An author who was one of the most widely read and copied through to the Renaissance at last has a critical text which scholars can use with confidence. John Briscoe has brought to Valerius Maximus the thoroughness and accuracy we have come to expect from his work on Livy over some twenty years.
The compendious catalogue compiled by Dorothy Schullian bears witness to Valerius’ enormous popularity — over 800 codices are extant, although none dates from before the 9th century. The signal contribution of Briscoe is to make available all the readings of manuscript G, an 11th century codex from Gembloux (Brussels Bibliothèque royale 5336). The significance of this manuscript was first appreciated by Schullian in 19371 and then by C.J. Carter in his unpublished Cambridge 1968 doctoral thesis, but Briscoe’s is the first published text to incorporate readings from G. Although there have been translations and texts of Valerius published since 1937,2 they have reproduced Kempf’s old Teubner text with the occasional incorporation of later conjectures. The most disappointing missed opportunity is seen in the two Budé volumes published since 1993 (one more to emerge): Combès has produced an excellent, fluent translation and serviceable notes, but his text ignores completely the existence of G and any idea of its significance. The appearance of parochialism is increased because Combès, emeritus professor of the university of Montpellier, gives weight to a manuscript from his own university library.3
The only critical text of Valerius worth discussing before Briscoe’s is that of his predecessor in the Teubner series, K. Kempf, whose second edition appeared in 1888. Kempf’s text relied on the two oldest codices of Valerius A and L, which he recognised were so close as to be almost twins. Kempf’s great contribution was to recognise the importance of A (Berne Burgerbibliothek 366) for his 1854 edition and then the importance of L (Flor. Med. Laur. Ashburnham 1899) for the 1888 second edition. However, as Briscoe’s text shows, Kempf’s readings are occasionally inaccurate (e.g. 6.1. ext. 2: the reading of L is namque, not nam neque; Marshall’s damning ‘frequently wrong’ [n. 1 p. 428] seems excessive); in fact Kempf never collated L himself. Nonetheless Kempf’s edition was not without merit — indeed it is only the appearance of Briscoe that supersedes Kempf’s work. The reader who wishes to know the readings of some later manuscripts, all of which Briscoe subsumes under “dett.” will still have to consult Kempf. Briscoe’s policy cleans up his apparatus criticus admirably but it is a minor loss not to see separately the readings of G in particular where the influence of G can be felt.
Following Schullian, Briscoe argues that G has an authority independent of A and L and of their immediate source. Briscoe argues that it is highly unlikely on the many occasions where G presents the correct reading, against A and L, that these are the result of conjectures either by the scribe of G or by his source (e.g. 2.8.5: veriori G; iustiori AL; 3.6.1: diutius G, dium A c, diu munum L).
A casual reading of almost any page of Briscoe’s text will reveal instances where the reading of G agrees with the corrected text of A and L (A c, L c) and is indubitably correct. For example in 5.6.5, eight lines of Teubner text, there are three such readings: devovit A c L c G, devolvit AL; strage A c L c G, stra AL; insperata A c L c G, inseparata AL. G is in very good company here, as we know that the excellent Carolingian scholar Servatus Lupus was responsible for the corrections to A.4 As the corrections to A and L are independent of each other, it is probable that when G makes the same correction it should not be thought to be dependent on either. The independence of G makes it a unique source for readings from its ultimate common origin with L and A which were corrupted by the immediate source of L and A and gives G a value equal to that of A and L in establishing the original reading.
Briscoe uses 4.3. praef. to refute the alternative to this, that A and G share an immediate ancestor, on the grounds that there are only 18 instances where A and G agree in the same error against L, where the latter preserves the correct reading. In his review of Combès Briscoe now regrets that he did not collate Avranches Bibliothèque municipale 157, which derives from A before it was corrected by Servatus Lupus and as such could provide evidence as to the uncorrected reading of A often concealed now by the correction.5
The main benefit derived from the readings of G occurs where G does not agree with corrected A and L, i.e. where new evidence is presented. As Schullian showed, the occasions where G is correct (or where Briscoe considers G correct) fall into two categories: (1) wholly new readings and (2) readings anticipated by Kempf and other scholars either by conjecture or through preference for the reading of Gamma, but which now have a greater weight.
1) 1.5.6:
2) 1.1 ext. 4: quam, quem AL; 1.5.2: transferri, transferre AL (Briscoe here follows AL, Kempf G); 1.6.3: possent, possint LA; 1.6 ext.1: ostentorum, ostentatorum LA; 1.6 ext. 1: vix tandem, iuxta indem AL; 1.8.6: ancipit, ancipitis: LA; 1.8.10: praecucurrit, percuccurrit AL; 2. praef.: felicem, felice AL; 2.1.1: desierint, desierit LA; 2.2.5: relatis Q., relatisque AL; 2.3.1: duxit, duxi AL; 2.3.3: Navio, Naevio AL; 2.4.2: passus, passos AL; 2.4.4: arcessendi, arcersendi AL; 2.4.5: dati, dari LA; 2.4.6: Petreius, preteius AL; 2.6.2: eorum, earum AL; 2.7.6: imbecillitatem, imbecillitate AL; 2.7.6: a, ab AL; 2.7.15: Fulvio, Furio ALP; 2.7 ext. 1: adiutorio, adiutorium AL; 2.7 ext. 2: Lacedaemoniorum, Lacedaemonium L; 2.8.2: si, si ea AL; 2.8.7: externo, hesterno AL; 2.9.2: nam illo, illo nam AL; 2.9.4: ipsae, ipse AL; 2.10.8: imago, inl imago AL; 2.10 ext. 2: dicere, diceret AL; 3.1.1: Lepidus, Laepidus AL; 3.2.21: flagrantissimi, fraglantissimi AL; 3.2 ext. 1: destringere, distringere AL; 3.2 ext. 5: auctius, Athius AL; 3.3 ext. 2: quemquam, quempiam A c; 3.3 ext. 5: fidiculas, fidicula AL; 3.3 ext. 7: comprehensus, compressus AL; 3.3 ext. 7: examinandum, exanimandum AL; 3.4.2: invidioso, indiviso AL; 3.4.4: alito, salito AL; 3.6.1: diutius, diu(m) AL; 3.7.1c: percontatus, percunctatus AL; 3.7.1e: ut, spirantem, sperantem AL; 3.7.5: senatu, senatum LA; 3.7.10: Ostiam, Hostiam AL; 3.7 ext. 6: Prusiam, Plusiam AL; 3.8.2: amore, amorem AL; 4.1.12: tam, tot AL; 4.3.5: perfectissimum, perspectissimum AL; 4.4.5: rubori, roburis AL; 4.5.1: dignitati, dignitatis AL c; 4.6.5: ille, illae AL; 4.6 ext. 3: uxoriae, uxori A, uxoris A c L; 4.7.4: conspectu, conspectum AL; 4.7 ext. 1: iuxerant, iuxerunt AL; 4.8.5: Isthmicum, Isthimicum L; 4.8.5: audierant, audierunt A c L; 5.1.6: speciosa, speciose AL; 5.1.7: ingressum, ingressus AL; 5.1 ext. 2: volente, volenti AL; 5.1 ext. 4: fructum, strictum AL; 5.2.1: noluissent, voluissent AL; 5.2.5: quam Q. quamque AL; 5.2.7: nec, ne AL; 5.3 ext. 2: summa, summae A c L; 5.5.4: imbelli, et belli, AL; 5.6.1: mortifero, mortifico AL; 5.6.4: adfirmassent, adfirmasset AL; 5.6.8: functos, functus AL; 5.6 ext. 1: Athenis, sed, Athenienses AL; 5.8.3: petiit, petit AL; 5.8.3: tam, tamen AL; 5.9 praef.: sed ut, sed et ut AL; 5.9.4: hoc, hunc AL; 6.1.4: cum praesertim, praesertim cum A c, cum praesenti L; 6.4 ext. 1: subicit, subiacet AL; 6.4 ext. 1: dedidisset, dedisset AL; 6.6.5: ac non, annon AL; 6.8.1: iret, irent AL; 6.8.4: ne, nec AL; 6.9.3: religione, regione AL; 6.9.6: se, AL omit; 6.9.14: Arpini, Arpina AL; 6.9.4: sio … felices, inter felicissimos AL; 6.9 ext. 3: coegit, cogit AL; 7.2.6: debet, deberet AL; 7.2 ext. 1: teque totam, et quae totam L; 7.2 ext. 2: consummat, consumat AL; 7.2 ext. 9: consulebat, consulebatur A, consulabatur L; 7.3.6: conatus, conatum AL; 7.3.9: offudit, effudit AL; 7.3 ext. 2: Oebaris, o e baris AL; 7.3 ext. 4: vero urbis, urbis vero AL; 7.3 ext. 7: superior, superiori L c; 7.5.1: epuli, epulis A, aepulis L; 7.5.3: animi bonis, animis boni AL; 7.6.2: qui ne a, quin ea A, quinea L; 7.7 praef.: quo, qua AL; 7.7.4: geris, egeris AL; 7.7.5: movit, monuit AL; 7.7.6: quo, quod AL; 7.8.9: super quae, superque AL; 8.1 abs. 7: aeque, eque AL; 8.1 abs. 10: licere, liceret AL; 8.1 damn. 2: pependit, perpendit AL; 8.1. damn. 6: nocturnus, no tunus L c; 8.3.1: motusque, partesque A c, mortuusque L; 8.5.5: flagitium, meflagitium AL; 8.6.3: consulatus, coñs AL; 8.7.4: qui aetatis, qui et aetatis A c, qui etatis L; 8.7 ext. 2: mature, maturae AL; 8.8.2: quietae, quiestis A, quiete L; 8.11 praef.: patebit, patebunt AL; 8.13 praef.: spei, spe AL; 8.13 ext. 2: Isocratis, Hisocratis AL; 8.14 ext. 3: late, latae AL; 8.14. ext. 5: G omits in; 8.15.6: Crassus, Crassius AL; 8.15.8: Hispaniam, Hispania AL; 8.15 ext 4.: Berenices, Benetrices AL; 9.1.3: Gc has tenderet, tenderent AL; 9.2 ext. 9: Phalaridis, Phaleridis AL; 9.2 ext. 10: vivorum, virorum AL; 9.3.5: congratulationem, cum gratulationem AL; 9.3 ext. 1: Lysimachus, Lysymacus AL; 9.3 ext. 4: Semiramis, Samarramis A, Saemiramis L; 9.4.1: Minucius, Minutius AL; 9.5.2: despexit, dispexit AL; 9.7.1: et, A omits; 9.7.1: animorum, amorum A; 9.7.2: impudentia, imprudentia A; 9.7 mil. Rom. 1: ferat, fecerat A; 9.8.1: Syphacem, Sypacem A; 9.9.2: rediit, redit A; 9.10.1: ademerat, adimeret A; 9.10 ext. 1: utrem, utre A; 9.10 ext. 1: Berenice, Berenicae A; 9.10 ext. 1: satellitem … nomine, A omits; 9.11.2: Scaevola, Sevola; 9.11 ext. 4: torporem, torpore A; 9.12 praef.: evager, evagare A; 9.12.6: supplicio, A omits; 9.12.8: Etereius, A omits; 9.12 ext. 1: torqueant se, torquanse A; 9.12 ext. 6: anhelitu, hanelitu A; 9.12 ext. 8: faventem, favente A; 9.12 ext. 9: eam, eum A; 9.13.3: praecepit, praecipit AL; 9.13 ext 3: numine, numinum L; 9.13 ext. 4: Syracusanorum, Siraracusanorum L; 9. 13 ext. 4: longa fabula, longam fabulam L; 9.13 ext. 4: Aristomaches, Aristhomacles A, Aristhomaches L; 9.14 praef.: parvum, parum AL; 9.14 ext. 1: Laodice, Laudice L; 9.15 ext. 2: impendere, inpende AL.
How should Briscoe’s text be characterised? By my calculations Briscoe differs from Kempf’s 1888 text on 513 occasions (excluding proper names, place names and different orthographic conventions). This somewhat crudely indicates the scale of change Briscoe has thought necessary. I would characterise Briscoe’s text as conservative in that he rejects many of the conjectures printed by Kempf and prefers either the reading of the consensus of ALG or indicates a defective text by use of the obelus. I count 214 instances of conjectures or the readings of later mss. rejected in favour of the consensus of ALG, and 97 instances where Briscoe obelises but Kempf proffers some solution. This conservatism is seen also in relation to conjectures made after 1888, notably those of Watt and Shackleton Bailey.6 Their efforts are duly noted but never followed (at 6.8.1 Watt’s corpore is considered ‘fort. recte’).
Occasions where Briscoe prints his own conjectures are relatively infrequent and seem to relate particularly to place and proper names. The most interesting of these are: 1.6.2: septem for mss. VIII, to agree with Livy; 1.7.6: Coelius for mss. Caelius; 1.8.1: Vatienus for mss. Vatinius; 2.6.4: Areos for mss Areios; 3.2.23: Cassius for mss Caesius, following Suetonius and Plutarch; 3.7.10: Bariam for mss Badiam; 3.7.5: L. for mss. P.; 4.3.3: the omission of Augusto after fratri as a gloss; 4.7 ext. 1: et qui eam; for 5.1.1d: Misacenes for LG’s Musachanes; 5.1.7: illa for mss. Viri; 5.6.8: ne for mss. ne; 6.1 ext. 2: Ortiagontis for mss. Orgiangontis; 6.3.10: Gali for mss. Galli (cf. 8.1 abs. 2, 8.11.1); 7.6.1: a for mss. et a; 8.1 abs. 6: C. for L., following Syme; 8.3.2: Carfania for mss. C. Afrania; 9.8 ext. 1: Petelia for mss. Petilia.
Marking a return to the practice of Kempf’s 1854 edition, Briscoe cites testimonia for V.’s exempla. This makes his edition far more useful to the scholar. As Briscoe notes (p. xxx) this is an endless task, but I note only one possible oversight — at 1.1 ext. 6. V. takes his exemplum from Cicero Verr. 2.1.48.
The presentation of the text and apparatus criticus is excellent and the proof-reading of a very high standard. I have found only four errors.7
Although one might take issue with Briscoe’s decisions in individual instances,8 such dissension should not detract from the excellence of his work and from the debt that all future readers of Valerius owe him.
Notes
1. D. Schullian, ‘A neglected manuscript of Valerius Maximus’, Classical Philology 32 (1937), 349-59. Without consulting G I cannot determine whether Briscoe or Schullian is at fault when they disagree on readings of G: e.g. 1.5.6: Schullian reads cursum, Briscoe cursu; 1.7 ext. 7: somnium S, somnum B; 2.6.7: existimant S, existimans B; 4.7.2: potuerunt S, potuerant B; 5.2 ext. 4: reservarent S, reservare B; 6.8.1: etenim S, etiam B; 7.5.4: et S, ac B; 8.15.praef.: grate S, gratiae B. See also P.K. Marshall in L.D. Reynolds (ed.), Texts and Transmission (Oxford, 1983), 428-30. Briscoe gave a preliminary report of his conclusions in ‘Some notes on Valerius Maximus’, Sileno 19 (1993), esp. 395-98.
2. E.g. I. Faranda, Valerio Massimo: Detti e fatti memorabili (Turin, 1971), U. Blank-Sangmeister, Valerius Maximus: Facta et dicta memorabilia (Stuttgart, 1991).
3. R. Combès, Valère Maxime: Faits et dits mémorables. Livres I-III (Paris, 1995), Livres IV-VI (Paris, 1997). See the cool review by J. Briscoe, Classical Review 49 (1999), 76-9.
4. See J. Schnetz, Ein Kritiker des Valerius Maximus in 9 Jahrhundert (Neuberg, 1901).
5. Classical Review 49 (1999), 78.
6. W.S. Watt, ‘Notes on Valerius Maximus and Velleius Paterculus’, Klio 68 (1986), 466-78, and “Notes on Valerius Maximus’, Euphrosyne 23 (1995), 237-42; D.R. Shackleton Bailey, ‘Textual notes on lesser Latin Historians’, HSCPh 85 (1981), 158-67, and ‘On Valerius Maximus’, RFIC 124 (1996), 175-84.
7. P. xl, 1st item by Shackleton Bailey, for 678 read 67; p. 55, apparatus to line 29: for Kellenbauer read Kellerbauer; p. 598, apparatus to line 65 — there is a reference to the reading of L, but L is not extant for this portion of V.; p. 618, apparatus to line 23: for Trasumemum read Trasumennum.
8. H.F.-O. Müller ( Exempla tuenda: religion, virtue and politics in Valerius Maximus (Diss. UNC Chapel Hill 1994, p. 30 n. 47) defends the manuscript reading of alacritatis (1 praef.) which editors since Kempf (including Briscoe) have emended to claritatis in order to remove any idea of emotion from Roman imperial cult (cf. 8.15 praef. and 8.13 praef.).