). Conversely, unknown to Marc-René Jung was Archivio di Stato di Vicenza, Collegio dei Notai, busta 48, c. 22v (= Vi), for which see Lorenzo Tomasin, “Un disperso estratto del Roman de Troie all’Archivio di Stato di Vicenza”, Quaderni Veneti, 6/1, 2017, p. 9–26; for its location in the stemma, see Matteo Cambi, “Prime indagini sulla circolazione veneta del Roman de Troie di Benoît de Sainte-Maure”, Quaderni Veneti, 5/1, 2016, p. 1–22. Lastly, three more fragments that I haven’t been able to properly classify should also be added to the table: Amsterdam, Universiteitsbibliotheek (= A3), S4 from the private collection of Martin Schoyen and Tübingen, Wilhelmsstift Bibliothek (= T4) reported in Françoise Vielliard, “Marc-René Jung et le Roman de Troie”, Vox Romanica, 75, 2015, p. 241–255, p. 245–246.
27Benoît de Sainte-Maure, Le Roman de Troie…, op. cit., vol. VI, p. 74, 91 and see the stemma p. 105.
28Ibid.
29Jung, La légende de Troie…, op. cit., p. 24–25.
30Ibid., p. 26.
31Ibid., p. 27–37.
32Some of these narrative passages are illustrated in Vielliard, “Marc-René Jung…”, art. cited, p. 249–250.
33Jung, La légende de Troie…, op. cit., p. 30–32.
34The episode can be found in book XIX, p. 161–162 of Griffin’s edition (Guido delle Colonne, Historia destructionis Troiae, op. cit.). Léopold Constans notes however how the sentence Ex guerra enim amor procedere numquam potest (p. 162), pronounced by Hector in his last reply, seems to follow the line Amor de guerre pas ne vient from the « long version » of the meeting (referring to a bipartite classification of the episode now surpassed by Marc-René Jung’s studies), Benoît de Sainte-Maure, Le Roman de Troie…, op. cit., vol. VI, p. 320.
35Jung, La légende de Troie…, op. cit., p. 32–33 and 35.
36Guido delle Colonne, Historia…, op. cit., book XXI, p. 171: De quibus Dares et earum aspectibus multa descripsit, que magis instar habent inanium sompniorum quam certitudinem ueritatis, licet dictus Dares fuerit professus ea uera fuisse, et ideo de eis obmissum est in hac parte (“Dares fully describes them and their appearance which seem to be empty dreams rather than factual truths, and therefore they are omitted in this place”, translation by Mary Elizabeth Meek; see Guido delle Colonne, Historia destructionis Troiae, Bloomington/London, Indiana University Press, 1975, p. 164).
37Ibid., book XXVII, p. 208: Eius autem sepulture preciosam formam et modum describere superfluum uisum est (“It appears superfluous, however, to describe the costly appearance and fashion of the tomb”, transl. Meek, p. 200).
38Chap. IV, see Dares Phrygius, De excidio Troiae Historia, éd by F. Meister, Leipzig, Teubner, 1873, p. 6.
39Benoît de Sainte-Maure, Roman de Troie…, op. cit., v. 2930–2958; see in particular regarding Andromache v. 2949–2950: Des treis filles ot non l’ainz nee / Andromacha: mout fu senee.
40Ibid., VI, p. 32.
41See Benoît de Sainte-Maure, Le Roman de Troie. Extraits du manuscrit Milan, Bibliothèque ambrosienne, D 55, édités, présentés et traduits par Emmanuèle Baumgartner et Françoise Vielliard, Paris, Le livre de poche, 1998, p. 637.
42These are Paris, BnF, fr. 783 (= D), Paris, BnF, fr. 794 (= E), Firenze, Biblioteca Riccardiana, 2433 (= F1), ’s-Gravenhage, Koninklijke Bibliotheek, 75 G 68 (= G1), Paris, BnF, fr. 1450 (= H) and Nottingham, Wollaton Library Collection, WLC/LM/6 (= N4); M1 is Montpellier, Bibl. interuniversitaire, sec médecine., H.251 (Jung, La légende de Troie…, op. cit., p. 27).
43Paris, BnF, Arsenal 3340 (= A1), London, British Library, Add. 30863 (= L2); for descriptions see Jung, La légende de Troie…, op. cit., p. 134–139 and 100–110.
44Guido delle Colonne, Historia destructionis Troiae, op. cit., book V, p. 45 (« The first of the daughters was named Creusa. She is supposed to have been the wife of Aeneas », transl. Meek, p. 43).
45Jung, La légende de Troie …, op. cit., p. 134.
46Patricia Stirnemann, “Some Champenois Vernacular Manuscripts and the Manerius Style of Illumination”, in Keith Busby et al. (dir.), Les manuscrits de Chrétien de Troyes, Amsterdam/Atlanta, Rodopi, 1993, vol. I, p. 195–226, p. 205 and 209. At least this codex in particular could not have been available to Guido due to its localization and circulation’s peculiarities.
47Quant conquis furent li chastel / Tenedon et Lauriëntel, Benoît de Sainte-Maure, Roman de Troie…, op. cit., v. 6065–6066.
48Jung, La légende de Troie…, op. cit., p. 27. Manuscripts: Paris, BnF, fr. 1610 (= J), Paris, BnF, NAF 6774 (= P), Paris, BnF, fr. 2181 (= K).
49Guido delle Colonne, Historia destructionis Troiae, op. cit., p. 101–102.
50Benoît de Sainte-Maure, Roman de Troie…, op. cit., v. 8337–8342: Patroclus le fiert en l’escu / De tel aïr, de tel vertu / Qu’outre en passe li fers bruniz / E l’enseigne de vert samiz: / Sor le hauberc la lance archeie; / Escliz en volent, si peceie,.
51Benoît de Sainte-More et le Roman de Troie, ou les métamorphoses d’Homère et de l’épopée gréco-latine au moyen-âge, ed. by Aristide Joly, Paris, Franck, 1870–1871, v. 8337–8342: Hector le fiert parmi l’escu / De tel force et de tel vertu / Qu’oltre en passe li fers bruniz / Et l’enseigne de vert samis. / Patroclus sor la sele pleie, / L’arçon detriès brise et peceie. In its edition, prior to Léopold Constans’, Aristide Joly prints the text of K of the second family.
52Jung, La légende de Troie…, op. cit., p. 26.
53Guido delle Colonne, Historia destructionis Troiae, op. cit., p. 133: Quem ut presentit Patroclus, qui prime aciei Grecorum, ut dictum est, ductor erat, contra Hectorem in equi sui cursu uelociter dyametro recto procedens Hectorem acuta lancea sua petit, et cum ea uiriliter ipsum inpellens scutum quod sibi Hector opposuit uiriliter perforauit […] (“When he perceived this, Patroclus, who, as was said, was the leader of the first battalion of Greeks, advanced swiftly, charging on his horse in a straight line against Hector, and sought him with his sharp spear, and since he hurled it manfully, he pierced the shield which Hector manfully opposed to it […]”, transl. Meek, p. 128).
54Jung, La légende de Troie …, op. cit., p. 26.
55Manuscripts: Paris, BnF, Arsenal 3340 (= A1), London, British Library, Add. 30863 (= L2), Paris, BnF, fr. 1553 (= I), Sankt Peterburg, Rossijskaja Nacional’naja Biblioteka, fr. F. v. XIV. 6 (= S1).
56Manuscripts: Città del Vaticano, BAV, Reg. Lat. 1505 (= R), Paris, BnF, Arsenal 3342 (= A2).
57Benoît de Sainte-Maure, Roman de Troie…, op. cit., v. 16649–16719.
58For a general examination of the subject of Hector’s tomb, see still Marc-René Jung, “Hector assis”, in Georges Lüdi, Hans Stricker, Jakob Wüest (dir.), Romania ingeniosa. Mélanges offerts à Gerold Hilty à l’occasion de son 60e anniversaire, Bern/Frankfurt am Main/New York/Paris, Peter Lang, 1987, p. 153–169; see also id., La légende de Troie…, op. cit., p. 33–34.
59Enz en la voute de desus / L’ont gentement posé et mis / E dedenz la chaeire asis, Benoît de Sainte-Maure, Roman de Troie, ed. Constans, v. 16766–16768.
60Jung, « Hector assis », art. cited, p. 156. Manuscripts: Paris, BnF, fr. 12600 (= L), London, British Library, Harley 4482 (= L1), Napoli, Biblioteca Nazionale Vittorio Emanuele III, XIII c. 38 (= N), Paris, BnF, fr. 903 (= G), Paris, BnF, fr. 821 (= F).
61For the descriptions see Jung, La légende de Troie…, op. cit., p. 274–287 and 212–214.
62Guido delle Colonne, Historia destructionis Troiae, op. cit., p. 178, emphasis added (“Thus, when the design of this tabernacle had been completed, they placed around it an enclosure of ebony wood which could be closed sometimes and opened sometimes, so that the body of Hector, so placed, might be seen more plainly by those wishing to see how it looked”, transl. Meek, p. 171).
63Benoît de Sainte-Maure, Roman de Troie, ed. Constans, op. cit., v. 16786–16793: Quant la chose fu a chief traite, / Si ont une image levee / Qu’a merveille fu esguardee: / De fin or fu resplendissant / E a Hector su resemblant / Que nule chose n’i failleit. / Un brant d’acier tot nu teneit, / Grezeis par signe manaçot.
64Benoît de Sainte-Maure, Roman de Troie, ed. Joly, op. cit., v. 16740–16745: Quant la chose fu a chief trete, / Si ont l’ovre aïtant menée / Que merveilles fu esgardée. / En la chaiere Hector seeit, / Un brant d’acier tot nu teneit / Grezeis par signe manaceit. P3 is Paris, BnF, fr. 1420.
65Quant la chose fu a chief traite, / Si ont l’uevre si haut levee / Qu’a merveille fu esgardee. / La guille fu trop preciouse / Et trop riche et trop mervillouse. / Maint marc de fin argent et d’or / I mist li rois de son tresor. / Sour la guille levee en alt, / Illuec endroit u ele faut, / Ont une ymage tresjetee, / Ki le l’ost pot estre esgardee. / D’or le fisent resplendissant / Et a Hector si resamblant / Que nule chose n’i failloit. / .I. brant d’achier tout nu tenoit, / Grigois par signe manechoit, reading of I, from Jung, « Hector assis », art. cited, p. 158.
66Guido delle Colonne, Historia destructionis Troiae, op. cit., book XXII, p. 177, emphasis added: “On high, above the peak of the canopy of this tabernacle, the aforesaid master craftsmen fixed a certain golden statue, representing a likeness of Hector, with his naked sword in his hand, and his glance and expression were toward that place where the Greek army was staying in its tents, so that he seemed to offer threats to the Greeks with his naked sword”, transl. Meek, p. 170.
67Léopold Constans argues that “Guido se servait d’un ms. de la 2e famille” (Le Roman de Troie…, op. cit., vol. VI, p. 328), but the few findings he has made seem isolated and not always completely convincing. In “Dialogando…” (art. cited, p. 192), Stefania Cerrito observes that the exchange of names between Andromache and Creusa is a likely proof that “il manoscritto di Guido si collochi in questo ramo testuale [the second family]”.
68Eleven if one considers that is also Italian N, whose origin is disputed. For an updated survey on the Italian witnesses of the Roman de Troie see Ducati, La prosa latino-francese…, op. cit., p. 7–37.
69Manuscripts: Milano, Biblioteca Ambrosiana, D 55 sup. (= M2), the already mentioned F, Paris, BnF, fr. 782 (= C), Wien, Österreichische Nationalbibliothek, 2571 (= W), Venezia, Biblioteca Nazionale Marciana, fr. XVIII (= V2) and Sankt Peterburg, Rossijskaja Nacional’naja Biblioteka, fr. F. v. XIV. 3 (= S).
70Cambi, “Prime indagini…”, art. cited.
71The already mentioned F1 is from Florence, while the precise localization of R and P is uncertain.
72Jung, La légende de Troie…, op. cit., p. 287–291.
73Keith Busby, Codex and Context. Reading Old French Verse Narrative in Manuscript, Amsterdam/New York, Rodopi, 2002, p. 607–608; Ilaria Molteni, “Il mito troiano e i signori d’Italia: decorazione e ricezione dei Roman de Troie istoriati trecenteschi”, in Nicolas Bock, Ivan Foletti, Michele Tomasi (dir.), Survivals, Revivals, Rinascenze: Studi in Onore di Serena Romano, Roma, Viella, 2017, p. 365–377, p. 371 ff.
74Günther Holtus, Peter Wunderli, Franco-italien et épopée franco-italienne (GRLMA, vol. 3, t. 1/2, fasc. 10), Heidelberg, Winter, 2005, p. 210; Gabriele Giannini, Produzione e circolazione manoscritte del romanzo francese in versi dei secoli XII e XIII in Italia, doctoral dissertation, La Sapienza Università di Roma, 2003, p. 241–242; Sebastiano Bisson, Il fondo francese della Biblioteca Marciana di Venezia, Roma, Edizioni di Storia e Letteratura, 2008, p. 81. See also Costanza Cipollaro, “Invenzione e reinvenzione negli esemplari miniati del Roman de Troie tra Francia e Italia”, in Costanza Cipollaro, Michael Viktor Schwarz (dir.), Allen Mären ein Herr/Lord of all Tales, Wien/Köln/Weimar, Böhlau Verlag, 2017, p. 19–56, here p. 46.
75Vielliard, “Marc-René Jung…”, art. cited, p. 248.
Auteur
-
Jacopo Fois
Università Ca’ Foscari Venezia – École pratique des hautes études
Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Marquer la ville
Signes, traces, empreintes du pouvoir (xiiie-xvie siècle)
Patrick Boucheron et Jean-Philippe Genet (dir.)
2013
Église et État, Église ou État ?
Les clercs et la genèse de l’État moderne
Christine Barralis, Jean-Patrice Boudet, Fabrice Delivré et al. (dir.)
2014
La vérité
Vérité et crédibilité : construire la vérité dans le système de communication de l’Occident (XIIIe-XVIIe siècle)
Jean-Philippe Genet (dir.)
2015
La cité et l’Empereur
Les Éduens dans l’Empire romain d’après les Panégyriques latins
Antony Hostein
2012
La délinquance matrimoniale
Couples en conflit et justice en Aragon (XVe-XVIe siècle)
Martine Charageat
2011
Des sociétés en mouvement. Migrations et mobilité au Moyen Âge
XLe Congrès de la SHMESP (Nice, 4-7 juin 2009)
Société des historiens médiévistes de l’Enseignement supérieur public (dir.)
2010
Une histoire provinciale
La Gaule narbonnaise de la fin du IIe siècle av. J.-C. au IIIe siècle ap. J.-C.
Michel Christol
2010
