BMCR 2018.01.45

C. Iuli Caesaris Commentariorum Libri III De Bello Ciuili; Studies on the Text of Caesar’s ‘Bellum civile’

, C. Iuli Caesaris Commentariorum Libri III De Bello Ciuili. Recognovit brevique adnotatione critica instruxit C. D.. Oxford; New York: Oxford University Press, 2015. cx, 227 pages. ISBN 9780199659746.
, Studies on the Text of Caesar's 'Bellum civile'. Oxford: Oxford University Press, 2015. vi, 329 pages. ISBN 9780198724063.

Preview Libri III de Bello Civili
Preview Studies on the Text

The publication by Cynthia Damon1 of a new critical edition (henceforth: OCT) of Caesar’s Bellum Civile (BC), accompanied by a volume of Studies (St.),2 is in itself important news, given that the last Oxford edition appeared in 19003 and the two reference editions are still those of Alfred Klotz (1926)4 and Pierre Fabre (1936),5 the deficiencies of which have been pointed out on numerous occasions.6

After these editions, Wolfgang Hering (1963)7 and Virginia Brown (1973)8—Damon’s edition is dedicated to her memory—made pertinent suggestions to clarify the manuscript tradition of the BC, but there was yet no new edition that was based on an exhaustive critical study of the work and that assessed their claims.

Damon’s edition and study, which appear amid renewed interest in the works of Caesar, are based on in-depth research that has focused on four aspects of the BC: 1) a new collation of the manuscripts that Damon considers most relevant for the establishment of the text; 2) an analysis of their filiation and her proposal of a stemma codicum; 3) the constitutio textus, accompanied by the corresponding critical tools (a critical apparatus, an Appendix critica, and an Appendix orthographica); and 4) a commentary on the passages Damon regards as the most problematic, which enhances our understanding of the text and of Damon’s editorial decisions.

In all these matters Damon does a rigorous job, not easy in view of the challenges facing a new edition of the BC, especially given the sneaking suspicion that the composition was originally unfinished and the noticeably deteriorated manuscript tradition, which often makes it a very complex task to determine the reading of the archetype. These difficulties are even more significant because the importance of Caesar’s works as one of the paradigms of classical language meant they were subject to normalization for literary or educational reasons, as is mentioned by Damon (St. 100).

The volume of Studies (St.) consists of three parts. In the Prolegomena (3-95), Damon justifies a new edition of the BC, addresses the history of the text (10-15) and its transmission (16-54, the history of the stemma, the analysis of the evidence of BHisp regarding β and ν, the investigation into ν, and Damon’s proposed stemma), analyzes the witnesses (55-95), offering a precise description of the characteristics of the archetype, of the hyparchetypes μ and ν, and of π and of the five mss. used for constituting the text (MUSTV), with close attention to the corrections and innovations documented in them. The second part (97-126) focuses on the passages where the transmitted text presents a range of problems (omissions, innovations, signs of incompleteness, the inaudita atque insolentia readings and their attribution to Caesar or to the vicissitudes of transmission). The third offers a detailed commentary on 74 critically problematic passages (19 from Book 1, 12 from Book 2, and 43 from Book 3).

The OCT edition starts with a Praefatio in English (XI-CVIII) summarising the aspects covered in more detail in St. : the history of the text, the stemma, the archetype, hyparchetypes, and manuscripts used. There follow the guidelines for using the edition, a bibliography, a conspectus editionum setting out Damon’s differences from previous editions,9 and a list of sigla. The critical edition (1-163) is capped by an Appendix orthographica (165-178), an Appendix critica (179-220), and an Index nominum (221-227).

When addressing the history of the text, Damon discusses the subscriptiones of the eight books of BG and upholds the usual interpretation of legi tantum of the subscriptio to BG 8, which suggests that “the ‘reading ceased’ at the end of the BG ” and that “Constantinus was aware that there was more to the corpus” (OCT XIII, St. 13).10

For the readings of the main manuscripts Damon relies on the collation of nine mss. (SLNMURTV) by Virginia Brown, whose codicological studies are mostly taken for granted by Damon, and on Damon’s first-hand reading of the five she considers relevant to determine the text (MUSTV), as well as of m and Vall. for the lost beginning of M.11

In general, the readings of these manuscripts are reported faithfully and exhaustively, especially compared to the incorrect readings in previous editions, particularly in those of Klotz and Fabre: e.g., Damon registers the correct reading of ms. M in 3.103.4 (despiceret μπ), and of V in 1.27.1 (eodem portauerant [-erunt ] mV). Great care is taken in distinguishing different levels of correction in some mss.: e.g., at 3.22.4 inspection confirms Damon’s sensible decision in attributing habebant to V and not to V2 (Klotz); at 2.20.7 Damon distinguishes cui erit Vac from the correction cui praeerat Vc, not recorded in Klotz or Fabre.

To decide which mss. have source value, Damon uses the eliminatio that was proposed convincingly by Hering and Brown for three manuscripts to which earlier editions had attributed some value: two derivatives of S (N, Neapolitanus IVc.11 and L, Lovaniensis, London, BL, Add. 10084), and a descendant of U (R, Riccardianus 541).12 The elimination of these mss., which have no stemmatic value, means that the origin of some conjectures is not identified accurately: for example, some plausible corrections are attributed to the editio princeps, although they already appear in these mss., such as dissimulari (2.31.5), present in R., or segniores (1.3.1), documented in N r.

Damon also accepts the eliminatio of about 162 recentiores that had been proposed by Brown. This decision is based on plausible evidence drawn from a partial collation of these witnesses, and it seems to me that, save for some valuable corrections, they are practically useless for editing the text. It is likely that the complete collation of these mss. is of interest primarily for better determining the origin of many conjectures, like 1.15.5 fugientem, attributed by Damon to Aldo, but already located by Fabre in the (now severely damaged) ms. Dresdensis 122 (Dc 167, 14th century).

As for the relationship among the principal mss. (St. 16-54), Damon undertakes a detailed stemmatic analysis in which she studies the possible relevant innovations and the correctable errors in order to clarify the connections among families and mss., attempting to determine the readings of the archetype. After a meticulous and well-founded analysis, Damon argues for a bipartite stemma (μ ν), close to that of Hering, in contrast to Brown’s tripartite one, and draws three main conclusions:

a) The ms. S does not represent an independent branch of the tradition, as Brown had believed. It derives from the hyparchetype ν independently of π (T and V). The conclusive argument that Damon puts forward is the absence of evidence of agreement between μ and π in significant innovation where S presents an archetypical reading, so she rightly rules out the existence of a “β branch” (μ ν) against S.

b) The ms. V is an independent copy of v based on a hyparchetype π (St. 66-69; 92-95), pace Hering, who considered V a descendant of T, but not a direct copy. V could be valuable to reconstruct the archetype when the other two ms. witnesses of ν (T and S) are independently in error and only V displays evidence of agreement with the μ family. In addition, V offers numerous successful innovations ( St. 94-95). Damon convincingly demonstrates V’s contribution to the reconstruction of ν, identifying some cases in which V is in agreement with S against MUT (cf. 3.19.3 and St. 69, n. 134).

c) Damon identifies signs of horizontal transmission through an “asystematic comparison” between μ and ν (specifically in a generation between ν and π) that would explain the lack of significant omissions between μ and ν and the major discrepancy between π and S, which would not have been affected by the contamination. This hypothesis is supported by a small group of readings where π and μ share an attractive but spurious innovation against S, which preserved a difficult or corrupt reading of the archetype (cf. St. 51-54). The argument is well founded, but we might ask if some of these readings of S really go back to the archetype or if they are innovations of this ms.: e.g. 1.76.5 terror ablatus S; 3.93.1 cursum S om. (this omission is not in the critical apparatus and is relegated to the Appendix critica, at OCT 215).

As regards the archetype (ω), Damon characterizes it as written in a pre-Carolingian or a Carolingian minuscule around the 8th century (St. 55; OCT XXII-XXVI), and reconstructs some of its features such as abbreviations, word division, inversions, variants and corrections, glosses, and book division. Damon forcefully defends the division into three books documented by MUTV, unlike S,13 whose text “is badly out of order” (OCT XXVI; St. 86).

The critical apparatus, following the criteria in OCT (LXIV-LXXII), comprises only the critically relevant readings of the mss. used for the constitutio textus, as well as later corrections Damon deemed important for that end, disregarding the inflated number of conjectures from the later tradition (cf. Meusel’s Tabula coniecturarum). After a reading recorded in the apparatus, Damon also often quotes parallel passages of Caesar or other Latin authors in order to offer more information to support the choice of variants, which is useful especially in the case of passages that receive no comment in St.

In addition to the critical apparatus, Damon records a large number of singular but not relevant readings of the main mss. in the Appendix critica (OCT 179-220, cf. LXXII), although the merely graphic variants are relegated to the Appendix orthographica (OCT LXXIII, 165-178), where she describes the normalization criteria that have been followed in the edition.14

As regards the constitutio textus, Damon’s suggestions are consistent with her stemmatic proposal and reflect an intelligent interpretation of the text in the choice of variants and the proposal of conjectures, offering a focused review of punctuation, with a reduced use of commas. Evidence of the contribution of Damon’s edition can be found by comparing it with most important editions of the 20th century, via the conspectus editionum (OCT LXXXVI-CVII), where Damon gathers the 683 passages in which her text differs from that of the editions of Kraner, Hofmann, and Meusel (here: KHM), Klotz, and Fabre (I will use “ edd. ” in shorthand for the three editions). Her text differs most often from the oldest edition, KHM (577 divergent readings), although Damon agrees with KHM on 106 occasions against Klotz and Fabre. Damon disagrees with Klotz’s Teubner edition around 368 times, and with that of Fabre she lists 327 discrepancies.15

These divergences show that the text edited by Damon is noticeably different from these previous editions. Let us see what trends dominate in this new edition:

1. Damon recovers a large number of readings of the archetype compared to her three reference editions (edd.), which prefer to accept innovations of the recentiores or corrections from editions or critics.16 The preference for the readings of ω is motivated by the stemmatic analysis and by their plausibility (OCT LXVIII n. 85). Damon’s proposal is consistent for the main part.

One instance that might be debatable is 1.51.1, where Damon accepts comitatus (ω), instead of the conjecture commeatus (Nipperdey edd.), in reference to a convoy of troops described in detail in the passage; in 1.48.4 Damon correctly assumes, with edd., Beroaldus’s correction, commeatus, (“extensive supplies,” LCL 73) instead of comitatus (ω), and keeps the reading of the archetype in 1.54.5 ( commeatus ω: comitatus Manutius, “the supplies,” LCL 81). In the case of 1.51.1, one might object that comitatus is not documented in BG and only twice in BC, but with the specific sense of “company” or “entourage” (BC 3.61.1 magno comitatu, “company”, LCL 279, cf. H. Merguet, Lexicon zu den Schriften Cäsars, Jena 1886, Hildesheim 1963, 184, “Gefolge”); 3.96.4 comitatu equitum XXX, in reference to Pompey’s entourage (“escort” LCL 331). commeatus, in addition to its usual sense of “provisions” or “supplies,” can also refer to an expedition or convoy of soldiers (cf. BG 5.23.2).

2. In Damon’s selection of readings of a branch of the tradition or of one manuscript or several in comparison to Klotz and Fabre:

a) Damon tends to adopt μ readings more often (3.51.7 tormentumue μ Damon : –to SV edd.), and to a lesser extent ν (3.75.1 haec ν Damon : ac μ edd.), in line with the stemmatic analysis that shows that ν presents more erroneous readings than μ ( St. 64-65).

b) In the selection of variants based on the consensus among mss. from different families, Damon shows preference for μS readings instead of π,17 or the consensus of a μ ms. with ν or some of its mss. (1.24.3 procul Uν Damon: quae procul m edd.). 18 c) As regards ms. V, Damon tends to reduce its weight to a certain extent in comparison with edd., as is shown by the fact that she does not adopt any of the singular readings of V that have been accepted by edd.,19 while she only follows V in some cases (3.84.5 Egum V Damon : unum S Klotz : Aecum Fabre).

d) As for the ms. S, which has confounded scholars to such an extent, Damon diminishes its weight for the constitutio textus, in line with its position in the stemma,20 as is shown by the large number of examples in which Damon follows μπ instead of S edd. This, in our view, is one of Damon’s main achievements, to adopt more subtle and less expected readings than those of S, accepted by the three edd. 21

3. Damon maintains a balanced position in relation to the conjectures of editors and other scholars: she judiciously rejects a large number of these corrections in favor of the reading of the manuscript tradition, although she allows some of the conjectures accepted by a majority of the Caesarian tradition and in some cases adopts more than edd.:22

For example, 2.10.4 she follows Nipperdey’s conjecture superstruantur (St. 184-188), citing Sen. Con. 1, pr. 21. The proposal is an intriguing one although this verb is not documented in Caesar’s time; the other plausible correction, super musculo struantur (Manutius, edd.), offers the expected singular against super musculos struantur ω, but it also raises the question of congruence with the language of the period, given that super is documented in Caesar with the accusative (2.10.6 super lateres), but the ablative starts being documented only in Augustan poetry.

It is of note how Damon handles the early printed tradition, including the editio princeps (1469) and the first Aldine edition (1513), which offer many corrections, to which Damon takes a measured approach: as regards the princeps, Damon accepts some of the corrections that have been accepted by the edd.,23 but disregards many others. 24 However, one should note the plausibility of her choice in the small group of innovations of ed. pr. that she follows versus ω, Klotz, and Fabre.25

Something similar can be said of the Aldine edition, whose intense conjectural practice is assessed sensibly by Damon, who rejects additions that have been accepted by edd. 26 and accepts some others.27

Some of the relevant corrections that Damon attributes to Stephanus (Paris 1544) actually come from the corrections included at the beginning of Aldus’ edition of 151328, reproduced in the second Aldine edition (1519): 1.61.5 miliaque; 1.82.1 educunt; 1.85.8 praesideat; 1.87.1 restituatur. The same occurs with a correction attributed to Manutius (Venice 1571, 1597): 3.28.4 hic, which is documented in Aldus’ corrections 1513 and in 1519.

4. Another major aspect of the changes of this edition has to do with Damon’s own conjectures, the result of intelligent reflection on the text, looking in general for parallel passages in Caesar’s work and in other classical texts. Many of them are explained in St., others are simply proposed in the text.

An example is 3.73.5 se notum.29 This correction should be explained: Damon adduces 3.66.2, given the unsatisfactory reading of the mss. (secum μST: se eis V) and of the conjecture se aequum (Victorius Fabre).

Some of the proposed cruces are appropriate (1-35.4 † Gallias †, as against the conjecture of Glandorp, Sallyas, common in the previous editions), but in other cases it might be preferable to accept some corrections and, in fact, Damon has put many of these into the new text of LCL (E.g.: 1.3.3 † ius † : ipsum Hug; 3.11.1 † copiis † : oppidis Lipsius; 3.49.3 † ad specus † : ut specus Menge).

This excellent edition makes serious contributions to the reconstruction of the text, and its careful and deep reading of the text of BC and the close study of its textural tradition is accompanied by an insightful commentary on troublesome passages that brings to light the enormous complexity of a text that has been transmitted in such a deficient way, offering suggestive new proposals that will encourage reflection on the reading and interpretation of the work of Caesar. ​


1. Thanks are due to the Spanish Ministry of Economics and Competitivity for their financial support through the project FFI2015-67335-P. I am grateful to Daniel Kiss for his thoughtful reading of this review.

2. As well as a volume in the Loeb Classical Library, Caesar: Civil War. Edited and translated by Cynthia Damon. Cambridge Mass., London; Harvard University Press: 2016 (LCL).

3. Renatus Du Pontet (ed.), Caesar: Commentarii. 2. Bellum Civile, cum libris incertorum auctorum de Bello Alexandrino, Africo, Hispaniensi. Oxford; OUP: 1900 (1922).

4. Alfred Klotz (ed.), C.I. Caesaris commentarii, vol. II: Commentarii belli civilis. Leipzig; Teubner: 1926 (1950 2)

5. Pierre Fabre (ed.), César, Guerre civile Paris ; Les Belles Lettres: 1936 (revised ed. A. Balland, 2006).

6. Virginia Brown, The Textual Transmission of Caesar’s Civil War. Leiden; Brill: 1972, 9; Michael Winterbottom, in Leighton D. Reynolds (ed.), Texts and Transmission. A Survey of the Latin Classics. Oxford; Clarendon Press: 1983, 35 n. 1.

7. Wolfgang Hering, Die Recensio der Caesarhandschriften. Berlin; Akademie-Verlag: 1963, as well as the considerations on the mss. of the BC in his edition of the Bellum Gallicum (BG), C.I. Caesaris Commentarium rerum gestarum. Bellum Gallicum. Leipzig; Teubner: 1987.

8. Brown, The Textual Transmission (see n. 6).

9. Friedrich Kraner, Friedrich Hofmann, Heinrich Meusel, C. Iulii Caesaris Commentarii de bello civili. Berlin; Weidmann: 1906 (reprinted in 1959 with supplements by H. Oppermann).

10. There is no mention of Cameron’s interpretation of tantum in the sense that Constantinus read only his copy without comparing it to another, cf. Alan Cameron, The Last Pagans of Rome. Oxford, New York; OUP: 2011, 460-461 n. 18.

11. Stephen Oakley has recently rightly demonstrated that Vall. is a direct copy of m in his review of D.’s work (CR 67, 2017, doi:10.1017/S0009840X17000452).

12. Damon includes under the initial ϛ the readings of these mss., although it might have been advisable to have pointed out in the preface or in St. the plausible innovations attributed to them. E.g.: 1.58.4 comminus L; 1.67.1 sentiretur L; 2.29.4 nonnulla L; 3.2.2 equites L; 3.6.3 postridie S2 L; 3.27.1 recipiebat LNr; 3.51.6 magnam L.

13. This question has recently aroused the interest of Luca Grillo, The Art of Caesar’s Bellum Civile. Literature, Ideology, and Community. Cambridge; CUP: 2012, 181-184.

14. It would have been advisable to indicate the origin of the reading when discussing spelling variants adopted which are not documented in the main mss., as in the case of Boeoti- / boeti- (OCT 169): Boeotia (3.4.2) Aldus 1513 (Boetia ω), not included in the critical apparatus or in the Appendix critica.

15. However, the text edited later in LCL (XLIII-XLV) also differs from the OCT text in 65 places.

16. Cf. OCT LXVIII, n. 85. E.g. : 2.18.1 Pompeioque ω Damon: Petrei- ed. pr. edd.; 3.8.3 diligentiae ω Damon: indiligentiae Stephanus edd.; 3.18.3 ubi primum rursus ω Damon: ubi primum e re uisum est Elberling edd.; 3.97.3 spe ω Damon: re ed. pr. edd.; 3.112.8 reliqua ω Damon: regia Morus Klotz Fabre.

17. E.g.: 3.93.5 ex cohortium numero μS Damon : ex cohortibus π. However, I have not located any instances in which she follows π when not accepted by Klotz and Fabre. However, in all the cases in which Damon adopts a reading of the hyparchetype π against μ, Klotz and Fabre have already accepted the reading of π: e.g. 1.32.7 hortatur; 1.54.2 ex leui; 1.56.4 hae; 1.58.3 armamentorum (cf. OCT XXXI-XXXII).

18. Some of these proposals have been reconsidered in Damon’s own translation for Loeb (LCL XLII-XLV).

19. E.g.: 3.55.4 amicitia Caesaris μST Damon; 3.67.5 e loco μST Damon; 3.87.4 ex colonis μST Damon.

20. Damon adopts in OCT 30 readings that are transmitted only by S (cf. the list in St. 90-91), which have also been accepted by Klotz and Fabre. Other readings that could be added to Damon’s list: 1.15.1 progressus S Damon; and 1.48.5 ac ciuitates S Damon (“one of its canny innovations,” St. 157).

21. E.g.: 1.34.5 [ in ] omnibus ϛ Damon: ex omnibus S edd.; 1.61.6 muniuntur μπ Damon: muniunt S edd.; 1.85.4 hominum μπ Damon: hominibus S edd.; 3.19.31 ut inter μπ Damon: inter S edd.; 3.62.2 aberant μπ Damon: aberat S edd.; 3.68.3 coniuncta μπ Damon: coniunctam S edd.

22. E.g.: 1.6.4 consul Madvig; 3.9.5 quare missis Brutus; 3.19.2 tuto Vossius; 3.53.5 uestiariis Nicasius; 3.58.5 frus Buecheler; 3.63.3 [munitiones] Ciacconius; 3.85.2 <ille> Meusel; 3.95.1 dari Vascosanus; 3.102.6 arcem captam esse Oudendorp.

23. E.g.: 1.14.4 sese; 1.28.3 notisque; 1.35.3 <Romanum>; 1.41.3 <per> Afranium; 1.41.5 hos; 1.82.5 resistere; 2.19.4 Cordubam; 2.20.4 Hispalim; 3.45.4 receptus; 3.101.2 Pomponianam; 3.102.6 <civitates>.

24. E.g.: 2.9.2 eaque ω Damon: easque ed. pr., edd.; 2.18.1 Pompeioque ω Damon: Petrei- ed. pr., edd.; 2.27.2 conspectu ω Damon: conspectum ed. pr., edd.; 3.46.5 obtecti ω Damon: obiecti ed. pr., edd.; 3.97.3 spe ω Damon: re ed. pr., edd.

25. 2.25.1 [a] theatro ed. pr. Damon: a theatro V Klotz Fabre: atheatro μST; 3.15.1 <cum> classe ed. pr. Damon: in classe ϛ (LN) Klotz Fabre; 3.84.3 electos milites ed. pr. Damon: electis milites ω Fabre: electis [milites] Nipperdey Klotz.

26. 3.42.5 prouidebat ω Damon: providerat Aldus edd.

27. 1.54.2 primum ac Aldus Damon; 1.65.4 intra montes se recipiebant Aldus Damon.

28. Errata, quae uel inter impressionem contigerunt, uel impresso uolumine deprehendimus, Venice 1513. Cf. A. Moreno, Epos 26, 2010, 33-50.

29. “I gave you a familiar position for the fight. . . ” (LCL 297). ​