Bellum Alexandrinum

2.5

Dallas Simons and Cynthia Damon

Praeterea alias ambulatorias totidem tabulatorum confixerant subiectisque eas rotis funibus iumentisque obiectis derectis plateis in quamcumque erat uisum partem mouebant.°

“In addition, they had cobbled together moveable towers with the same number of stories (i.e., ten). After mounting these on rollers, with ropes and pack animals placed opposite, they would move them on level streets into whichever part (sc. of the city) they wanted.”

This sentence describes the last and most impressive of the defensive measures adopted by the Alexandrians against the newly arrived Caesar, namely, the construction of moveable siege towers. Obiectis is the principal problem in this passage: its authenticity has long been questioned, with Scaliger excising it and other critics offering emendations. Confixerant, too, has been suspected.

2) Obiectis, which occurs later in the description of these moveable military towers, makes little sense in context: if pack animals are pulling the towers, it is hard to see the relevance of their having been “placed opposite,” “in front of,” or “in the way of” the towers. The parallelism between funibus and iumentis is also peculiar.

Castiglioni emended obiectis to adiectis, assuming a palaeographically plausible innovation. This gives the sense of “fastened upon” or “added to.” Forms of adicere occur ten times in the corpus, but it usually means to add on or adjoin something abstract or to describe geographical space (cf., e.g., 28.3 unum latus [sc. castrorum] erat adiectum flumini Nilo). Here it would imply that the animals are added on to a larger contraption, which does not yield perfect sense.

Klotz proposed obtectis to describe the animals, which would be “covered” or “protected.” Caesar uses obtectus to refer to military protection, which would suggest that the pack animals are shielded from opposing attack (cf. BC 3.19.7 quae [sc. tela] ille obtectus armis militum uitauit and 3.54.1 uineis eam partem castrorum obtexit). The context of our sentence does not necessitate this idea, and the necessity of taking iumentis in tandem with funibus makes it unlikely.

Contexts involving pack animals towing something frequently contain the verb iungere. There are two potential problems with an emendation based on iungere, however. First, iunctis is paleographically further removed from obiectis than either adiectis or obtectis. Second, iungere usually refers to joining animals to one another rather than to other objects. However, the verb does occur with the etymologically related iumentum (Nepos Timol. 4.2, Varro RR 1.52.1; cf. Var. LL 5.136) and in one passage of Vitruvius a yoked team of oxen pull a rope (Vitr. 10.2.14 bubus iunctis funem ducebant [sc. boues]). So iunctis is worth consideration here, even if funibus is a loose end.

Compounds of iunctis are also worth considering. Cornelissen suggested subiunctis, making the ropes and pack animals “subjoined” to the towers. He cites as parallels a line from Columella in which bullocks are attached to a plow (Col. 6.2.8 uacuo plostro subiungendi) and a line from Virgil in which tigers are joined to a chariot (Virg. Ecl. 5.29 curru subiungere tigres). Neither passage offers a parallel for funibus, which sits oddly with the prefix sub-, and the verb itself appears nowhere else in the corpus. The form adiunctis, from a verb used eight times in the Bellum Alexandrinum and twenty-five times in the corpus overall, is perhaps a better repair. Aulus Gellius uses adiunctis absolutely with iumenta (Gel. 20.1.28 uectaculum … quod adiunctis pecoribus trahebatur), and Horace uses it jokingly to describe yoking mice to a cart (Hor. Sat. 2.3.247 plostello adiungere mures). However, nowhere in the corpus does adiungere describe affixing something to an object.

It therefore seems best to leave obiectis in its imperfection with the understanding that the syntax and general sense are clear, even if the original verb is not.

7.2–3

Cynthia Damon

Erat autem magna multitudo oppidanorum in parte Caesaris, quam domiciliis ipsorum non mouerat quod ea se fidelem palam nostris esse simulabat et desciuisse a suis uidebatur, ut mihi defendendi essent Alexandrini—neque fallaces esse neque temerarii—multaque oratio frustra absumeretur. (3) Cum uero uno tempore et natio eorum et natura cognoscatur aptissimum esse hoc genus ad proditionem dubitare nemo potest.°

“Moreover, there was a large crowd of townspeople in Caesar’s district, whom he had not removed from their homes because they were making an open show of being loyal to our men and seemed to have cut their ties with their own side, so that I would have to defend the Alexandrians, saying that they are neither deceitful nor impetuous, and much verbiage would be spent to no purpose. (3) But when their community and character are considered simultaneously, no one can doubt that this race is perfectly prepared for treachery.”

In the passage quoted above the string ut … absumeretur is excised by Gruterus (along with the following sentence), emended by many, and daggered by Andrieu. The general problem is the loose logical connection between the Alexandrians’ apparent loyalty to the Romans (uidebatur) and the idea that they require a defense from the author (ut mihi defendendi essent). What we want to be told instead, as Madvig observed (1873, 281), is why the Roman fears expressed in 7.1 (alii … extimescerent … quod neque celari Alexandrini possent in apparanda fuga, … neque illis imminentibus atque insequentibus ullus in naues receptus daretur) were justified. A further problem is that the hypothetical defense is contradicted by later critiques centered on Alexandrian deceptiveness (24.1 Caesar etsi fallacem gentem semperque alia cogitantem, alia simulantem bene cognitam habebat, 24.3 regius animus disciplinis fallacissimis eruditus, ne a gentis suae moribus degeneraret; cf. also 24.6 fallaciis pueri).

Editors who accept the basic authenticity of the transmitted text have to contend with a series of problems within the string itself, the most significant of which lie in the connectives. As transmitted the relevant part of the passage reads as follows:

The connectives string together three dissimilar propositions: the opening reference to the necessity of a defense stands in parallel to neque fallaces essent neque temerarii, which looks like part of the hypothetical defense and therefore seems to warrant subordination, and also to a declaration that defense would be a waste of breath, where the apparently adversative sense needs more than a linking -que. The following sentence compounds the problem by abandoning the idea of defense and instead indicting the Alexandrians for their aptitude for treachery. (Here, however, we do get a nice robust uero.)

Gruterus’ excision is a drastic response to this assemblage of problems of sense and usage. They look less unacceptable if considered one by one.

The first of these innovations creates the subordination desiderated above, the second improves the logic somewhat but leaves multa … absumerent rather isolated. Neither tum nor absumerent is essential to the repair and they were quickly dropped when better manuscript evidence became available, as in the edition of Vascosanus, who also retained -que and repunctuated:

… desciuisse a suis uidebatur: at mihi, [si] defendendi essent Alexandrini …, multaque oratio frustra absumeretur.°

The second ellipsis is present in Madvig’s article, so it is not clear whether or not he accepts Nipperdey’s esse. Presumably so. See further Gaertner/Hausburg 2013, 68 n. 185.

Nipperdey addressed the quod-clause problem by substituting esse for essent2:

… desciuisse a suis uidebatur: ut mihi defendendi essent Alexandrini neque fallaces esse neque temerarii, multaque oratio frustra absumeretur.°

Klotz merges Aldus’ conditional with Nipperdey’s indirect statement: … desciuisse a suis uidebatur: ut mihi, [si] defendendi essent Alexandrini neque fallaces esse neque temerarii, multaque oratio frustra absumeretur. But the genesis of the three-fold innovation is hard to explain.

On balance it seems best to accept this minimally emended text. The logic is murky and the language awkward indeed, but the very murkiness of the logic is, as Madvig saw, an argument against ut … absumeretur (let alone ut … potest) being an interpolation.

12.1–2

Tom Vozar and Cynthia Damon

Eo detrimento adeo sunt fracti Alexandrini, cum iam non uirtute propugnatorum, sed scientia classiariorum se uictos uiderent, quibus et […] superioribus locis subleuabantur, ut ex aedificiis defendi possent et materiam cunctam obicerent, quod nostrae classis oppugnationem etiam ad terram uerebantur. (2) Idem, posteaquam Ganymedes in concilio confirmauit sese et eas quae essent amissae restituturum et numerum adaucturum, magna spe et fiducia ueteres reficere naues accuratiusque huic rei studere atque inseruire instituerunt.°

“The Alexandrians were so crushed by this defeat, since they now saw themselves conquered not by the courage of the fighters but by the skill of the mariners, for whom both […] in/from (?) the higher locations they were being supported, so that (?) they could be defended from the buildings and interposed all the wood as a barrier, because they feared an attack by our fleet even on land. (2) The very same people, after Ganymedes assured them in council that he would both restore those (sc. ships) that had been lost and increase their number, began with great hope and confidence to repair the old ships and quite diligently to apply and devote themselves to this matter.”

The first sentence of this passage, as transmitted, has a number of glaring problems. The most substantial is the incompatibility of the main clause, which is about a loss of morale (adeo sunt fracti Alexandrini), and the apparent result clause, which is about defensive capability (ut ex aedificiis defendi possent). The antecedent of quibus is another puzzle: does the relative pronoun refer to Roman classiarii, Alexandrians, or something that can function in tandem with locis? And then there is the difficulty of making sense of materiam cunctam in light of the earlier assertion that Alexandrian buildings contained no materia (1.3 [ab] incendio fere tuta est Alexandria, quod sine contignatione ac materia sunt aedificia et structuris ac fornicibus continentur tectaque sunt rudere aut pauimentis). These problems are interconnected, which complicates the task of seeing and explaining solutions, but the context, especially the subsequent sentence (quoted above), provides something of a framework to use in approaching them.

The passage we are concerned with reports the beginning of the aftermath of an unexpected naval defeat suffered by the Alexandrians, who had hoped to catch Caesar in a moment of vulnerability (10.3 magnam sibi facultatem fortunam obtulisse bene gerendae rei crediderunt) but barely avoided losing their entire fleet (11.5 Quod nisi nox proelium diremisset, tota classe hostium Caesar potitus esset). In the sentence immediately before our passage they return to Alexandria terrified (11.6 perterritis). The immediately following sentence, introduced with an emphatic idem, shows the resilience of their military commander, Ganymedes (see 4.2), and the revival of hope and confidence in the Alexandrians themselves. Details follow in 12.3–13.4. This is the context into which our problematic sentence needs to fit.

It will be helpful to begin by reviewing the (relatively) secure ground in our sentence. As was mentioned above, it begins with the Alexandrian loss of morale (adeo sunt fracti Alexandrini). It ends with a causal clause concerning their fears about a possible fleet-based attack on the city (quod nostrae classis oppugnationem etiam ad terram uerebantur). Verebantur might be a restatement of fracti, but it might also show that the Alexandrians have overcome their panic and started taking sensible precautions. In the latter scenario, our sentence mirrors the turnaround reported in 12.2. In the former scenario, our sentence stands in contrast to 12.2. The emphatic idem is compatible with either scenario (parallel: BG 6.13.5, BAfr 22.2, 28.2; contrasting: BAlex 18.3). According to the cum-clause, the morale problem was particularly acute because the Alexandrians had been defeated in the arena where they felt most confident, the sea, and now needed to defend themselves on land, a situation that seems to underlie both aedificiis defendi possent and materiam … obicerent.

Early editors left the spot alone, presumably taking the antecedent for quibus to be uirtus and scientia and quibus itself as an instrumental ablative parallel to locis (cf. BG 1.40.6 quos … usus ac disciplina … subleuarent). The lack of materia in Alexandria is perhaps made less worrisome by a later passage about roof beams (13.2 Porticus gymnasia publica aedificia detegebant, asseres remorum usum obtinebant) that encourages taking 1.3 sine contignatione ac materia as a hendiadys, “without wooden decking.” But the larger sense problem eventually attracted attention and a variety of approaches.

Manutius proposed a simple solution to the aforementioned incompatibility, replacing ex with the negation uix: the Alexandrians are so dispirited by the naval disaster that they can scarcely feel secure in the city itself. It is easy enough to supposed that uix was misread as ex before aedificiis, and a similar scene plays out later on Pharos Island with an explicit reference to the situation in Alexandria: 18.1 Neque uero diutius ea munitione se continere potuerunt etsi erat non dissimile atque Alexandriae genus aedificiorum, ut minora maioribus conferantur, turresque editae et coniunctae muri locum obtinebant.

But neither Manutius’ repair, nor Dinter's subsequent suggestion that uix fell out before ex, sits at all well with cunctam. And yet the two subjunctive expressions seem to be aligned in content as well as form.

Davies postulated a lacuna in the text before quibus. But therafter he rather throws up his hands, “cum de locis tam mutilis nihil certo possit definiri.” Nipperdey placed the lacuna after quibus, a position preferred by subsequent editors since it allows one to avoid specifying the identity of the antecedent: it could be the Romans or the Alexandrians, or for that matter uirtus and scientia. Both lacunas leave a space in which the meaning of materia cuncta might have been made clear. But we will argue below that the expression locis subleuabantur is problematic.

First, however, another approach. Dinter rewrote the passage:

13.5

Tom Vozar and Cynthia Damon

Caesar Rhodias naues VIIII habebat—nam X missis una in cursu litore Aegyptio defecerat—Ponticas VIII, Lycias V, ex Asia XII.°

“Caesar had nine Rhodian ships—for although ten had been sent one had foundered off the Egyptian coast en route—eight Pontic ships, five Lycian, twelve from Asia.”

The text enumerates the naval forces that Caesar had under his command at the beginning of the battle in the Eunostos harbor of Alexandria in 47 BCE. The paradosis has been questioned on both philological and historical grounds.

1. The awkward language of the parenthesis introduced by nam has drawn the most critical attention. Larsen, feeling that the ablative absolute decem missis is intolerably harsh, supplies de, producing a well paralleled construction (u. TLL 5.1.58.74–59.28 ). But Incertus uses a similar expression at 15.5 progressis ultra uadum IIII, so emendation does not seem necessary.°
   If that is how the opening of 15.5 should be construed. See the apparatus notes ad loc.
   The string in cursu litore Aegyptio has likewise been queried. Nipperdey asserted that in cursu and litore were contradictory — “cursus enim in littore Aegyptio erat confectus” (1847, 190) — and therefore excised litore as a scribal gloss on the unusual expression in cursu Aegyptio. It is indeed unusual: the common expression in cursu never takes an adjectival toponym. Such adjectives are, however, used with litus (cf., e.g., Cic. Sest. 140 in litore Dyrrachino). Larsen proposed adding sub before litore, comparing BG 5.57.3 sub castris eius uagabatur. That after cursu the preposition sub might be overlooked is obviously possible, but the absence of the expression sub litore elsewhere weighs against this emendation. Furthermore, the paradosis is paralleled at BG 4.23.6 aperto ac plano litore naues constituit, where the bare ablative has a locative function and, as it appears from the context, in which Caesar’s ships cannot land on the beaches, litore does not signify the shore itself but the waters just offshore (for this sense of the word see TLL s.u. litus 7.2.1537.52–1538.10 “respicitur litoris pars uda, tam area undis et aestu affecta quam mare terrae uicinum”).°
   Are we perhaps to understand that the Rhodian ship ran aground upon the treacherous uada Aegyptia mentioned at Luc. 8.539–40 (perfida qua tellus Casiis excurrit harenis/ et uada testantur iunctas Aegyptia Syrtes)?
   The peculiarity of deficere for a ship “foundering” is highlighted by the fact that its nearest parallel is Verg. Aen. 6.354, where Palinurus tells Aeneas he fears less for himself than that tua … / deficeret tantis nauis surgentibus undis.°
   R. G. Austin, 1986, P. Vergili Maronis Aeneidos Liber Sextus, Oxford, ad loc. cites 13.5 as a parallel for the Vergil passage.
   An additional problem with defecerat is that in a civil war context it could easily be construed as a reference to defection (see TLL 5.1.327.79–328.7), something perhaps better passed over in silence in an account such as ours. So different verbs have been proposed for the spot. Ciacconius’ suggestion of decesserat, based on BC 3.112.3 naues … suo cursu decesserunt, is rebutted by Davies on the grounds that the senses required in the two passages are different (“deviate” in the latter, “perish” in the former); it is also harder to take litore as a locative ablative with this verb. Siesbye proposed desederat, from a verb used of sinking ships at Sen. Nat. 6.6.4 nauigia desidunt. However, in defense of the paradosis one can cite BC 3.2.3 multi galli tot bellis defecerant, “many men had succumbed (or defected) in the numerous wars,” which shares with our passage both the ablative and the possibility of misconstrual.
2. There is also a problem of some historical consequence here. Scholars were content with the transmitted reading l(i/y)(c/t)ias = Lycias until Schneider proposed the emendation Cilicias, with the addition of some unknown number after Syrias, citing 1.1 as evidence: Caesar Rhodo atque ex Syria Ciliciaque omnem classem arcessit. Schneider’s conjecture has since been adopted by Kübler, Klotz, and Andrieu, among others.

In sum, it appears that the basis of Schneider’s conjecture is unreliable at best, and that the paradosis, far from being historically objectionable, may actually illuminate the condition of Romano-Lycian relations in the time of Caesar.

17.1–3

Cynthia Damon

Hoc ne sibi saepius accidere posset, omni ratione Caesar contendendum existimauit ut insulam molemque ad insulam pertinentem in suam redigeret potestatem. (2) Perfectis enim magna ex parte munitionibus in oppido, [et hanc] et illam [urbe(m)] uno tempore temptari posse confidebat. (3) Quo capto consilio cohortes X et leuis armaturae electos quosque idoneos ex equitibus Gallis arbitrabatur in nauigia minora scaphasque imponit. In alteram insulae partem [[…]] distinendae manus causa constratis nauibus aggreditur praemiis magnis propositis qui primus insulam cepisset.°

“So that this could not happen to him repeatedly, Caesar thought that he ought to strive to bring the island and the jetty extending to the island into his power, whatever the cost. (2) For since he had mostly completed his works in the town, he was confident that an attempt could be made on [both the latter] and the former at the same time. (3) Having adopted this plan, he placed in small vessels and skiffs ten cohorts and the picked men belonging to his light-armed force and whichever of the Gallic horsemen he deemed suitable. On one side of the island [he landed his footsoldiers, on the other side] he attacked with decked ships in order to divide the (enemy) forces, with great rewards offered to whoever seized the island first.”

Aldus proposed a relatively simple solution to the problem(s) of illa and provided the desired doubling:

Another approach would be to assume that the missing parallel for et illa(m) was lost in an eye-skip from et to et. The pronoun’s obscurity would be clarified if what was lost was hac or hanc. Incertus’ usage elsewhere suggests that et ha(n)c preceded et illa(m) (cf. 6.3 haec propior … illa inferior, 13.1 has … illas, 19.2 hunc … illum). As between the accusative and the ablative, the accusative seems preferable given the aforementioned objections to seeing urbem as a subject for temptari. This leaves the problem of urbe(m): where did it come from? Perhaps it was spawned when the final m of illam was misread as in (as was mentioned above), or perhaps — after the loss of et hanc — urbem was added as a gloss for the now lonely illam based on an antithesis with in oppido.

With these repairs the sentence reads:

As the narrative continues, Caesar attacks the island (17–18) then the jetty (19–21). This does not square perfectly with uno tempore, but the misfit is at least partly a result of the gaps in the narrative in the following sentence and at 17.6.

The second gap pertains to the two-pronged attack itself. In the passage as transmitted we hear that Caesar, hoping to split the enemy force, sent decked ships to attack one part of the island, and that he encouraged the men in his own divided force to compete with one another for the rewards that would be given to whichever group first took control of the island:

The construction in … insulae partem … aggreditur is problematic — although for a similar anacoluthon see 26.2 — but the bigger problem here is the absence of any reference to the other part of the island, presumably the part that the large force of men (ten cohorts plus light infantry and select cavalry, presumably on foot) embarked on the smaller vessels and skiffs was to attack. Since this area is the focus of the main battle narrative starting at 17.5, its omission here is odd.

Aldus repaired the smaller problem by excising the unwanted preposition, Oudendorp by converting it into an adverb, inde. Vascosanus, who accepted Aldus’ emendation, addressed the larger problem by adding a connective before alteram and repunctuating so as to join this sentence to its predecessor:

… cohortes X … in nauigia minora scaphasque imponit: in alteram[que] insulae partem distinendae manus causa constratis nauibus aggreditur praemiis magnis propositis qui primus insulam cepisset.°

Vascosanus’ repair presupposes the loss of -que, which may be related to the appearance of a surplus -que later in the sentence: praemiis ed. pr. : praemiisque MUSTV.

Castiglioni combined the two approaches, replacing in with et:

Another approach is to assume another eye-skip here resulting in a lacuna that one might fill thus:

This supplement makes use of the transmitted preposition in and provides a second alteram, thereby specifying the attack’s two approaches, which the enemies withstand pariter in 17.4 (ac primo impetum nostrorum pariter sustinuerunt). Its language has acceptable parallels, too, since Incertus uses the expression copiae pedestres twice, once in an antithesis with naval forces (44.1 partim classe per se, partim pedestribus copiis per barbaros, cf. also 64.1), and mentions troop landings twice (10.2 remiges in terram … exposuisset, 19.3 cohortium instar in terram exposuerat). But at least two aspects of the supplement give pause. First, the hypothesized skip goes from partem to partem, yet the second partem is unnecessary and probably unwanted: in comparable constructions Incertus omits the second substantive (1.5 alterius rei … alterius, 11.4 una hostium triremis … altera, 19.5 altero opere effecto … alter instituto, 28.3 unum latus … alterum). However, the pleonasm is paralleled at BG 1.2.3 una ex parte … altera ex parte (cf. also BG 4.3.3). Second, one misses the connective provided by earlier repairs. So the supplement can only be offered exempli gratia.

On balance, however, it seems better to indicate a gap than to apply a superficial repair.

22.1–2

Cynthia Damon

Hoc detrimento milites nostri tantum afuerunt ut perturbarentur ut incensi atque incitati magnas accessiones fecerint in operibus hostium expugnandis. (2) In proeliis cotidianis, quandocumque fors obtulerat, procurrentibus et erumpentibus Alexandrinis, manum […] comprehendi multum operibus et ardentibus studiis militum. Nec diuulgata Caesaris hortatio subsequi legionum aut laborem aut pugnandi poterat cupiditatem, ut magis deterrendi et continendi a periculosissimis essent dimicationibus quam incitandi ad pugnandum.°

“Our men were so far from being thrown into confusion by this defeat that, roused and provoked, they made extra efforts in attacking the enemy’s fortifications. (2) In the daily battles, whenever a chance had come their way, with the Alexandrians running forward and making sorties, […] combat […] to be checked greatly by fortifications and the burning eagerness of the soldiers. Nor was Caesar’s widespread exhortation able to attain the level of the legions’ effort or desire for fighting, so that they had to be deterred and restrained from exceptionally dangerous forms of combat rather than worked up for fighting.” (The two lacunas in the translation reflect a single lacuna in the Latin that contains at a minimum the main clauses of both the sentence to which “combat” belongs and the sentence that ends “by fortifications and the burning eagerness of the soldiers.”)

The personal use of afuerunt and the shift from secondary to primary sequence in the first sentence prompted Vielhaber (1869, 558–60) to replace the main verb with an impersonal afuit and deal with fecerint and the subsequent problem spots by excising incensi … ut. His explanation for the genesis of this long stretch of text is not particularly plausible, however, and the shift of sequence is paralleled at Suet. Tib. 50.1 (Iuliae uxori tantum afuit ut relegatae … offici aut humanitatis aliquid impertiret ut … commercio hominum frui uetuerit; cf. also Suet. Vesp. 14.1). Afuerunt remains anomalous, but it is hard to believe that so strange a usage arose as a gloss or scribal emendation. It is best left alone.

Stephanus proposed putting a lacuna before manum:

In proeliis cotidianis, quandocumque fors obtulerat, procurrentibus et erumpentibus Alexandrinis, […] manum comprehendi multum operibus et ardentibus studiis militum.°

Klotz and Andrieu credit Stephanus with two lacunas and misplace both. The lacunas in their texts are Nipperdey’s: “lacunas ego significaui” ad loc. He explains them in his prefatory “Quaestiones Caesarianae” (1847, 193).

To fix the former problem Nipperdey indicated a lacuna between manum and comprehendi, proposing [conserentes] or something similar in the first lacuna to govern manum, which now refers to combat, not troops:

The sense is good, the expression, although more Livian than Caesarian, does occur in the corpus (71.2 ut celerius omnium opinione manum consereret, BC 1.21.3), and the lacuna can be explained as originating in an eye-skip from con- to com-. As for the remaining words, Nipperdey suggested that the first three originated in something like Caesar comprehendi multum operibus noluit. This required the second lacuna shown above, to separate operibus and studiis, since the latter makes no sense with comprehendi. Nipperdey’s text is accepted by subsequent editors, including Klotz and Andrieu. However, his explanation of comprehendi multum operibus seems unlikely, in that the most recently mentioned opera were protecting the Alexandrians, not Caesar. Furthermore, it would be nice to have one problem spot instead of two. So it is worth trying to come up with a better explanation for the words after Nipperdey’s first lacuna.

1) manum [conserere optabant] (cf. Hirt. 8.41.2 hoc fonte prohiberi posse oppidanos cum optarent reliqui, from a passage that has much in common with the one under consideration and will be cited again below). Or:

2) manum [conserendam conclamabant] (cf. BG 3.18.5 conclamant omnes occasionem negotii bene gerendi amittendam non esse, ad castra iri oportere)

Neither is quite right in a sentence that begins In proeliis cotidianis. Better:

If the lacuna did originate in an eye-skip, it is difficult to determine its length. But the words that come after the lacuna provide some clues as to its contents.

1) If we make Caesarian soldiers the subject of the verb in the lacuna, we also need something that changes the subject, or at least the focalization, since these eager soldiers are later referred to in the genitive (ardentibus studiis militum). In the Hirtius passage just mentioned (8.41.2), which continues Caesar unus uideret (sc. eos prohiberi posse), the focalization shifts from reliqui to Caesar, as it often does in passages in which Caesar overrules the impetuous desires of his men (e.g., BC 1.71–72 id … ex omnibus partibus ab eo flagitabatur … Caesar in eam spem uenerat se sine pugna et sine uulnere suorum rem conficere posse; cf. BG 6.34.7, BC 3.74.2–3, BAfr 82.2–3 and, with the roles reversed, BC 1.64.2–3).

4) The mysterious diuulgata in the following sentence needs a textual anchor; at present it is hard to see anything that it might refer to, since the previously mentioned Caesarian exhortations mentioned occurred in earlier battles, one a success (15.5), one a failure (20.1, 21.1). It ought to point to something that Caesar says in this episode, something that aligns with the resulting labor and pugnandi cupiditas, and can perhaps be located in the word that governs comprehendi.

So one might fill the lacuna thus, exempli gratia:

The verbal parallel is not perfect but the sense is decent and there is a similar combination of opus, munitio, and milites at BG 1.8.4 operis munitione et militum concursu et telis repulsi (sc. Heluetii), a parallel that gives a slight edge to hostes over loca as the subject of comprehendi. All of our desiderata have now been met.

This extended discussion has shown that Nipperdey’s second lacuna is in the wrong place and may not be necessary. The remaining options seem to be these:

• Place the second lacuna after multum to give room for the word it modifies. Or
• Obelize multum. The argument for emending it to munitionum is a bit wobbly, hence the obelus.

As between these two, our preference goes to the latter as being more likely to stimulate further work on this spot, which has been untouched since 1847. With our exempli gratia supplement it reads as follows:

26.1–2

Marcie Persyn and Cynthia Damon

26 (1) Sub idem tempus Mithridates Pergamenus, magnae nobilitatis domi scientiaeque in bello et uirtutis, fidei dignitatisque in amicitia Caesaris, missus in Syriam Ciliciamque initio belli Alexandrini ad auxilia arcessenda, cum magnis copiis, quas celeriter et propensissima ciuitatium uoluntate et sua diligentia confecerat, itinere pedestri, quo coniungitur Aegyptus Syriae—(2) Pelusium adducit. Idque oppidum firmo praesidio occupatum Achillae propter opportunitatem loci (namque tota Aegyptus maritimo accessu Pharo, pedestri Pelusio uelut claustris munita existimatur), repente magnis circumdatum copiis, multiplici praesidio pertinaciter propugnantibus, et copiarum magnitudine, quas integras uulneratis defessisque subiciebat, et perseuerantia constantiaque oppugnandi quo die est aggressus in suam redegit potestatem praesidiumque ibi suum collocauit.°

“Around the same time, Mithridates of Pergamum, a man of great prestige in Pergamum and of great wit and valor in war, and loyalty and dignity in his friendship with Caesar, having been sent into Syria and Cilicia at the start of the Alexandrian War for the sake of summoning reinforcements, with a great number of troops (which he had put together quickly thanks to both the most eager desire of the citizens and his own diligence), (2) brought (sc. the troops) to Pelusium by a land route, where Egypt is joined with Syria. And as for this town, held by Achillas’ strong garrison because of the location’s convenience (for all Egypt is deemed to be fortified, as if by locks, with respect to naval attack by Pharos, with respect to land attack by Pelusium), it was surrounded suddenly with a great number of troops, and, although the multifarious garrison was fighting back stubbornly, Mithridates brought it under his authority on the very day of the attack and installed his own garrison there, thanks to both the number of his own troops—he kept substituting intact troops for the wounded and the weary—and his perseverance and steadiness for fighting.”

1. Paragraph 26 of the Bellum Alexandrinum presents four primary difficulties, all indicated by problems of syntax or style rather than by discrepancies in the manuscript tradition: first, the tense of adducit, which stands alone as a present-tense verb amid perfect-tense verbs both in the previous section and in the following clauses of paragraph 26; second, the absolute usage of adducit, which is found in all the manuscripts but not used absolutely in an active form elsewhere in the Caesarian corpus; third, the strange placement and syntax of Achillae, a proper noun that contributes little except a reminder of the identity of the enemy in the passage (and, because of this mnemonic value, may have originated as a helpful gloss provided by a later hand); fourth, the difficult phrase multiplici praesidio, which must either be understood in a constructio ad sensum as an ablative absolute with a plural participle propugnantibus or be suspected as a gloss.

However, a parallel for the mixture of tenses linked by a -que is found elsewhere within the Bellum Alexandrinum, as for example at 61.4:

Both insequitur and confert are present-tense main verbs, and they are linked—after an ablative absolute and a relative clause—by a -que with the perfect passive est deductus, clearly in secondary sequence as it is followed by imperfect subjunctives.

But why the mixture of tenses? For paragraph 26, the most probable explanation is the vividness of the present tense, which complements the dramatic staging of the introductory phrase sub idem tempus. This same opening hook is used five times in the Bellum Alexandrinum (at 26.1, 28.1, 40.1, 42.1, 64.2), and two of these passages have clear present-tense verbs (28.1 proficiscitur, 40.1 concurritur), one includes an ambiguous form (64.2 uenit), and one a perfect passive (42.1 acceptum est). In those cases where sub idem tempus introduces a verb of present (or potentially-present) tense, the surrounding sentences include a medley of perfect and imperfect tense verbs, as occurs in paragraph 26. Additionally, adducit, being itself a verb of motion, seems to have a kinship with the first three examples of the present/potentially-present variety, offering the action anticipated by the opening phrase.

There is clearly, then, no problem with the tense of adducit in this passage, and Aldus' emendation may be set aside, though it remains a helpful diagnostic emendation.

The following four emendations adapt the syntax according to the first method.

a. Davisius, in his second edition of 1727, first suggested that adducit be replaced with aduenit, an intransitive verb. This suggestion was followed by both Schneider (1888) and Klotz (1927), and is attractive for its simplicity, but it must be noted that aduenio is not found as a finite verb in the Bellum Alexandrinum, and is only found twice in participial form (32.3 dignum adueniens fructum uirtutis et animi magnitudinis tulit and, later in the same passage, aduenienti Caesari occurrerunt seque ei dediderunt). In neither of these instances is a destination designated, although parallel—albeit, again, participial—forms of aduenio with attendant accusatives of place do appear at Liu. 42.56.3 (M. Lucretius, Chalcidem adueniens … nuntium, praetoris uerbis qui abscedere eum inde iuberet, misit) and 44.30.9 (Carauantium in Cauiis Durnium oppidum aduenientem benigne accepit; see further TLL 1.831.40–50). Overall, participial forms are far more common in the Caesarian corpus (participial forms occurring at BC 2.32.12, twice at BAlex 32.3, BAfr 11.2, 23.2, 74.2, 89.2; only once in a finite form at BAfr 7.3). The infrequency of this verb with an accusative of destination, as well as the apparent preference for the verb in its participial form, together justify editors in continuing to search for alternative solutions.

b. Vielhaber’s emendation, though made over a century later, closely corresponds with that of Davisius, and thus should be treated next. He emended the text to read ad Pelusium uenit, arguing that there was a transposition of the preposition ad, which later led to a misconstrual of the verb: ad Pelusium uenit became Pelusium ad uenit and, because the verb uenit would have been abbreviated as uēit, the text would resemble Pelusium aduēit, easily (mis)read as “Pelusium aducit.” Vielhaber cited 36.3 as a model for this passage (magnis et continuis itineribus confectis cum aduentaret ad Nicopolim). His paleographic explanation is appealing, and, like Davisius’, his suggested emendation neatly sidesteps the lack of direct object by replacing it adducit with an intransitive verb.

Venit is, additionally, far more common in the Bellum Alexandrinum than Davisius’ aduenit, and is found in finite and infinitive forms twenty times in the text, and in participial forms three further times. Parallels from elsewhere in the text that can be used in support of Vielhaber’s emendation include 62.1, where the ablative of accompaniment cum copiis shares the same function as in this paragraph (paucis diebus Q. Cassi litteris acceptis rex Bogus cum copiis uenit), and 25.5, where uenit is used with ad (cum ad Canopum uentum esset).

c. Fleischer proposed Pelusium accedit, and cited four parallels from the Bellum Alexandrinum of accedo used with an accusative of destination without a preposition: 37.3 Domitius postero die propius Nicopolin accessit castraque oppido contulit, 38.1 propiusque ipse Alexandriam per Syriam accederet, 56.6 ipse classem quam parabat ut inspiceret, Hispalim accedit ibique moratur, 67.1 cum propius Pontum finesque Gallograeciae accessisset. Though the number of parallels seems impressive, Fleischer himself notes that three of his examples include the adverb propius, while paragraph 26 does not.

d. A more radical repair to the verb that attempted to emend the problems of tense and syntax was proposed by Hoffmann, who altered adducit to adductis and excised cum, resulting in the following:

This is, as is made clear from the text above, a three-part emendation: 1. Hoffmann deleted the cum of cum magnis copiis, 2. he emended adducit to adductis (a reasonable transposition), and 3. he deleted the -que of idque, reducing the number of finite, indicative verbs in this lengthy sentence to two and the number of -que’s to one. The distribution of these emendations, which requires a number of small errors to coordinate throughout the passage, makes his suggestion less likely; the great separation between magnis copiis and propugnantibus, forty-six words apart from one another, is another reason to doubt Hoffmann's emendation. Yet the repair is worth consideration because, although with each intervention his proposal becomes more unlikely, each individual emendation is plausible and his overall approach to altering the verb is of a completely different character from those that preceded him.

There are also emendations that resolve the difficult syntax of adducit without altering the form of the verb, three of which are discussed below.

e. Vascosanus replaced cum magnis copiis with magnas copias; Stephanus made the same emendation almost simultaneously. The emendation is based upon the overall sense of the passage: the ablative of accompaniment makes it clear that it is (at least in part) the copiae that Mithridates led into Asia, so the deletion of cum and modification of the cases of magnis copiis simplifies this complex sentence. But this emendation, too, requires a two-step corruption, wherein first the accusatives shifted to ablatives, then cum was inserted (or the same steps in reverse order); these changes could have occurred accidentally through transcription (though there is no nearby cum that could have resulted in dittography), or the distance between the accusative and its transitive verb could have led a particularly intrusive scribe to alter the text to its detriment. Distance alone, however, was little enough cause to prompt such an innovation, as it is unclear why the ablative of accompaniment would have been inserted in place of the accusative.

f. Landgraf in 1891 chose not to alter the paradosis at all, but rather to justify the anomalous use of adducit by means of anacoluthon. Landgraf explains the lack of a direct object for adducit as an error original to the text resulting from the ungainly length of the sentence, which resulted either in the author’s losing track of the syntax of his sentence or choosing rather to start again with emphasis—a subtle distinction that either demeans the style of the author of the Bellum Alexandrinum or underscores its similarity to other prose works incorporating rhetorical figures. This is an attractive solution, as it requires no alteration of the text at all, and can be used to explain the rambling structure of this sentence with logic rather than by assuming random clumsiness of the author and ignorance of all subsequent scribes and readers. Even with anacoluthon, however, the verb adducit strongly desires a direct object, even if only a resumptive deictic hearkening back to the ablative of accompaniment.

g. Following upon Landgraf’s supposition of anacoluthon, it is additionally possible, and, indeed, my suggested emendation, that there is a lacuna of one or more words before the verb adducit that included its original direct object. The grammatical problem of adducit is subtle enough that no subsequent scribe would necessarily suspect a lost direct object, or indeed look for or supply a correction. The sheer length of the sentence is a likely cause for such an omission: adducit occurs only at the conclusion of the long opening sentence; a lacuna may indeed have arisen precisely because of the string of genitives, participial phrases, and relative clauses that in the end masks the syntactic trouble with adducit.

It is possible that the lacuna included a number of troops, or a military unit such as a legion (cf. the list of troops provided at BC 3.4.3–6). Mithridates, although a Pergamene, may have led legions alongside the copiae he had rallied together; something similar occurs at 62.1 (rex Bogus cum copiis uenit adiungitque ei legioni quam secum adduxerat conplures cohortes auxiliarias Hispanorum). Likewise, commonly abbreviated words, such as nostros or deictics such as eas, could easily have been lost. A lacuna doesn’t fix everything, of course, since anacoluthon is still present.

3. The third problem in this passage is the word Achillae, which, in form, could be either a genitive or a dative, and is problematic both in syntax and in its position within the sentence. There are two possible grammatical functions served by this word: either it is a dative of agency with the perfect passive participle occupatum (uncommon and therefore unlikely), or it is a possessive genitive depending on praesidio. The latter is a more likely reading, but the intervening word occupatum makes for an awkward word order.

The name Achillas only occurs two other times in the Bellum Alexandrinum, and here the word serves only to remind the reader of the identity of the stronghold’s leader. Achillas himself plays no direct role in the action of the passage. The possibility must therefore be entertained that Achillae was supplied by a scribe as a gloss on the text. We therefore suggest excision as a possible repair.

4. The final problem involves the unusual phrase multiplici praesidio. As with adducit above, the issue is not one of confusion within the manuscripts, for multiplici praesidio is undisputedly the parodosis. Suspicion arises, however, because of the seeming tautology that arises if propugnantibus is construed with copiis as a drawn-out ablative absolute with a repetitive phrase contained within it. The phrase multiplici praesidio complicates the surrounding ablative phrase without, at first glance, contributing additional information about the fighting—we have already been informed by magnis … copiis that there are a large number of troops fighting. So, if we read copiis as the noun of the ablative absolute in agreement with propugnantibus, multiplici praesidio, whether understood in apposition or as a locative ablative, is superfluous—or, at the very least, the term multiplici is. At BC 1.69.4 one finds the ablative praesidio used as a locative within an ablative absolute (… atque omnes copiae paucis praesidio relictis cohortibus exeunt rectoque ad Hiberum itinere contendunt), but multiplex occurs nowhere else in the Caesarian corpus. Gemoll therefore proposed that multiplici praesidio should be removed, arguing that it was inserted into the text as a gloss. Gemoll is correct in positing that multiplici praesidio provides little additional information: yet the fact that the phrase supplies no necessary data can also be used to rebut the probability of this phrase acting as a gloss in the first place. Glosses are meant to provide helpful information to a reader, not repeat what is already stated in a synonymous phrase; thus no scholarly scribe would find the need or desire to add it in the first place.

But Landgraf (1888) defended the parodosis, retaining the phrase multiplici praesidio but construing it, not as a parallel or subordinate detail to magnis copiis, but rather as the noun paired with the participle propugnantibus in the ablative absolute. Landgraf cited Liu. 32.32.4 as a parallel for praesidium used as a collective singular (arcem regium tenebat praesidium neque ut decederent inde aut Opuntiorum minui aut auctoritate imperatoris Romani perpelli potuerunt). Not only can praesidium function collectively, but multiplex, too, may do so, and one can complement Landgraf’s defense with parallels elsewhere in Latin (see TLL 8.1592.48–60 ‘spectat ad quantitatem,’ where this passage is listed along with fourteen others).

This interpretation resolves the matter, too, of the meaning of propugnantibus, a defensive verb that generally means “to make a sortie” or “to go out to fight,” often explicitly from walls or battlements. This is an action that befits the troops of Achillas but not those of Mithridates, who are storming a well guarded city, and thus we should not expect the copiae (who belong to Mithridates) to be the subject of propugnantibus. The meaning of the participle, therefore, makes the ablative absolute extending from magnis to propugnantibus unlikely, and favors instead Landgraf’s explanation of the paradosis. Inserted in a sentence about the operations of Mithridates’ troops, multiplici praesidio pertinaciter propugnantibus reminds the careful reader of both sides of the battle and should be understood as an ablative absolute with adversative force.

27.5

Cynthia Damon

Quorum impetum Mithridates magna cum prudentia constantiaque uirtutum et Alexandrinorum imprudentia consuetudine nostra castris uallatis sustinuit. Cum uero incaute atque insolenter succedere eos munitionibus uideret eruptione undique facta magnum numerum eorum interfecit.°

“Mithridates withstood their attack with great prudence and steadfastness of his virtues and the imprudence of the Alexandrians, his camp having been fortified according to our custom. But when he saw them approaching the fortifications recklessly and immoderately, he made sallies from every side and killed a great number of their men.”

The ungainly sentence presented by the paradosis here accumulated substantive innovations in the process of transmission and has accordingly been the object of a variety of interventions. As a first step towards establishing a text it will be important to state the textual evidence clearly, since, as will soon become clear, most editors have resorted to excision and consequently shown little interest in the problems raised by the excised text.

The version of the sentence found in S appears in editions from the editio princeps onwards (the only early exception is the edition of Beroaldus, who prints the paradosis). That is, it made its way into the vulgate long before S itself was used in the constitution of the text, via N, a descendant of S; Brown (1972, 48–49) describes the source of the editio princeps as a manuscript that contains “a text based on a union of M and N … that was the basis of succeeding editions for nearly four hundred years.” So although the phrase constantiaque uirtutum et Alexandrinorum imprudentia has been confronting texts that exclude it since the first time the readings of M and N were joined in a single manuscript, only Beroaldus (1504) admitted it into the text. Nobody wants it. The long string of ablatives is clumsy, and the juxtaposition of moral terms such as prudentia, constantia, and imprudentia with the matter-of-fact consuetudine nostra castris uallatis is jarring.

A different solution was suggested tentatively by Dübner and embraced by Landgraf (1891b), who suggested that constantiaque belonged in the text and the rest arose from a marginal gloss of the following form: uirtutes Mithridatis. Alexandrinorum imprudentia, a shorter form of the gloss discussed above. Dübner’s solution supplies a possible origin for the problematic uirtutum, which started life as a nominative plural referring to prudentia constantiaque and was converted into a genitive when it was moved into the text. Klotz' point about the apparent irrelevance of constantia remains, but the pair prudentia constantiaque can perhaps be explained by the antithesis with the following incaute atque insolenter, itself an unusual collocation in the corpus.

On balance it seems that Dübner’s emendation best accounts for the features of this problematic passage, particularly the origin of uirtutum. The string of ablatives is still awkward, but there is a comparable collection in the preceding paragraph (25.2 magnis circumdatum copiis multiplici praesidio pertinaciter propugnantibus et copiarum magnitudine … et perseuerantia constantiaque oppugnandi quo die est aggressus). The somewhat irrelevant praise of Mithridates can be ascribed to political exigency, and the fact that S makes one of its characteristic eye-skips here is coincidental.

31.3

Cynthia Damon

Quorum perturbatione nostrorum animi adeo sunt incitati ut paene eodem tempore ex omnibus partibus, primi tamen editissimum castrorum locum caperent.°

“The spirits of our men were so incited by the enemy’s confusion that they captured positions at nearly the same time from all directions, the stonghold’s highest point, however, first.”

The problem here lies in the change of subject at primi tamen, which leaves the implicit subject in the first part of the result clause, presumably nostri, without a verb. Kübler accordingly posited a lacuna before primi and suggested filling it with impetus fieret; the expression is paralleled at BG 1.22.3: ut undique uno tempore in hostes impetus fieret. Klotz prints this in preference to Schneider’s subsequent suggestion inrumperent, imperfectly paralleled at BG 4.14.3 quorum timor cum … significaretur, milites nostri … incitati in castra inruperunt. Inrumperentur improves on impetus fieret by supplying a subject that can stand more easily in antithesis with primi tamen, but even Schneider claims no more for this repair than that it suits the sense of the passage (“Hinter partibus fehlt ein Verbum von der Bedeutung ‘vorstürmen’ oder ‘einbrechen’; das von mir eingesetzte inrumperent genügt dem Sinne nach.”). Andrieu leaves an unfilled lacuna in his text but supplies [l’assaut se produisait] in the translation.

35.3

Maria Kovalchuk and Cynthia Damon

Cuius itineris has esse certas [constat] opportunitates, quod in locis superioribus nullus impetus repentinus accidere hostium poterat et quod Cappadocia his iugis subiecta magnam commeatus copiam erat subministratura.°

“And [it is agreed] that these are the indubitable advantages of this route: the fact that on higher ground no sudden attack of enemies could occur, and the fact that Cappadocia, adjacent to these ridges, was going to supply a great abundance of provision(s).”

In chapters 34–35 Domitius orders Pharnaces to withdraw from Cappadocia and Lesser Armenia, which had been the kingdoms of Deiotarus and Ariobarzanes, Roman allies. Pharnaces withdraws from Cappadocia but not from Lesser Armenia. Consequently, Domitius marches towards Armenia with troops, traveling W-E along a ridge between Cappadocia (to the south) and Armenia (to the north). Incertus specifies two advantages of this route in a sentence whose syntax is incomplete. The main clause is transmitted as follows:

This is printed in many editions, including those of Scaliger (1635) and Nipperdey (1847), presumably on the assumption that a uerbum noscendi could be understood from instituit two sentences earlier. But it is not obvious that iter facere instituit will launch a train of thought, and if it does, the thought is immediately interrupted by an eighteen-word topographical parenthesis (Nam … Armenia).

In some recentiores and the editio princeps the sentence is rewritten thus:

A less drastic repair is to supply a finite verb to govern the indirect statement, e.g.:

cuius itineris has esse cernebat opportunitates Vielhaber (1869, 570)°

Although cernere usually denotes visual perception rather than realization in the corpus Caesarianum, it does govern indirect statement at BC 3.69.4 cum … Pompeium adesse … cernerent. For Vielhaber’s argument see note .

Larsen (1886, 18–19) proposed a different construction:

In support of his emendation he cited BC 1.1.3 sin Caesarem respiciant atque eius gratiam sequantur as a parallel for the abstract object of sequor and Cic. Off. 1.35 sed credo aliquid secutor, opportunitatem loci maxime for the pursuit of opportunity. By replacing esse certas with est secutus Larsen also addressed the problem of certas. However, Larsen’s paleographical argument (-us misread as -as) is rather weak.

Landgraf (1891b, 12–13) developed Larsen’s emendation in two ways:

36.4–5

Maria Kovalchuk and Cynthia Damon

Quibus ex castris cum locus angustus atque impeditus esset transeundus, Pharnaces in insidiis delectos pedites omnesque paene disposuit equites, magnam autem multitudinem pecoris intra eas fauces dissipari iussit paganosque et oppidanos in his locis obuersari (5) ut, siue amicus siue inimicus Domitius eas angustias transiret, nihil de insidiis suspicaretur cum in agris et pecora et homines animaduerteret uersari tamquam amicorum aduentu, sin ut in hostium fines ueniret, praeda diripienda milites dissiparentur dispersique caederentur.°

“Since a narrow and obstructed place had to be traversed (sc. by anyone approaching Nicopolis) from this camp, Pharnaces set selected infantry and almost all his cavalry in an ambush. Moreover, he ordered a large quantity of livestock to be scattered within the pass and the peasants and townsfolk to linger in these places, (5) so that, whether Domitius traversed this narrow spot as friend or enemy, he would have no suspicion of an ambush when he noticed that both livestock and men were moving about in the fields as if at the arrival of friends, or, if he came as if into enemy territory, his soldiers would be scattered in plundering booty and slaughtered piecemeal.”

The sense is fine up to suspicaretur, and even up to aduentus provided that one supplies esset to complete the syntax of the tamquam clause. But thereafter things become murky. What kind of ut clause comes next? How are dissiparentur and caederentur connected to the rest of the sentence? And why is ut in hostium fines so far from siue inimicus?

The friend/enemy antithesis and the siue … siue pair have suggested various approaches to repairing the text.

Taking his lead from a reading found in the editio princeps, Aldus excised siue inimicus and replaced the adjacent siue … siue pair with a more distant si … sin uero pair:

ut siue amicus siue inimicus Domitius eas angustias transiret, nihil de insidiis suspicaretur, cum in agris, et pecora, et homines animaduerteret uersari, tamquam amicorum aduentu, s[in uero], ut in hostium fines ueniret, praeda diripienda milites dissiparentur, dispersique caederentur.°

Square and pointed brackets are supplied here and below for the sake of clarity. Aldus also omitted the word greges that appeared in the editio princeps (see the previous note, ).

He thereby clustered the two references to friendly behavior and the two references to hostile behavior, and simplied the syntax of tamquam with aduentu as a circumstantial ablative. The resulting sense is good but the repair requires at least three innovations, whose sequence is hard to understand. Did si amicus generate siue inimicus, which then generated siue in place of si? Did uero fall out before ut, and aduentu acquire a final s, whereupon the orphaned in was excised as nonsensical?

The difficulty of tracing a path from the original text to what is found in the manuscripts prompted Nipperdey (1847, 195) to try a different approach. He transposed siue inimicus to later in the sentence, simultaneously changing the syntax of aduentus by dissolving its final s into the transposed phrase:

With this text the puzzling ut introduces an elliptical comparison ut in hostium fines parallel to tamquam amicorum aduentu. The initial ut clause expresses Pharnaces’ double purpose: that the Romans either fail to suspect the presence of the enemy (ut … nihil … suspicaretur) or scatter in pursuit of booty and get killed (ut … dissiparentur dispersique caederentur). The two-part purpose is supported by parallel siue clauses, siue amicus … transiret and siue inimicus … ueniret, a construction for which Nipperdey cites two Ciceronian parallels (Phil. 14.13, Fin. 1.3). Clarity, it seems, has been restored, and this is the text printed by Klotz. But the explanation of how clarity was lost is still complicated: siue inimicus was omitted, the omission was noticed, a repair was attempted by writing the omitted words elsewhere on the page, and they were then reincorporated into the text at a plausible but incorrect position following siue amicus. The alteration of aduentu to aduentus in this scenario is poorly motivated: siue inimicus seems to have left its initial letter behind and regenerated it during the course of its travels. Forchhammer (1852, 92) adds that siue inimicus contributes nothing when placed in front of ut in hostium fines; he prefers to see it as a gloss motivated by the simple siue that originally preceded ut in hostium fines, but his repair is as complicated as Nipperdey’s.

Andrieu adopts a simplified version of the repair proposed by Aldus, with si in place of sin uero.

The series of innovations is now clearer: aduentu acquired a final s, the left-over minim was omitted, and si amicus was expanded into siue amicus siue inimicus to make up for the loss of the second si clause and apply the friend/enemy antithesis for the verb transiret; ut … ueniret would then mean “supposing he came.” But this is still a three-part innovation involving two spots in the text.

Hoffmann printed the same text in 1890 and argued against Nipperdey’s emendation (why move siue inimicus out of the first conditional clause if deceiving Domitius would have been even more important if his arrival was hostile? 1890, LIV). The adversative sin is used twice elsewhere in the Bellum Alexandrinum (24.1, 42.4), the sequence siue … siue … sin in Cicero (De orat. 2.292, Att. 7.3.2). Hoffmann’s repair implies a two-part innovation (aduentu =⟩ aduentus, in is omitted) but neither stage is per se implausible and the fact that they involve the same spot in the text is a distinct advantage over the repairs discussed so far. Dübner adopted it, and so do we.

40.2

Cynthia Damon

At Pontica ex altera parte legio, cum paulum aduersa hostibus cessisset, fossam autem circumire ac transcendere conata esset ut aperto latere aggrederetur hostem, in ipso transitu fossae confixa et oppressa est.°

“But on the other side the Pontic legion, when it had withdrawn a small distance facing the enemies but had tried to circumvent and get across the trench so that it could attack the enemy on its open flank, was stopped and overwhelmed in very act of crossing the trench.”

When Domitius and Pharnaces finally fought, the Romans were decisively defeated. Three of the four legions involved in the battle suffered catastrophic losses, and Domitius retreated to Asia Minor with the veteran Thirty-sixth and a few other survivors (40.4–5). The battle itself must have been chaotic, but Incertus seems to have been aiming for a narrative as clear as Caesar’s usually are. The battle lines are described in some detail (38.3–4, 39.2), then the events on the Roman right (40.1), then those on the left (40.2), then those in the middle of the line (40.2). After the Roman middle collapses, we return to the right wing (40.3) and finish with the “cost-of-battle” assessment (40.4–5). The framework is clear. But problems in the transmitted text prevent us from seeing exactly what happened on the Roman left:

As Nipperdey says (1847, 197), “Verba ‘acies secundo’ quin corrupta essent, recte nemo dubitavit.” But those two words are not the only problems in this passage. Here is a fuller list:

• The unqualified reference to an acies here is baffling.
• The adverb secundo appears nowhere else in the corpus, and one would expect it to be preceded by primum (uel sim.).
• It is difficult to see which verbs go with which nominative. Word order suggests that legio is the subject of cessisset, acies the subject of the others, but since we learn later that the Pontic legion was almost completely annihilated in this battle (40.4 Pontica legione paene tota amissa) it seems likely that legio is the subject of confixa et oppressa est. However, since it is the acies, not the legio, that is mentioned in connection with the fossa, the relevance of in ipso transitu in that scenario is unclear.
• The term transitu is per se a puzzle, since Incertus has described the troop movement as “going around” (circumire) not “going across.”
• The expression fossam circumire is not only in conflict with in transitu but also hard to reconcile with the battlefield described here, where Pharnaces’ compact battle line is protected on the flanks by 4-foot trenches (38.3–4) and in the rear, as becomes clear from 40.1, by the walls of Nicopolis. How then could one go around the trench to attack the enemy from the rear? (See further below, under B).

The context provides some parameters for solutions to these problems.

• According to 39.2, Domitius’ battle line consists of a right wing (the Thirty-six legion), a left wing (the “Pontic legion,” i.e., the unit newly recruited in Pontus [34.5]), and a media acies (two legions contributed to the Roman war effort by the Galatian king Deiotarus [34.4], separated by a narrow gap). Domitius also placed an unspecified number of cohorts in unspecified positions from which they could serve as reinforcements.
• The phrase at … ex altera parte sets the narrative of events on the left wing in antithesis to that of events on the right wing, where the attack of the Thirty-sixth proceeded as follows: the legion advanced outside the trench (extra fossam) on Pharnaces’ left, reached the city walls (moenibus oppidi succederet…), crossed the trench (fossamque transiret), and attacked the enemy from the rear (auersosque hostes adgrederentur). A battlefield success-story, as Incertus says (40.1 secundum proelium).
• In 40.2 both auersa and secundo echo words used in 40.1 (see B above).

Lipsius, according to Oudendorp, addressed problems (1) and (2) as follows:

Ciacconius (cited from Jungermann 1606) addressed all five problems as follows:

Madvig (1873, 283–284) created a similar scenario with a simpler repair:

Excision has also been applied to the problems of this passage, without much plausibility since it is unclear what would motivate the addition of words so plainly alien to the context as acies and secundo.

The efficiency of Nipperdey’s emendation, which like that of Ciacconius deals with all five problems at once, makes up for its lack of palaeographical plausibility:

His ac transcendere crisply eliminates acies and secundo and justifies transitu. The verb is used of crossing trenches at BC 3.46.3 ut fossas transcenderent. In our view this is the best of the available repairs, particularly when combined with Madvig’s aduersa. If it is right, acies and secundo may be further persistence errors: forms of the former occur eight times in the preceding three paragraphs, and secundum, as was mentioned above, was used in 40.1, where it is preceded by the word adeo. But perhaps the problem lies deeper, as Klotz suggests by daggering the passage.

46.1

Cynthia Damon

Vatinius cum animaduerteret neque nauium se magnitudine neque numero parem esse, fortuitae dimicationi fortunae rem committere maluit. Itaque primus sua quinqueremi in quadriremem ipsius Octaui impetum fecit.°

“Although Vatinius realized that he was inferior both with regards to the ships’ size and their number, he preferred to entrust the matter to a chance engagement. Thus, he attacked first, driving his quinquereme against the quadrireme of Octavius himself.”

Chapter 46 is part of the narrative of Vatinius’ res gestae after the battle of Pharsalus (43.3–47). This narrative starts and ends in Brundisium, but our passage belongs to Vatinius’ naval operations on the coast of Illyricum, where he was vying with the Pompeian fleet commander M. Octavius. Vatinius’ “fleet” was a motley collection of a few warships and a larger number of light vessels retrofitted with rostra (44.2–4), and Octavius knew it (45.1). Catching Vatinius off guard in a storm Octavius decided to attack, whereupon Vatinius rapidly prepared his fleet for battle. The general urgency of the task of stopping Octavius’ depredations was made clear earlier (44.1–3), but Vatinius’ reasoning at this critical moment is given in the following sentence, which contains a puzzling string of nouns, a murky cum clause, and an elliptical verb:

Vatinius preferred battle—to what? Evasive action, presumably, but Incertus does not specify. If specification is wanted, one could emend maluit, as the scribe of T did, or fill in the blank, as Kübler did:

Kübler’s repair also includes tamen, which clarifies the concessive nature of the cum clause; others had added it elsewhere. But this too, although an improvement, seems unwarranted, since many of Incertus’ cum clauses, especially in chapters 34–47, are murky owing to postponed conjunctions and complicated syntax (see, e.g., 34.2, 35.2, 36.3, 39.1, 40.2, 42.3, 44.1, 44.3). The antithesis between calculation (magnitudine, numero, parem) and chance (fortuitae, fortunae) may suffice to indicate the concession that Incertus intended.

But the expression fortuitae dimicationi fortunae rem committere is distinctly unlovely. Rem committere could go with dimicationi (cf. BC 2.33.2 proelio rem committere; 2.38.2, 3.74.2) or fortunae (cf. Liu. 9.12.11 committere rem fortunae) but not with both, so fortunae must be genitive. But if fortunae is a subjective genitive dependent on dimicationi (for the construction see TLL 5.1.1197.21–34), fortuitae is tautological. And if fortuitae modifies fortunae, it is otiose. In short, something has to go. Suspicion has fallen on fortuitae owing to the word’s scant presence in the corpus Caesarianum (BG 7.20.2 omnia fortuito et sine consilio accidere; cf. BAfr 3.5 fortuitu). Fortuna, of course, is a favorite Caesarian concept (⟩70x, including recently at 43.1 and 43.3).

47.2

Tim Warnock and Cynthia Damon

Capit ex eo proelio penterem unam triremes duas dicrotas VIII compluresque remiges Octauianos.°

“From that battle he gained possession of one penteris, two triremes, eight two-banked galleys, and several of Octavius’ oarsmen.”

Stadler (1907, 1119) proposed adding one more Greek-derived boat term to this list, emending remiges to celoces. But the capture of rowers with their boats is reported in another “cost-of-battle” summary earlier in the text (16.6 capitur hoc proelio quinqueremis una et biremis cum defensoribus remigibusque; cf. BC 3.24.3 quadriremem cum remigibus defensoribusque suis ceperunt, BAfr 63.3 triremem … onustam remigum epibatarumque cepit), so the repair seems superfluous.

49.1

Tim Warnock and Cynthia Damon

Cassius legionibus in hiberna dispositis ad ius dicendum Cordubam se recepit contractumque in ea aes alienum grauissimis oneribus prouinciae constituit exsoluere. Et, ut largitionis postulat consuetudo, per causam liberalitatis speciosam plura largitori quaerebantur.°

“Having established the legions in their various winter quarters, Cassius went to Corduba for the purpose of hearing legal cases, and he decided to pay off debt contracted in the province by means of extremely heavy burdens on the province. And as the custom of largesse demands, the specious pretext of generosity was used, but the intended beneficiary of the majority of the requests was the donor.”

A variety of substitutes for the offending demonstrative have been suggested.

Jurinius (1742, ad loc.) supplied an adjective:

Madvig (1873, 284), noting that if in ea refers to Corduba, the expression ought to be in ea urbe, suggested two replacements for it, an ablative noun and a temporal adverb:

Alea has little to recommend it apart from its final two letters, since nothing in the narrative connects the cause of Cassius’ indebtedness with the measures he took to to remedy it; gambling would be a stray detail. Antea is a little better, since Incertus did mention Cassius’ earlier office in Spain, and the temporal adverb does not point specifically to Corduba. However, it adds nothing to the sense, since the participle contractum itself indicates priority.

Menge (1889, 126) suggested a different temporal adverb:

Interea is closer to the paradosis, and Menge (ibid.) explains it as a reference to the time “während er Recht spricht.” But the word order associates it closely with contractum and makes it point to an unspecified period of time between Cassius’ quaestorship and praetorship and events nowhere mentioned in the commentarii.

On this reading, Cassius accumulated this debt during his time in Spain as Pompey’s representative, and he used his recent largesse to the soldiers as an excuse for extortion and financial chicanery (described in the following sentences), some of the proceeds of which went to canceling his debt.

49.2–3

Tim Warnock and Cynthia Damon

Pecuniae locupletibus imperabantur, quas Longinus sibi expensas ferri non tantum patiebatur sed etiam cogebat. In gregem locupletium simul et causae tenues coniciebantur, neque ullum genus quaestus aut magni et euidentis aut minimi et sordidi praetermittebatur quo domus et tribunal imperatoris uacaret. (3) Nemo erat qui modo aliquam iacturam facere posset quin aut uadimonio teneretur aut in reos referretur. Ita magna etiam sollicitudo periculorum ad iacturas et detrimenta rei familiaris adiungebatur.°

“Funds were requisitioned from the wealthy, and Longinus not only permitted but even forced these to be set down for him (i.e., in his account books) as expenditures. At this same time frivolous lawsuits, too, were launched against the crowd of wealthy men, and no kind of profit, either great and apparent or small and sordid, was neglected, such that the commander’s residence and tribunal were uninvolved with it. (3) No one who could make at least some outlay was not either constrained by a guarantee or listed among the accused. Thus great anxiety about personal peril, too, was added to expenses and damage to family fortunes.”

Chapter 49 concerns the finances of Caesar’s governor in Further Spain, Q. Cassius Longinus, particularly his extortion of money from the Spaniards. The first sentence supplies necessary context: one of Cassius’ goals was to reduce his personal debt (49.1 contractum in ea [sc. prouincia] aes alienum … constituit exsoluere). Incertus devotes four sentences to the governor’s procedures, but problems in the transmitted text of the first part of the second sentence obscure the details (the rest of the passage is quoted above for context):

Some basic orientation can be derived from the insistent binaries in the passage quoted above: requisition vs. prosecution, large and visible profit vs. small and sordid profit, domus vs. tribunal, financial risks (uadimonio … iacturas … detrimenta rei familiaris) vs. criminal charges (in reos … periculorum). Incertus’ Cassius seems to have applied pressure both directly and indirectly in pursuit of money to pay his debts.

Causa, the reading of the nu-branch, is a preposition-like ablative with simultatium (cf. Var. L. 5.155 litium causa). With this reading the subject of coniciebantur is the substantival tenues, “people of modest means.” Neither simultatium causa nor tenues coniciebantur is easy to understand, however. Nipperdey (1847, 197) paraphrases the sentence thus: “homines tenues, quia Longino erant inuisi, eadem passos esse, quae locupletes.” But simultatium causa seems like an odd way to communicate Cassius’ hostility, and translations are tellingly disparate. Nipperdey’s interpretation of causa, with its causal (“because of”) rather than purposive (“for the sake of”) meaning, is reflected in some translations, e.g.:

• Andrieu (1954): “en cas de démêlés avec lui, le petit peuple était jeté dans la masse des riches.”
• Carter (1997):“Men of modest means were lumped in with the rich because of private quarrels.”

• Ciaffi-Griffa (2008 [1953/1972]): “includeva poi le persone di condizione modesta nella massa de ricchi per confondere le cose.”
• Raaflaub (2017): “People of modest means were classed with the rich arbitrarily in order to stir up social tensions.”°
  “Classed with the rich” is an effective way of dealing with in gregem … coniciebantur (cf. BHisp 22.7 siqui ex nostris transfugerent in leuem armaturam conici), but “arbritrarily” has no basis in the text.
• Landgraf: “gegen das Corps der Reichen wurden simultatium causa die Armen aufgewiegelt” (see below for the verb translated here).

The difficulty of simultatium causa prompted Schneider to emend the genitive:

Klotz (1927, ad loc.), who prints simulationis causa, explains it as follows: “i. in numerum eorum quibus pecuniae imperabantur, etiam pauperes referebantur, ne locupletium divitias peti appareret.” But Incertus’ Cassius is unconcerned about his public image—for another of his flagrant abuses see note in this section—indeed he so provokes the provincials that he is the target of two unsuccessful assassination attempts and a successful defection.

Furthermore, the context makes it problematic to take tenues as the subject of coniciebantur, since it is clear that Cassius favored wealthy targets: 49.3 Nemo erat qui modo aliquam iacturam facere posset quin aut uadimonio teneretur aut in reos referretur. To address this problem Landgraf (1888, 49) emended the verb:

With this reading the tenues are not Cassius’ targets (real or simulated) but his tools: he arouses them to initiate prosecutions that they will profit from, but intends to profit from their profits himself. Landgraf (1889, 13) later claimed that the passage should be viewed as “definitiv geheilt.” His repair does permit a logical connection to the preceding sentence about largesse (49.1 Et, ut largitionis postulat consuetudo, per causam liberalitatis speciosam plura largitori quaerebantur): these prosecutions will count as largesse for the tools. But in a context where actual largesse has been mentioned (48.2–3) and will be mentioned again (52.1) the connection seems rather tenuous, and Klotz (1927, ad loc.) deems Landgraf’s verb ill-suited to the context.

Simultatium is still a problem, however, indeed a worse problem than it was with causa. The awkwardness of its position next to locupletium was mentioned above, and with causae the genitive is ambiguous. The other occurrence of simultates in the corpus Caesarianum refers to the contests between Pullo and Vorenus (BG 5.44.2 summis simultatibus), and at BC 2.25.4 the singular refers to a dispute involving Curio and Pompey. But “trivial causes of contests (or disputes)” seems to give Cassius’ victims too much agency: why give them reasons to fight him? The term can also denote lawsuits (e.g., Liu. 39.5.2 alienarum uero simultatium tribunum plebis cognitorem fieri turpe … esse), but neither “causes of lawsuits” (objective genitive) nor “cases, i.e., lawsuits” (defining genitive) seems likely here.

A long-neglected 16th-century emendation improves the sense signficantly:

60.2

Isabella Reinhardt and Cynthia Damon

Marcellus cum confligere miserrimum putaret, quod et uictoris et uicti detrimentum ad eundem Caesarem esset redundaturum, neque suae potestatis esset, legiones Baetim traducit aciemque instruit.°

“Marcellus, although he thought that combat was utterly wretched, because the loss for both the victor and the defeated would similarly redound on Caesar, and it was not within his official capacity, led his legions across the Baetis and drew up his battle line.”

In chapter 60 the rebellion against Cassius has reached a point where armed conflict seems inevitable. Of the five legions originally controlled by Cassius two and a half have mutinied (57.1–5) and put themselves under the leadership of the quaestor M. Marcellus (59.1). Marcellus’ forces are based in Corduba itself, Cassius’ are on high ground across the Baetis River, about four miles distant (59.1). Cassius has taken the initiative by ravaging Cordoban fields and buildings outside the city (59.2), and Marcellus’ troops demand action (60.1). In the present sentence Marcellus does act, but with considerable reluctance, the grounds for which are only partly clear in the text as transmitted, which is printed by Nipperdey and other editors (here with Nipperdey’s punctuation):

A relatively simple emendation was already available in the recentiores:

Another fairly simple repair was suggested by Ciacconius, who added an infinitive after esset. Jungermann’s report of the repair (1606, 333) does not indicate whether Ciacconius intended the verb to be part of the cum clause (with neque connecting putaret and esset) or part of the quod clause (with neque connecting esset redundaturum and esset):

or

In the cum clause esset requires an awkward shift from concessive sense (“although he thought”) to causal (“since it was not in his power to prevent”). In the quod clause it is something of a non sequitur after confligere miserrimum putaret.

Larsen (1886, 25) left the matter of what was not in Marcellus’ power open, placing a lacuna after esset:

Marcellus cum confligere miserrimum putaret, quod et uictoris et uicti detrimentum ad eundem Caesarem esset redundaturum, neque suae potestatis esset […], legiones Baetim traducit aciemque instruit.°

Klotz (1927, ad loc.) and Andrieu (1954, ad loc.) oddly print the recentiores’ esse and Larsen’s lacuna.

Another possibility would be to find a supplement that would place esset in the extant quod clause, along the lines of Ciacconius’ emendation, discussed above. As an exempli gratia supplement based on the subsequent course of events we considered:

With this supplement Marcellus would consider combat utterly wretched “because the loss for both the victor and defeated would similarly redound on Caesar and retreat was not in his power.” The phrase esset receptus could be misconstrued as a pluperfect verb form, of course, but the same is true at BC 1.45.6 Hac nostris erat receptus and BC 3.45.5 erat per decliue receptus. However, the difficulties of the terrain are not mentioned until later in the narrative (60.4–5, esp. cum … quid transitus fluminis uitii difficultatisque haberet cognitum esset), and the two explanations for miserrimum are both strangely dissimilar and strangely arranged, with the proximate and practical worry about retreat placed after the future and general worry about what the battle would cost Caesar. And even if one solves the latter problem by placing a comma before neque and making esset [receptus] parallel to putaret, the fact that we don’t know about the difficulties of the terrain gives pause.

In the end we decided that no supplement (and hence no lacuna) is needed here, and that the transmitted text, despite the awkward syntax of confligere and the ambiguity of potestatis, was acceptable. Marcellus’ worry about the limits of his authority aligns well his subsequent decision to withdraw his troops at the first opportunity (60.3) and with Caesar’s obiter dictum at BC 3.51.4: aliae … sunt legati partes atque imperatoris. Marcellus’ loyalty to Caesar is stressed at 59.1, his deference to authority at 63.2 (huic [sc. Lepido] uenienti sine dubitatione Marcellus se credit), and his reluctance to fight is mentioned more forcefully than in the present passage at 61.4 (quoted above). The repairs reported in the apparatus indicate our lingering dissatisfaction with the paradosis and suggest possible approaches for improving it.

63.5

Brian Credo and Cynthia Damon

Non tantum indutiis factis sed prope iam constituta opera [cum] complanarent custodiaeque munitionum essent deductae, auxilia regis in id castellum Marcelli quod proximum erat regis castris necopinantibus omnibus—si tamen in omnibus fuit Cassius: nam de huius conscientia dubitabatur—impetum fecerunt compluresque ibi milites oppresserunt.°

“With not only a truce arranged but when they (sc. Lepidus and Marcellus) were leveling their now nearly constructed earthworks and the fortifications’ guards had been removed, the king’s auxiliary troops made an attack against the stronghold of Marcellus that was closest to the king’s camp—when no one was expecting it, if in fact that ‘no one’ included Cassius, for there was doubt concerning his complicity—and overpowered many soldiers there.”

Aldus (1513, ad loc.) made the obvious repair, supplying the conjunction cum before complanarent:

The postponement of cum is a feature of Incertus’ style (see Gaertner and Hausburg 2013, 36–39). However, this repair makes the non tantum … sed structure oddly imbalanced, with non tantum introducing an ablative absolute and sed a subordinate clause.

Nipperdey (1847, 200) addressed the new problem by adding pace to create a second ablative absolute with constituta:

The opera are now unmodified, but their identity is clear from the preceding narrative (61.5 operibus in circuiti oppidi continuatis, 62.2 munitiones … ab operibus); Cassius had demanded their dismantlement (63.4 postulat uti munitiones disicerentur). Parallelism has been restored, but it is hard to explain a two-part innovation here: either there were two unrelated omissions or a scribe tried to repair a text that was nonsensical after cum was omitted before con-. But in the latter case the excision of pace is not the most obvious move.

Castiglioni (1924, 239) suggested that cum and pace were adjacent when omitted:

Non tantum indutiis factis sed prope iam constituta [pace cum] opera complanarent custodiaeque munitionum essent deductae, auxilia regis … impetum fecerunt.

This simplifies the innovation but weakens the paleographical justification for the omission of cum. This repair was adopted by Klotz (1927). But the expression constituta pace is problematic, since constituere pacem appears nowhere in the Caesarian corpus, despite the fact that Caesar refers to many peace negotiations; Caesarian expressions include pace facta (BG 2.29.5, 3.1.4), pace confirmata (BG 4.48.1), pace concilianda (BG 7.55.4, BC 1.26.3, 1.85.2), pace petita or petenda (BG 4.13.1, 4.27.4). Of the small number of passages listed in the relevant spot of the TLL (4.515.58–61), the best parallel comes from Cicero: Q. fr. 1.1.25 pacem tota prouincia constitutam. So we considered two nouns that are used with constituta.

As for complanarent, a verb that appears nowhere else in the corpus Caesarianum, the transmitted active form is hard to construe (no subject) and parallel (the best is Sen. Dial. 1.5.9 quae … conplanet ipse). In emending the verb to a passive form Kübler makes the syntax simpler, but here again the best parallel comes from outside the corpus (Cic. Dom. 101 domus est complanata), so it is hard to feel confident that he is right. And it seems rash to emend when the subject of complanarent could have fallen out with cum.

66.3–4

Brian Credo and Cynthia Damon

Ibi rebus omnibus prouinciae et finitimarum ciuitatium constitutis cupiditate proficiscendi ad bellum gerendum non diutius moratur magnisque itineribus per Cappadociam confectis biduum Mazacae commoratus Comana […] uetustissimum et sanctissimum in Cappadocia Bellonae templum, quod tanta religione colitur ut sacerdos eius deae maiestate imperio potentia secundus a rege consensu gentis illius habeatur. (4) Id homini nobilissimo Lycomedi Bithyno adiudicauit, qui regio Cappadocum genere ortus propter aduersam fortunam maiorum suorum mutationemque generis iure minime dubio, uetustate tamen intermisso, sacerdotium id repetebat.°

“When all the affairs of the province and the neighboring cities were settled there, owing to his desire to set out to make war he delayed no longer, and stopping at Mazaca for two days after forced marches through Cappadocia, Comana […] the shrine of Bellona, the oldest and holiest in Cappadocia, which is worshipped with such great fervor that the priest of this goddess is considered second to the king by the consent of that people in majesty, command, and power. (4)  He awarded it to the very noble Lycomedes, a Bithynian who was born in the royal family of the Cappadocians and was seeking that priesthood again, his right (sc. to it) being not at all in doubt but nevertheless long discontinued on account of the adverse circumstances of his ancestors and a change in the royal line.”

The problems in this passage about Caesar’s journey from Cilicia to Pontus through Cappadocia are both historical and textual. Both center on the word Comana.

The main historical problem is the identity of Comana. There are two cities called Comana in Cappadocia and Cappadocian Pontus, but it is hard to say which is meant here, since the information supplied by the reference to the famous temple of Bellona, which points to the Cappadocian city, conflicts with the route deduced from the prior reference to Mazaca, the capital of Cappadocia: a journey to Comana after two days at Mazaca would have taken Caesar away from Pontus and the coming fight with Pharnaces, towards which, as the passages stresses, he was hastening eagerly (66.3 cupiditate proficiscendi ad bellum gerendum non diutius moratur; 66.5 simili uelocitate). A secondary problem is the apparently unmotivated two-day stop in Mazaca, which is hard to ignore given the incompatibility of non diutius moratur and commoratus.

The main textual problem is the absence of a main verb for the second part of the opening compound sentence (magnisque … habeatur). This obscures the syntax of Comana uetustissimum et sanctissimum in Cappadocia Bellonae templum. Are Comana and templum nominative? accusative? in apposition? The secondary textual problems indicated in the apparatus involve the asyndetic list of ablatives of respect (66.3 maiestate imperio potentia) and the position of the ablative absolute (66.4 iure minime dubio, uetustate tamen intermisso); neither needs discussion here.

An early repair to the main textual problem, the addition of uenit before Comana, completed the syntax and made Comana and templum accusative. But it left the two in uncomfortable apposition. Nipperdey’s placement of uenit after Comana instead of before allows one to explain the verb’s omission as a jump from uen- to uet-. But the apposition is still a problem, and both repairs exacerbate the historical problems by apparently having Caesar detour to Cappadocian Comana after an unproductive stay in Mazaca. Madvig declared the supplement misguided and the text sound, taking Comana and templum (still in apposition) as objects of adiudicauit and explaining the closer object id as a resumptive pronoun necessitated by the distance between Comana and its verb (31 words). His analysis of the syntax is not entirely persuasive—why id, not ea, which would remove the apposition by making Comana and templum parallel objects in asyndeton?—but on his reading the problem of Caesar’s route through Cappadocia disappears (1873, 285): “Mazacae, non Comanis, de templo statuit.” Caesar made his decision about Comana and the temple without going there. The solution is particularly appealing because the following sentence, although somewhat opaque owing to its lacuna, clearly concerns decisions pertaining to Cappadocian royalty, whose seat was in Mazaca. Madvig’s explanation also supplies welcome information about why a hurrying Caesar spent two days in Mazaca: two days for two decisions about the future administration of a powerful priesthood and a kingdom. But syntax of the transmitted text remains unsatisfactory, and subsequent editors, including Klotz and Andrieu, have posited, with Kübler, a multi-word lacuna between Comana and uetustissimum.

Desiderata include (1) a verb, (2) a syntactic differentiation between Comana and templum that accommodates the demonstrative id, and (3) content that keeps Caesar out of Cappadocian Comana, that allows for a reference to Comana Pontica, and that explains Caesar’s stop at Mazaca. Here are two possibilities for the missing syntax; the words in each lacuna are supplied exempli gratia:

Both of them make Comana the object of a verb in a subordinate clause whose conjunction is postponed and, after repunctuation, use adiudicauit as the desired main verb. Incertus postpones ut twice (45.3, 56.6) and cum frequently (Gaertner-Hausburg [2013] 36–39), so the stylistic oddity passes muster.

In the first version the lacuna contains a verb, confirmarentur, that can govern both Comana and templum, which are connected by et. There is a parallel for the double object at BC 1.29.3 (interim ueterem exercitum duas Hispanias confirmari … nolebat [sc. Caesar]), and Incertus had earlier used confirmare in a passage where Caesar was intent upon settling protracted disputes involving Eastern royalty (33.2 priusquam diuturnitate confirmarentur regum imperia). The omission would have originated in a skip from ut to uet-. With this supplement Caesar’s stop in Mazaca is explained by the purpose clause but it is still hard to see Comana as anything but the Cappadocian city, given the presence of in Cappadociae … templum and, later, regio Cappadocum genere ortus. And the supplement sits somewhat awkwardly with id, since the demonstrative points to a resolution of the temple issue but leaves the problem of Comana apparently unresolved.

Both supplements deliver some but not all of the desiderata listed above. Neither acquits Incertus of confusing the two Comanas—assuming that Strabo deserves our credence—but both provide a main verb, keep Caesar out of the Cappadocian city, and explain Caesar’s stop. The syntax of each repair makes a welcome distinction between the city and the shrine, but in neither version is id entirely unproblematic. So neither is good enough to go in the text, and we follow Klotz and Andrieu in indicating a lacuna.

67.1

Amelia Bensch-Schaus and Cynthia Damon

… quod, in ea parte positus terrarum quae nulla praesidia Caesaris habuisset [et coactus] exercitibus imperiisque, in Cn. Pompei castris fuisset.°

”… the fact that, as he was located in a part of the world that had no Caesarian garrisons and was under the compulsion of armies and orders, he had been in the camp of Gnaeus Pompey.”

This extract is part of Deiotarus’ speech as a suppliant before Caesar. His defense has two parts. Here, he explains that his military support of Pompey was a consequence of the absence of Caesarian troops in his region. In the following sentence he adds that he did not think he should be a judge of Roman conflicts. Caesar’s response is given in 68.1; he objects mainly to the latter sentence.

The main puzzle in this passage is the syntax of exercitibus imperiisque. In the paradosis neither word is obviously connected to any expression in the vicinity, and the plural forms make a peculiar pair:

Scholars have proposed numerous solutions.

Markland (1723, 129) proposed emending to excitusque imperiis, transposing the connective and converting exercitibus into a causal participle parallel to positus.

The instrumental ablative is a neat and attractive solution to the syntax problem, and one can hypothesize a two-step innovation: first the substitution of the common word exercitibus for the rare participle excitus, and second the transposition of the connective to connect the two nouns. But excitus is indeed rare. In fact the only extant example occurs at BHisp 4.4 itaque Cn. Pompeius … litteris fratris excitus … ad Cordubam iter facere coepit. So this solution is not entirely convincing.

In 1861 Kraner (LVIIII) suggested a slight modification to Markland’s repair, excising the connective and thereby linking excitus more tightly to fuisset.

Here the distinction between the two causal participles is hard to justify.

In 1873 Madvig proposed emending exercitibus to excitus uerbis, pointing out that excitus itself is too short to yield exercitibus:

The connective now links uerbis and imperiis in a hendiadys. But uerbis does not add much to the sense of what is happening, there are no close parallels pairing uerbis with imperiis, and the separation between positus and excitus remains hard to justify.

In 1886 Larsen (27) suggested a paleographically and semantically superior version of this repair:

Precibus contributes more meaning than uerbis, and the scribe of the archetype will have substituted exercitibus for a less common expression beginning with ex- and ending with -ibus. Larsen cites a passage from Cicero’s speech for Deiotarus as evidence for the different ways in which the king’s behavior could be characterized: Cic. Deiot. 13 uenit (sc. Deiotarus ad Pompeium) uel rogatus ut amicus uel arcessitus ut socius uel euocatus ut is, qui senatui parere didicisset. With Larsen’s reading Deiotarus appears to be covering himself with a similar range of options.

In 1887 Dinter took a new approach to the puzzle, adding coercitus before exercitibus imperiisque and leaving the noun phrase as in the paradosis:

Coerceo is not as rare as excio, although the perfect passive participle occurs nowhere in the Caesarian corpus. The verb itself is used once elsewhere in the Bellum Alexandrinum, of rulers who might prove ungrateful to Rome: 33.4 si essent ingrati, posse isdem praesidiis coerceri. Similarly, there is an ablative of means with the infinitive at BC 1.67.4: at luce multum per se pudorem omnium oculis, multum etiam tribunorum militum et centurionum praesentiam adferre, quibus rebus coerceri milites et in officio contineri soleant. On the whole, this suggestion is helpful in preserving the transmitted nouns and in proposing a more common verb, although the absence of any participial forms of this verb is troubling.

In 1889 Cornelissen tried a drastically different approach, inserting in before Caesaris:

The next year Hoffmann, who had printed the paradosis in his first edition, offered his own solution to the problem of exercitibus imperiisque, inserting coactus after the pair:

Cogere is a fairly common word in the Caesarian corpus, and this use of the participle with an ablative of means is paralleled in the Bellum Alexandrinum: 43.1 magnisque difficultatibus coactus. On the whole, this is a promising suggestion, although critics have felt that the line still needed further work.

In 1927 Klotz, retaining Cornelissen’s in, added aduersariorum before Hoffmann’s coactus:

Since none of the solutions discussed has been entirely satisfactory, we have adopted a different repair. We follow most editors in inserting a participle to govern the instrumental ablatives; Hoffmann’s coactus seems the most persuasive suggestion, since it is the most common in Caesarian usage. But we place the participle before exercitibus imperiisque and insert a connective to facilitate the coordination of positus and coactus:

The omission of et and the following word might have been prompted by the ending of habuisset. We have left exercitibus imperiisque intact, since the exercitibus of unspecified supporters of Pompey corresponds to the praesidia that Caesar has not provided. Admittedly, the unusual combination is problematic. Furthermore, if Deiotarus is claiming that he was forced to join Pompey by Pompeian armies, one might expect Caesar to point out that the king did not actually engage in combat with these armies. Instead, in his reply, Caesar focuses exclusively on Deiotarus’ subsequent claim that he did not know whose imperia to obey. To keep the focus on imperiis, we have suggested potestatibus as a possible alternative to exercitibus, a term that is found in combination with imperiis: cf. Cic. Leg. 1.23 si uero isdem imperiis et potestatibus parent, and Liu. 4.54.7: salii flaminesque nusquam … sine imperiis ac potestatibus relinquantur.

68.1

Amelia Bensch-Schaus and Cynthia Damon

Contra quem Caesar cum plurima sua commemorasset officia quae consul ei decretis publicis tribuisset cumque defensionem eius nullam posse excusationem eius imprudentiae recipere coarguisset, quod homo tantae prudentiae ac diligentiae scire potuisset quis urbem Italiamque teneret …°

“In reply to him, Caesar, after he had recalled his own very numerous services, which he as consul bestowed upon him by public decrees, and after he had argued that his (sc. Deiotarus’) defense could not contain any excuse of ignorance on his part, since a man of such great foresight and attentiveness was able to know who held the city and Italy …”

The transmitted text of this passage contains a bold personification and an apparent pleonasm: recipere has defensionem as its subject and excusationem as its object. A lesser objection is the fact that defensio has a military meaning in its seven other occurrences in the corpus Caesarianum.

Madvig condemned the construction in no uncertain terms (1873, 285 n.1 “hae sordes”) and praised the repair found in Clarke’s 1712 edition and in some manuscripts—replacing the objectionable defensionem with defectionem—on the grounds that “factum recipit aut non recipit defensionem excusationemque imprudentiae” (ibid.). However, the textual support for Madvig’s obiter dictum is rather thin. The best parallel for a “factum” as the recipient of an excuse is Cic. Agr. 3.5 inuidiosa lex … habet excusationem, where the verb is different. It is in fact clear from the TLL entry on excusatio that a person is normally the subject of such expressions, as at BG 6.4.3 libenter Caesar petentibus Haeduis dat ueniam excusationemque accipit (TLL 5.2.1299.73–82, with verbs including accipere, adimere, dare, habere, obtendere, parare, praebere; there are two examples of recipere). The “factum” is normally expressed with a genitive, e.g., imprudentiae in the present passage (TLL 5.2.1298.55–82).

Klotz suggests in his apparatus that the transmitted text can perhaps be explained as a pleonasm. Cicero uses a similarly pleonastic expression for “defense by excuse” at Rab. Post. 27: Rufum necessitatis excusatio defendet. The wording is different, but the concept is perhaps sufficiently similar to defend the paradosis, particularly given the problems inherent in the repairs currently on offer. The fact that defensio is exclusively military elsewhere in the corpus Caesarianum should not make us reluctant to accept its forensic meaning here, since the context contains other “courtroom” words: 67.1 reorum habitu, 67.2 iudicem esse, 68.1 de controuersiis tetrarcharum … se cogniturum esse.

72.2–3

Wes Hanson and Cynthia Damon

Circumpositi sunt huic oppido (sc. Zelae) magni multique intercisi uallibus colles, quorum editissimus unus, qui propter uictoriam Mithridatis et infelicitatem Triari detrimentumque exercitus nostri, superioribus locis atque itineribus paene coniunctus oppido, magnam in illis partibus habet nobilitatem nec multo longius milibus passuum tribus abest ab Zela—hunc locum Pharnaces, ueteribus paternorum felicium castrorum refectis operibus, copiis suis omnibus occupauit.°

“Many large hills separated by valleys surround this town. And the highest of these, a single hill that on account of Mithridates’ victory and Triarius’ misfortune and the defeat of our army ([sc. it is] practically connected to the town by relatively high ground and paths) has a great notoriety in those parts and is not much more than three miles away from Zela—(3) this location Pharnaces occupied with all of his troops after the old defenses of the camp that was lucky for his father were repaired.”

The main problems in the paradosis pertain to the stretch of text quorum … Zela.

• The clause introduced by quorum lacks a verb.
• The words superioribus locis atque itineribus paene coniunctus oppido, which pertain to terrain, seem out of place in the middle of a relative clause about the memory of an earlier battle. But the unusual expression (see below) is unlikely to have originated as a gloss. And the severity of this problem depends in part on what one takes to be the point of coniunctus oppido: is the hill’s connection to Zela relevant to the tactics of the coming battle, or to the memory of both battles? As it happens, the town per se plays no role in either battle: it is nothing more than a place name for battles fought in the vicinity. Accounts of the battle of 67 BCE situate it on level ground (see above, note ), while Incertus situates Caesar’s battle near a collis more than three miles from the town.
• The syntax of superioribus locis is puzzling: by itself it might be taken as a local ablative (cf. BC 1.65.1 locis superioribus constitit and possibly 12.1 superioribus locis subleuabantur; contra, 35.3 in locis superioribus) but with atque itineribus it seems to be a rather loose ablative either of route (cf. 35.3 locisque superioribus iter facere instituit) or means (cf. 30.3 castellum … brachiis … cum opere castrorum coniunxerat). Furthermore the adjective is unexpected given the prominence of editissimus—which places are higher than the highest hill?—and the expression superioribus … itineribus is unparalleled.°
  In the apparent parallel at BC 1.51.2 superiorum temporum atque itinerum licentia the adjective has a temporal sense. Klotz (1927, ad loc.) suggests adding [commodis] before itineribus.
• The grammatical connection created by nec between habet and abest seems arbitrary, given the divergent referents of the two verbs, reputation for the one, terrain for the other. The latter verb seems better suited to the clause introduced by quorum editissimus.
• The adjective unus is strangely emphatic. There is another emphatic unus at 25.6 Ita qui unus ex omnibus eo proelio bene rem gessit solus cum sua quadriremi uictrice periit, but in that passage unus prepares the way for solus. The adjective in our passage has no comparable role. It is hard to say whether its omission by S is deliberate. Another possibility is that unus originated in a misconstrued abbreviation, here possibly for est (cf. ad BC 3.19.4, where the original might have been primo).

Problem 4, too, responds to relatively simple approaches. Oudendorp (1737), for example, uses punctuation:

Problem 2 has been addressed by a variety of transpositions, all of them serial innovations: omission of the original string of words, addition of the string in the margin uel sim., insertion of the string into the text at an inappropriate spot.

Morus (1780, ad loc.), who prints the paradosis, explains in a note how to address problems 1, 2, and 4 by moving superioribus … oppido to after nobilitatem, making the relative clause a parenthesis, and altering nec to non:

But later (1869, 574) he abandoned this complicated repair and suggested instead that superioribus … oppido was an incorporated gloss that was subsequently corrupted.

Madvig (1873, 1.48–49) addressed problem 2 by moving superioribus … oppido to the end of the sentence.

This creates a useful, if tacit, antithesis between abest and paene coniunctus, strengthening the topographical significance of coniunctus. But as Vielhaber says (1869, 574) the term is “bei einer Entfernung von mehr als 3/5 Meilen auffällig.” And in this position superioribus is oddly lacking in point: how does the close connection on high ground qualify the absolute distance between the hill and the city?

Larsen (1886, 28–29) added a solution for problem 3 to Madvig’s repair by turning the transposed string into a terminal ablative absolute:

The resulting text bears a certain resemblance to the description of the terrain at Nicopolis: 36.3 quod oppidum positum (sc. Nicopolis) in Armenia minore est plano ipsum loco, montibus tamen altis ab duobus lateribus obiectis, satis magno interuallo ab oppido remotis. But the presence of tamen here makes the point of this terminal ablative much clearer.

Problem 3 is addressed head-on by Kübler (ad loc.), who rewrites the text thus:

74.4

Wes Hanson and Cynthia Damon

Cuius aliquamdiu Caesar irridebat inanem ostentationem et eo loco militum coartationem quem in locum nemo sanus hostis subiturus esset, cum interim Pharnaces eodem gradu quo in procliuem descenderat uallem ascendere aduersus arduum collem instructis copiis coepit.°

“For some time Caesar kept laughing at the empty show and the massing of soldiers in a spot into which no sane enemy would advance, when in the meantime Pharnaces, with the same pace with which he had descended into the downward-sloping valley, began to climb the steep hill opposing him with his troops in battle formation.”

The transmitted text of a sentence that shows an incredulous Caesar watching Pharnaces lead his army onto unfavorable ground needs repair in at least three spots:

The fact that the valley into which Pharnaces had descended was described as praerupta in the preceding sentence (74.3 At Pharnaces … descendere praerupta ualle coepit) suggests that the adjective may have been supplied here as a convenient repair for the semantic problem discussed above. On the other hand, eodem gradu suggests that Incertus wanted the two phrases read together here, and in any case he is not particularly scrupulous about avoiding repetition (see Gaertner and Hausburg 2013, 37 n. 38). But if in praeruptam was the original reading, it is hard to see how in proelium arose. It cannot have been an unconscious remembrance of a phrase used elsewhere in the corpus, since in proelium descendere occurs only at Fron. Strat. 2.1.10 ut in proelium cum coniugibus ac liberis descenderent, and in proelium is altogether absent from the corpus Caesarianum.

Nipperdey (1847, ad loc.) proposed an emendation that clarifies the origin of in proelium:

The noun will have arisen as a misreading of procliuem, an adjective that appears only three times elsewhere in the corpus (76.1 cum in procliui detruderentur hostes; BC 1.48.8 quibus erat procliue tranare flumen; BAfr 10.1 omnia sibi procliuia), in two of which it has its metaphorical meaning “easy.” Procliuis is not elsewhere applied to uallis, but it is used with topographical substantives such as pauimentum, solum, saxum, locus, and uia (see TLL 10.2.1537.68–1538.13 “de iis quae qui positionem uel formam inclinatam habent”). And it is used in antithesis with arduus at Liu. 5.43.2 (in arduum … per procliue) and Sen. Dial. 7.25.6 (in procliui … aduersus ardua).

Both in praeruptam and in procliuem have merits, and it is difficult to choose one over the other. In the end, however, we decided to print Nipperdey’s emendation, to indicate that we think that in praeruptam is more likely to be a scribal innovation based on 74.3 praerupta ualle than to be authentic text.