We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 11 February 2009
In recent times, a previously unchallenged and longstanding communis opinio concerning the extant manuscript tradition of Valerius Flaccus' Argonautica has been shattered by Prof. W.-W. Ehlers in his revelation that the fifteenth-century Laurentianus plut. 39.38, L, written by the Florentine scholar, Niccolò Niccoli, is independent of the much exalted oldest witness, Vaticanus Latinus 3277, V, copied in Fulda in the second quarter of the ninth century. With equally silent subservience to the hazardous and now discredited principle, vetustissimus et optimus, second place in the tradition had commonly been given to a manuscript of similar age to that of V, namely S, discovered in 1416 at the monastery of St. Gall by three humanist scholars, Poggio Bracciolini, Cencio Rustici and Bartolomeo da Montepulciano, and subsequently lost, except through reconstruction from six extant fifteenth-century apographa, each instantly recognizable from the fact that they contain only 1.1–4.317. Between V and S, however, as Ehlers and others before him have argued, a common exemplar, a, must be deduced to exist.
1 Ehlers, W.-W., Untersuchungen zur handschriftlichen Ueberlieferung der Argonautica des C. Valerius Flaccus (Munich, 1970), pp. 44 ff.Google Scholar; further information on L may be found in Cambier, G., ‘Attribution du Laur. 39.38 a Niccolò Niccoli’, Scriptorium 19 (1965), 236–43CrossRefGoogle Scholar; Ehlers, , Untersuchungen, pp. 11–12, 42ff., 83 ff.Google Scholar; Labardi, L., ‘Congetture del Niccoli eTradizione Estranea all' Archetipo sui Margini del Laurenziano 39,38 di Valerio Flacco’, Italia Medioevalee Umanistica 26 (1983), 189–213Google Scholar. The provenance of V was established by Lehmann, P. (Erforschung des Mittelalters, iii [Stuttgart, 1960], p. 180)Google Scholar and the date by Bischoff, B. (ap. Ehlers, Untersuchungen, p. 19)Google Scholar. For further details on V, cf. Courtney, E., C. Valeri Flacci Argonauticon Libri Octo (Leipzig, 1970), pp. x ff.Google Scholar; Ehlers, , Untersuchungen, pp. 18 ff., 102 ff.Google Scholar; Chatelain, E. L. M., Paléographie des Classiques Latins (Paris, 1884/1900)Google Scholar, plate CLXV.
2 The apographa are Matritensis Bibl. Nat. 8514 (X), Vaticanus Latinus 1613 (P; contains only 1.1^.199), Vat. Ottobonianus 1258 (O), Oxoniensis Reginensis 314 (Q), Malatestianus Caesenas S XII 3 (Mai) and Vaticanus Latinus 1614 (II). The principal literature concerning S and the relationships of its apographa is contained in Getty, R. J., The Lost St. Gall Ms. of Valerius Flaccus (Aberdeen, 1934)Google Scholar; Ehlers, , Untersuchungen, pp. 24ffGoogle Scholar..
3 cf. Getty, , op. cit. (n. 2), pp. 23 ff.Google Scholar; Ehlers, , Untersuchungen, pp. 32 ff.Google Scholar; Taylor, P. R., Gai Valeri Flacci Argonauticon Liber Quartus Edited with an Introduction and Commentary (Diss., Birmingham, 1987), pp. 30–2Google Scholar.
4 Discovered in 1961, this manuscript has been reproduced, transcribed, edited and published by V. Branca and M. Pastore Stocchi (Florence, 1972). The significance of its discovery with respect to the tradition of Valerius Flaccus is discussed in much greater detail by Schmidt, P. L., ‘Polizian und die Valerius-Flaccus-Ueberlieferung’, IMU 19 (1976), 241–56Google Scholar.
5 Untersuchungen, pp. 103–4.
6 cf. Schmidt, , art. cit. (n. 4), 244Google Scholar.
7 cf. ibid., 241–3.
8 The physical composition of γ has been ingeniously constructed on the basis of various duplications, transpositions and omissions in its apographa by Ehlers, , ap. Schmidt, art. cit. (n.4), 246 ffGoogle Scholar. For the nature of the script of this manuscript, see Taylor, , op. cit. (n. 3), p. 11Google Scholar.
9 Schmidt, , art. cit. (n. 4), 250 ffGoogle Scholar. He cites (p. 252 n. 2) as authority for his opinion on the use of the Argonaulica by Lovati, and Mussato, the article of Guido Billanovich, ‘Veterum Vestigia Vatum’, IMU 1 1958, 178 ffGoogle Scholar. The reliability of much of the earlier part of this article has however been the subject of great dispute; cf., e.g., Ludwig, W., ‘Kannte Lovato Catull?’, RhM 129 (N.F.) (1986), 329–30Google Scholar. Nevertheless, as regards the evidence that Billanovich provides for the influence of the Argonautica on Lovati and Mussato, the following resemblances seem, in my opinion, sufficiently striking as to afford strong indication of imitation: Lovati Ep. 3.8 ‘carmina barbarico passim deformat hiatu’, cf. Arg. 8.69; Lovati, Ep. 4.3–4Google Scholar ‘sic te socii iactura pericli exagitat”, cf. Arg. 8.403–4; Lovati, Ep. 4.41Google Scholar ‘cur rosee fraudatrix cruda iuvente’, cf. Arg. 8.257; Lovati, Ep. 5.8Google Scholar ‘seu sic Parcarum nevit iniqua colus’, cf. Arg. 6.645 (loose but surely clear verbal imitation); Mussato, , De obsid. 1.2–3Google Scholar, ‘…, Clio,…, ede virum’, cf. Arg. 3.15–16. No other classical or post-classical source, as far as I can ascertain by consultation of concordances to Catullus, Lucretius, Vergil, Ovid, Horace, Propertius, Tibullus, Juvenal, Persius, Seneca, Lucan, Silius Italicus, Statius, Martial, Claudian, Ausonius, Boethius and Prudentius (I have made no assumptions as to the number of poets known to Lovati and Mussato in this survey in order to examine the question as thoroughly as possible; however, it will be noted that there are unfortunately certain omissions in the list, e.g. Martianus Capella, where no concordance was available to me), may rival the claim of Valerius Flaccus to have provided the inspiration for these lines of the Paduan humanists.
10 Ven. Marc. lat. x 51 (3503), f. 18v; cited by Meneghel, R., IMU 16 (1973), 177–8Google Scholar.
11 The individual manuscripts and their contents are listed by Ehlers, , Gai Valeri Flacci Setini Balbi Argonauticon Libri Octo (Stuttgart, 1980), pp. xiv–xvGoogle Scholar; cf. also Ullman, B. L., CPh 26 (1931), 21–7Google Scholar; Ehlers, , Untersuchungen, pp. 22–3, 107–13Google Scholar.
12 Ullman, , art. cit. (n. 11), 25–6Google Scholar; Courtney, , op. cit. (n. 1), pp. xxx–xxxiGoogle Scholar; where Ullman and Courtney refer to errors that could not be corrected by medieval conjecture, except in the cases of 1.331 (‘deficiamus’ a), 1.579 (‘aperte’ V), 1.589 (‘oenatria’ V) and 1.595 (‘refringet’ a), understand y instead of V. Schmidt (Gymnasium 81 [1974], 263) also expresses his belief in the independence of the florilegia from the Italian branch of the tradition.
13 CPh 81 (1986), 58–60Google Scholar.
14 In the preface to the castigationes of the second edition, Carrio writes ‘quos [libros manuscriptos] certe in toto hoc nostro Belgio, praeter ilium meum, reperire adhuc potui nullos.’
15 I am grateful to the anonymous referee of this article for pointing out that Carrio may not have intended the word ‘sexcentos’ to be understood in the literal sense of ‘six hundred’ but in the common transfigurative sense which in modern parlance might be translated as ‘hundreds’. I accept, moreover, the referee's scepticism concerning the accuracy of the judgement ‘ante sexcentos annos conscriptum’, if a literal interpretation was intended, in view of the difficulty encountered even today of identifying a manuscript of the tenth century. Nevertheless, what is clear is that the sixteenth-century Belgian scholar is claiming to be in possession of a manuscript that was many centuries old.
16 N. Heinsius and F. Eyssenhardt, ap. Thilo, G., C. Valeri Flacci Setini Balbi Argonauticon Libri Octo (Halle, 1863), p. lxxGoogle Scholar; Giarratano, C., C. Valerii Flacci Balbi Setini Argonauticon Libri Octo (Milan, 1904), p. xxxivGoogle Scholar; Schmidt, P. L., Gymnasium 81 (1974), 263Google Scholar.
17 Thilo, , op. cit. (n. 16), pp. lxx–lxxxvGoogle Scholar; Schenkl, C., ‘Studien zuden Argonautica des Valerius Flaccus’, Kaiserliche Akademie der Wissenschaften. Philosophisch-Historiche Classe. Sitzungsberichte 68 (1871), 315 ff.Google Scholar; Krenkel, P., De codicis Valeriani Carrionis auctoritate (Luckae, 1909)Google Scholar; Kramer, O., C. Valerii Flacci Setini Balbi Argonauticon (Stuttgart, 1913), pp. xlvi ff.Google Scholar; Courtney, E., op. cit. (n. 1), pp. xxvi ffGoogle Scholar..
18 cf. Dolbeau, F., RecAug 13 (1978), 33Google Scholar; id., RecAug 14 (1979), 227.
19 Ehlers, , ‘Valerius-Probleme’, MH 42 (1985), 339–40Google Scholar.
20 cf. Schenkl, , art. cit. (n. 17), 317 ff.Google Scholar; Krenkel, , op. cit. (n. 17), pp. 22 ff., 41Google Scholar; Ehlers, , MH 42 (1985), 340Google Scholar.
21 Brux. Bibl. Roy. 10012.
22 The hand of Carrio is identified by a subsequent owner of Bodl. Line. 8° F 107, J. C. Geuartius (1593–1666), who has added at the beginning of the text of Bk 1: ‘Collatu cum membranaceo codice Gemblacensi manu Lud. Carrionis’. I wish to express my thanks to Prof. M. D. Reeve of Cambridge for informing me of the existence of this collation and to Prof. G. P. Goold of Yale for lending me a photographic reproduction of Brux. 10012 itself. For the purpose of the following investigation, I have examined Carrio's work on Book 1 alone, considering this a sufficiently large proportion of the total material to be considered representative.
23 It appears that Carrio completed his collation before 1569 on the evidence of the following information provided by Geuartius on a page added to Bodl. Line. 8° F 107 after the dedication: ‘Veteri codici Membranaceo, in quo rubrica enotatum (?) erat, Liber S11 petri Gemblacensis, adscriptum erat: “Hunc librum D. Iacobus Susius utendum recepit (?) Mechlinie a D. Ludovico Carrione, I.C. An. 1569”’. This and two other notes mentioned at the same time by Geuartius are no longer visible in Brux. 10012; I suspect that they were excised with the bottom of f. 99.
24 An exception occurs at 1.800, where the ‘vel’ symbol follows the reading ‘Augustus’.
25 On these occasions, the ‘vel’ symbol is placed over the original ‘v.’ and therefore often follows the collated reading.
26 
27 On two occasions, similarly, readings of G appear in Carrio's collation under the label ‘vel’: 1.144 vel 'rapacia‘ Carrio rap acia printed ed.; 1.800 ‘Augustus’ vel Carrio Augustos printed ed.. Both of these readings, however, correct trivial errors of the printed edition and are thus no doubt genuine, independent corrections of the collator.
28 1.7 hunc; l.lO canenda; 1.78 se quisque satis novisse putabant; 1.155 fluctusque natantes; 1.271 fulgens. I discount two examples where the unlabelled marginal note is clearly no more than a spontaneous correction of a trivial error in the printed edition:
1.59 percepta pr. ed,: praecepta G Carrio
1.275 atqui pr. ed.: at qui G Carrio
29 1.271 augusto v. fulgens
30 1.7 ff.… scribendum itaque, poscente sententia, versu primo: Hunc mihi tu Caesar palriaeprincepsq. paterque et mox, iubente libro Gemblacensi: viresque facis ad tanla canenda. 1.75ff.…eamqu e liber Gemblacensis in nobis probaret. in quo est:…se quisque salis novisseputabant. 1.155…scribe: -fluctusque natantes. cf. Pasquali, G., Sloria delta tradizione e critica del testo (Florence, 1952), pp. 70–1Google Scholar, who reports the findings of H. Blass with respect to the work of both Carrio and the contemporary scholar, Modius, on the extant Censorinus tradition. According to Blass, it is the practice of both scholars to introduce manuscript readings with a variety of terms, including scribe, emenda, malim etc. Whereas Modiu s is guilty of the occasional error in the attribution of a reading to a source, moreover, ‘Quant o a notizie positive il Carrion sarebbe in sè piú credibile del Modio, perché, se questi si sbaglia di rado, il Carrion non si sbaglia mai’ (p. 71).
31 1.271 angusto G: augusto Carrio; 1.795 claudu G: claudi Carrio (i.e. claudii, cf. p. 457 para. 2).
32 1.361 tune Carrio: turn G [it appears that Carrio intended to correct an identical error four lines below at 1.365, where G exhibits tune and the printed edition turn; turn in 1.361 was underlined accidentally]; 1.877 affetus Carrio: affectus G [Carrio was primarily concerned with the prefix, which appeared as ‘ef-’ in the printed edition].
33 I confess that I cannot be sure that this is what Carrio has written, since the last letter is virtually obscured by a second-hand correction to ‘rapti’.
34 Schenkl, , art. cit. (n. 17), p. 317Google Scholar.
35 cf. Ehlers, , MH 42 (1985), 340Google Scholar.
36 The siglum Carr. 1 indicates the reading adopted by Carrio in the text of the 1565-edition, Carr. 2 the reading adopted in the text of the 1566–edition.
37 Thilo, , op. cit. (n. 16), p. lxxviiGoogle Scholar.
38 cf. ibid., pp. lxxxiii–iv; Schenkl, art. cit. (n. 17), 317–20; Krenkel, , op. cit. (n. 17), pp. 16 ffGoogle Scholar.
39 cf. Thilo, , op. cit. (n. 16), pp. lxxi ff.Google Scholar; Krenkel, , op. cit. (n. 17), pp. 9 ffGoogle Scholar.
40 Untersuchungen, p. 86. Although Carrio actually makes no reference to his manuscript at any of the lines omitted by V (i.e. 1.45; 2.565A; 7.579–80), Ehlers' argument nevertheless stands ex silentio, since Carrio shows sufficient consciousness of other such omissions both in his own manuscript (e.g. 2.136; 5.542; 6.195–6; 8.106 ff.) and in the rest of the tradition (e.g. 1.830; 3.77; 7.633) for us to be confident that he would also have acknowledged these particular omissions, if they had been present. It is very unlikely, moreover, that in the period in which he was working he would have known or perceived the advantage of covering up certain errors whilst revealing others. It would be inappropriate therefore to charge him with engineering the facts.
41 These manuscripts are listed by Ehlers, in Gai Valeri Flacci Setini Balbi Argonaulicon (Stuttgart, 1980), pp. x–xivGoogle Scholar.
42 A list of errors introduced to the tradition by L has been compiled by Ehlers, , Untersuchungen, pp. 79–81Google Scholar. Missing from this list (although reported in his edition [cf. previous note] are the following: 1.38 timendis L: timendus V; 1.96 Macedumque L: macetumque V; 1.371 Caphareu L: Caphereu V. Ehlers also overlooks the first-hand correction of L at 2.569 to ‘erepta quae’ and at 6.413 misrepresents V, which reads ‘biiuges’ and not ‘biuges'.
43 The resemblance of C (‘ad fremitus’) to Lacl (‘ad fretus’; ‘adflatus’*) at 6.441 is likewise of no significance, since the independent witness of V is absent (om. 6.439–76). If V had exhibited ‘ad fletus’, then there would be reason to suspect that C was contaminated from L. V too however may have exhibited ‘ad fretus’, in which case it could be concluded that the error in all was derived from a common source earlier in the tradition.
44 cf. Thilo, , op. cit. (n. 16), pp. lxxxiii–lxxxivGoogle Scholar; Schenkl, , art. cit. (n. 17), 317 ff.Google Scholar; Krenkel, , op. cit. (n. 17), pp. 24 ff.Google Scholar; Ehlers, , Untersuchungen, pp. 87, 98–9.1Google Scholar discount the few examples where C apparently agrees in error with Bu alone (of which only ‘caelata’ (1.420) is of any significance), since there is no guarantee against an equally probable possibility that Burman himself confused the sources of his collations. Agreements with Ha alone (the collation recorded in Leiden, BPL 760 B 10) can also be ignored, firstly since there is clear evidence that this collation was compiled in several stages and secondly since there is no means of establishing a terminus ante quern for the date of any (cf. Taylor, op. cit. [n. 3], Appendix 2). It is not inconceivable therefore that some of the readings recorded are extracted from or inspired by those exhibited in Carrio's editions.
45 Krenkel, , op. cit. (n. 17), p. 17Google Scholar. He incorrectly cites the following as agreements in error:
1.678 pascit λ: pascet* CbB–1519 Eng. Col. Gryph.
7.165 sacro λ: sacris* C B–1498
7.501 si meritis noctis immemor λ: si meriti sim noctis et immemor* C: sim meriti noctisque nee immemor edd. ante Carr.
46 The following agreements cited by Krenkel belong to this category:
1.34 clusus α: clausus C B–1519 (also L)
2.464 flectus α: fluctus C edd. ante Carr. (also L)
7.579–80 om. V exh. C edd. ante Carr. (also L)
1.227 minsas ambige α: Minyas ambage C edd. ante Carr. (also L)
2.376 disertisque V: desertasque C edd. ante Carr. (also L)
3.25 prima α: primas C edd. ante Carr. (also L)
5.680 accipere V: accipite C B–1474 F–1481 F–1503 (also L)
47 Kramer, , op. cit. (n. 17), p. lxvGoogle Scholar.
48 Schenkl, , art. cit. (n. 17), 319Google Scholar.
49 Pasquali, , op. cit. (n. 30), p. 71Google Scholar.
50 Kramer, , op. cit. (n. 17), pp. lii–liiiGoogle Scholar.
51 Krenkel, , op. cit. (n. 17), p. 27Google Scholar.
52 Labardi, L., IMU 26 (1983), 201 ffGoogle Scholar.
53 cf. Krenkel, , op. cit. (n. 17), p. 10Google Scholar.
54 Untersuchungen, pp. 86–7.
55 That the scribe of C or a close contemporary was not averse to creating verses to fill lacunae in the text is revealed in the information that Carrio provides. After 1.830 and at 3.77, Carrio refers to lines contained in the margins of C, the second of which he describes as being ‘aeque tamen vetusta scriptura’. Characteristic of both is the ‘leonine’ rhyme, betraying the fact that they were composed in medieval times:
Iupiter et primae velit omnia reddere massae C i.m. post 1.830
hostis et exciti dent obvia praelia Colchi C i.m. 3.77
Both, however, conform to the rules of the dactylic hexameter, illustrating at least a reasonable degree of competence in their creator.
56 For the use of ‘visus’ to describe apparitions of substance rather than purely illusory apparitions, cf. Verg. A. 3.36
57 I attribute the loss of line 7.633 to the error saut du même au même; cf. Reynolds, L. D. and Wilson, N. G., Scribes and Scholars 2 (Oxford, 1974), p. 204Google Scholar.
58 For the quality of emendation achieved by the scribe of C, cf. below, p. 470.
59 Significantly C shares no corruptions with these witnesses. It would not be possible, therefore, to argue contamination from this source.
60 The precise relationship of f to C cannot be discerned conclusively from the meagre evidence provided by the former and the second-hand account given of the latter. Even the evidence of content, that is to say the presence in some of the florilegia of lines occurring after the point of mutilation in C (8.105), does not in itself rule out the possibility that C could be the parent of the florilegia, since, as Prof. Ehlers has recently pointed out (MH 42 [1985], 341 n. 37), the existence of 8.408 ff. in the florilegia could be attributed to the loss of lines 8.106 IT. from C after copies had been made and before it was discovered in the sixteenth century. There is however evidence to suggest that the exemplar of the florilegia was incorrectly bound, for the extracts from the first book are presented in an incorrect order (76ff., 39, 248ff.). Since there is no indication of any such dislocation in C and in the absence of any textual evidence indicating the dependence of the florilegia on C, I am inclined to believe that they are independent of C, but transcribed from a common exemplar.
61 cf. Krenkel, , op. cit. (n. 17), p. 52Google Scholar.
62 cf. Krenkel, , op. cit. (n. 17), pp. 14–16Google Scholar.
