Jump to ContentJump to Main Navigation
Cicero, Greek Learning, and the Making of a Roman Classic$

Caroline Bishop

Print publication date: 2019

Print ISBN-13: 9780198829423

Published to Oxford Scholarship Online: December 2018

DOI: 10.1093/oso/9780198829423.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 27 January 2020



(p.259) 6 Cicero
Cicero, Greek Learning, and the Making of a Roman Classic

Caroline Bishop

Oxford University Press

Abstract and Keywords

This chapter investigates Cicero’s desire to enshrine himself as a classic from a different perspective than the rest of the book. It analyses Cicero’s appropriation not of a classical Greek figure but of himself, by examining his self-quotation of his earlier poetry (primarily the Aratea) within his late philosophical dialogues De Natura Deorum and De Divinatione. While Cicero likely found the Greek source for his poem, Aratus, quoted within the Greek philosophical works he used as sources for these dialogues, quoting his own poetry obviously carried a different charge. The chapter concludes that by staging the reading of his earlier works within these dialogues, Cicero was modelling the proper way to read his corpus: namely, as a set of works every bit as authoritative as the Greek classics he had adapted throughout his literary career.

Keywords:   Cicero, Posidonius, De Natura Deorum, De Divinatione, Aratea, Aratus, Stoicism

‘Sed quo potius utar aut auctore aut teste quam te?’

De Divinatione 1.17

‘But what authority or witness can I better employ than you?’

Quintus Cicero utters these words immediately before launching into a nearly eighty-line quotation from his brother’s De Consulatu. His literary tastes closely resemble those of Quintus Lucilius Balbus, the Stoic interlocutor of De Natura Deorum, who quotes extensively from Cicero’s Aratea. In fact, Quintus’ and Balbus’ quotations from Cicero’s poetry share a number of similarities. In the two dialogues in which the men appear, Marcus Cicero also features as a character, and in both, Cicero-the-character listens in silence while his poetry is praised and then used as evidence for the various philosophical topics under discussion. It is a long time to be silent, since each dialogue features a substantial amount of Cicero’s poetry: over ninety lines of the translation of Aratus appear in De Natura Deorum, and more than a hundred from the poems Marius, De Consulatu, and Prognostica are quoted in De Divinatione.

De Natura Deorum and De Divinatione were composed in 45 and 44, respectively, as part of Cicero’s turn to philosophical pursuits under Caesar. But this was not the first time Cicero had returned to the published works of his younger self under these changed political circumstances. As I showed in Chapter 4 (sections ‘Brutus’ and ‘Orator’), in Brutus and Orator Cicero had already taken up the question of how to memorialize his earlier successes, though there it was his career as an orator and politician he sought to memorialize. In Brutus, he fashions a biography of his early career in which he resembles Demosthenes, while in Orator he actually turns to self-interpretation, quoting from his more notable speeches in a bid to create a canon of his greatest oratorical triumphs and provide the grounds on which they could be studied by scholars and students.1

(p.260) In those two works, as I showed, Demosthenes’ reception in the Greek world offered an attractive model for Cicero of his own potential reception as an orator. His use of the reception of a Greek classic in order to visualize his own reception was not new; he had already done the same in his adaptations of Aratus, Plato, and Aristotle (in De Oratore, at least). But the prominence he gives in Brutus and Orator to his posthumous legacy is striking, and is a sign of how decisively Caesar’s dictatorship had shifted his interests from literary or political success in the immediate present to success as a literary classic in a more nebulous future. Brutus and Orator, then, serve as a bridge between Cicero’s earlier adaptations of Greek figures and the even more bold self-interpretation on display in De Natura Deorum and De Divinatione. For in fact, in the latter two dialogues, the author whose reception Cicero takes up is not Greek, but Roman: it is Cicero himself.

This is not to say that these dialogues do not draw on Greek sources; like all of Cicero’s late philosophical works, they are creative adaptations of a variety of Hellenistic philosophical resources. But these Greek sources were not classics in the way that Plato, Aristotle, Demosthenes, and Aratus were classics; many were close contemporaries to Cicero, and they did not have the weight and authority of the earlier Greek works he had adapted. What they offered instead was a model of classical reception, one that Cicero realized could provide a guide as he set about creating a set of parallel Roman classics. As I have already discussed, classicism had been a driving factor in much Greek philosophy from the mid to late second century on, and focus was increasingly turning to the interpretation of key figures like Plato and Aristotle, who had begun their evolution into the paragons of peerless authority that they had become by later antiquity.2 But they were not the only classical authors whose words could be found in the philosophical works of this period: the majority of these treatises would also have contained quotations from classical Greek poetry to illustrate and, in some cases, even prove their philosophical points.

The role that this approach to poetry could have in the Roman world was clear to Cicero. By the late republic, Rome’s intellectual elite had begun to consolidate a canon of the Roman poetry of the third and second centuries, and much contemporary scholarship centred on equipping this canon with an interpretive apparatus.3 Cicero’s philosophical treatises most definitely participate in this canonization, frequently drawing on the comedy, tragedy, and epic of the second century for quotations, and he must have recognized that in (p.261) imitating the poetic quotations of his Greek philosophical predecessors, he was helping to create a parallel poetic canon.4 But it is the more self-interested project of canonization in these treatises—the canonization of Cicero’s own poetry—that will be my concern in this chapter.

It is clear that Cicero’s self-quotation in De Natura Deorum and De Divinatione continues the project he began in Orator: by staging the reception of his earlier works within these dialogues, Cicero provides his readers with a model of the proper way to read and to interpret him as a literary classic. The nature of this model is nowhere better summed up than in the two nouns Quintus uses in the quotation with which I began: auctor (‘author’ or even ‘authority’) and testis (‘witness’). Cicero demonstrates in these moments of self-quotation that he is in possession of the qualities implied by these two related terms: like the authors of the Greek classics whose ipsissima verba were being given increasing focus in Greek philosophy, Cicero too is possessed of a literary authority whose weight can add to the force of a philosophical argument. His poetry also has the heft of a witness’ legal testimony—the type that could be used to provide valuable evidence for either side of a debate. Furthermore, just like his Greek classics, he can also provide another sort of auctoritas: the kind that writers on grammar like Varro used to determine correct linguistic usage (GRF 1.289–90, no. 268).

The desire for literary and stylistic authority on display in these moments of self-quotation sits ill at ease with the larger motivations that drive the late philosophical works, grounded as they are in the mistrust of authority and desire to consider all sides of a question that was Cicero’s inheritance from the New Academy. The refraction of the quotations through two interlocutors who claim allegiance to Stoicism (a philosophical system much friendlier to the idea of poetic authority) is a way for Cicero to partially control for this inconsistency. In the end, though, there is only so much he can do when beset with two such contradictory impulses: his strong desire to eschew reliance on authority and his equally strong hope of becoming a literary classic.

Modelling Reception in the Philosophical Dialogues

Nowhere is Cicero more interested in his own reception that in the philosophical dialogues he composed during and after Caesar’s dictatorship. In fact, it is not only his poetry whose reception he seeks to model in these late (p.262) philosophical works; it is that of the philosophical works themselves. His preoccupation over who will read these works and how they will do so is evident throughout them: both in their prologues (many of which were composed as a group and only afterwards attached to separate treatises) and in the use of internal readers who react to earlier dialogues in later ones.5 These devices can be seen as explicit formulations of the same themes that implicitly characterize the quotation of Cicero’s poetry within the works themselves: an ongoing preoccupation with the way his literary labour will be received, and an attempt to model its correct reception.

While we might not normally connect Cicero’s poetry to his philosophy, their similarities become clear once we recognize that Cicero had good reason to fear for the reception of each. In fact, his poetry was vulnerable to precisely the same sort of attack he anticipates for his philosophical works: namely, that it was not of direct value to the Roman state in the way that his political career had been. Cicero’s attempts to show that his philosophical project had a tangible benefit to the state have been well documented, but it is worthwhile to outline briefly the nature of his defence of the project before considering how it might intersect with the more implicit defence he mounts for his poetry within these works.6

From the outset of this philosophical project, Cicero was at pains to justify its value. This work was even done on a large scale by the lost Hortensius, composed between March and April 45: in it, Marcus Cicero counters the objections to the study of philosophy by Hortensius, who represents here the quintessential Roman politician arguing that the study of rhetoric is a more valuable endeavour.7 Though it is lost, its companion piece Lucullus (composed in May of that year) survives, and the preface to that work provides the earliest extant indication of the critics Cicero believed his project would face. In this preface, he breaks his critics into three distinct groups: ‘there are many (1) who have no love of Greek literature at all, and even more (2) that dislike philosophy specifically, while the rest (3), even if they do not disapprove of such things, still do not think the discussion of them is becoming for the leading men of the state’.8

(p.263) Each group typifies a different critique to which Cicero believed his project was vulnerable. The first questions the value of translating Greek literature into Latin at all (certainly a criticism that had been applicable to many of Cicero’s literary works over the course of his career); the second disputes the value of philosophical translation in particular; and the third questions the validity specifically of Cicero’s involvement with such a project. Similar objections to the project are raised in the prefaces to later dialogues, although the idea of the xenophobic anti-Greek Roman (critic number 1 in this passage) is replaced, in the prefaces to De Finibus, Academica, and the Tusculan Disputations, with the Philhellene who despises any work written in his native tongue.9 To rebut these anticipated critics, Cicero aims in these prefaces to show that the transfer of Greek cultural material (philosophy in particular) is not only appropriate, but is also an appropriate form of otium for a former consul—and perhaps even a new form of negotium now that Caesar’s reign has removed the normal political functions of the statesman.

Cicero makes his case not just by mounting defences against these particular critics or by openly praising the merits of his philosophical project (as he does at Luc. 5–9, Div. 2.1–6, and Nat. D. 1.11), but also more subtly, by attributing positive statements about his philosophical works to other well-connected Romans (whether they be his dedicatees or his interlocutors) whose approval lends authority to the project.10 The most overt use of this trope comes in the prologue to De Finibus, where Cicero, addressing Brutus, suggests it was only the favourable response to the Hortensius by Brutus and other competent readers that induced him to continue the project (1.2); here, the assertion that respectable Romans approve of his work is a large part of what lends it respectability in the first place.

Cicero mounts a similar defence within the dialogue portions of the works by frequently introducing interlocutors who have read the earlier philosophical dialogues, and who comment favourably upon them. In De Fato, for example, Hirtius recalls reading the Tusculan Disputations (4), a very appropriate choice of dialogue for the moment at hand, since he has just asked to hear one of Marcus’ oratorical exercises, and Marcus has offered him a choice between that or an argument in utramque partem on a philosophical thesis. Hirtius says he would be willing to listen to the latter, since he knows from his reading of the Tusculan Disputations how effectively Marcus can argue against a philosophical thesis. In the Tusculans, of course, philosophy and rhetoric are explicitly linked by the practice of in utramque partem debate, and the (p.264) former is actually called ‘the declamation of my old age’ (Tusc. 1.7: senilis est declamatio). Hirtius’ comments thus not only confirm the close link Cicero had forged between rhetoric and philosophy in the earlier dialogue, they also further the fiction that these dialogues preserve Cicero’s living voice: Hirtius promises that he will judge what Cicero is about to say (ostensibly the De Fato) in the same way he has already approved of what Cicero wrote (the Tusculan Disputations) (Fat. 4).11

Hirtius is not the only internal reader of Cicero’s works; indeed, such readers can be found in Cicero’s dialogues as early as De Legibus, where, as I discussed in Chapter 2 (n. 61), both Quintus and Atticus are intimately familiar with De Re Publica (e.g. Leg. 1.27, 2.23, 3.12). But Cicero makes particular use of the trope in his later philosophical works: so, for example, the interlocutor in the Tusculan Disputations presses Marcus on the apparent inconsistency between his arguments in that work and the ones that he made in Book 4 of De Finibus (5.32),12 and in the authorial preface to De Amicitia, Cicero himself becomes a willing participant in the fiction, claiming that when he rereads De Senectute, he often forgets that it was himself and not Cato who made the arguments there (4).13 The trope is at its most marked in De Divinatione. In the preface, Cicero names the work a sequel to De Natura Deorum (1.7), and throughout the first book, Quintus is portrayed as a close reader of that earlier dialogue. At the outset, he tells his brother that ‘a little while ago, I carefully read the third book of your De Natura Deorum’ (Div. 1.8: perlegi…tuum paulo ante tertium de natura deorum), and he goes on to use several examples from it to make a case for the existence of divination.

As I have already discussed in Chapter 2 (section ‘Plato in Philo’s Academy’), these sorts of breaks in the fictive integrity of his dialogues allow Cicero to keep his readers from becoming too invested in the authority of the fictions he creates, and were part of the heritage of the New Academy. But when they involve his interlocutors discussing his own earlier works, they are also doing something more, and De Divinatione is the most notable example. Quintus’ engagement with De Natura Deorum in that dialogue is intimately connected to his use of Cicero’s poetry within it and to his method of argumentation; both are topics that I will take up in the final section of this chapter. But for (p.265) now I would simply like to note the significance of the verb he originally uses to describe his reading of that dialogue: perlegi, where the prefix denotes the thoroughness of the practice. Much like Brutus in Orator read and reread Cicero’s and Demosthenes’ speeches (Orat. 105: lectitas), thus validating and vouching for Cicero’s rhetorical style and the canonical value of these speeches, Quintus in De Divinatione is represented as the consummate reader of his brother’s philosophical and poetic works: as we find out shortly after he makes this pronouncement, he has studied both so closely that he can recite from them by heart.

The fact that Quintus builds his case using an earlier philosophical dialogue of his brother’s alongside earlier works of his poetry shows nicely that the two projects were in some sense coterminous for Cicero. And, as I have already suggested, a link between them is not difficult to find: both were types of writing that a Roman consular was not expected to produce. Much like philosophy, poetry—especially poetry of a lengthy, epic nature—was not the province even of literarily inclined Roman politicians.14 Prominent military and political figures had others to write their aggrandizing verses, and if they themselves composed poetry, it was of the light, nugatory sort, like the epigrams of Catulus and his circle that were discussed in Chapter 1 (section ‘Cicero’s Aratea’). Indeed, while Cicero may simply have been anticipating attacks on his philosophy that never came, his poetry had already been viciously attacked, the De Consulatu in particular.15 The bulk of this criticism seems to have come from Cicero’s political enemies (who would have used any stick with which to beat him), but it is likely that his poetry was especially susceptible to their attacks, at least in part because the composition of it was seen as outside the realm of acceptable senatorial activity.

It is interesting, however, that when Cicero responds to his poetic critics, he does so not just on political grounds, but on literary ones as well: he is especially contemptuous in his speech against Piso, when he claims that it was because Piso failed to recognize basic poetic figures like metonymy that he misinterpreted the infamous line cedant arma togae, ‘let arms yield to the toga’ (Pis. 72–5).16 Antony too is portrayed as misunderstanding this verse, and his (p.266) ignorance is described as part and parcel of a larger inability to understand literature at all (Phil. 2.20). Of course, in a different genre—and in different political circumstances—Cicero is willing to admit that Piso’s reading of the verse as promoting Cicero’s rhetorical skills over military prowess was basically correct (Off. 1.77–8). But it was useful to him, in public performances, to portray his enemies as culturally and literarily ignorant, bad readers who failed at interpretation.

The internal readers of the philosophical works, on the other hand, tend to be good readers: if not always ideal, at least almost always so (although the situation is more complicated with Quintus in De Divinatione).17 Though Cicero responds to numerous critics of the project in the prefaces, these critics are never portrayed as having actually read the dialogues in question: if they had, presumably their response would be as positive as the response of the internal readers within them. Perhaps, then, Cicero saw this realm of positive feedback and dutiful readers as a place where his poetry might not just survive, but even thrive.

Cicero considered his philosophical works congenial to poetry in another sense as well. In the prologues to both Academica and De Finibus, responding to philhellenic critics who consider it unnecessary to translate Greek philosophy into Latin, he links his philosophical project to the widespread adaptation of Greek poetry by the Roman poets of the third and second centuries.18 De Finibus 1.4–10 offers the fullest example of the analogy (though compare Acad. 1.10).19 Even among those who prefer Greek literature to Latin, what Roman is there, Cicero asks, who would reject Ennius’ Medea or Pacuvius’ Antiope for Euripides’ versions?20 Who has such hatred for their native tongue that they would prefer Menander over Caecilius or Terence? And, he adds, if Romans have no complaint about the Medea of Ennius, why would they object to having Plato or Aristotle in Latin?

Cicero then builds on the analogy. In fact, his philosophical project is not simply a translation of Plato or Aristotle; it is not, then, wholly similar to these straightforward adaptations of Greek drama by Roman playwrights. In fact, the sophisticated approach he takes to his Greek sources in these dialogues (p.267) actually surpasses the word-for-word translation of the poets: he must choose his source material judiciously, join it to his own opinions, and arrange the resulting adaptation into a new and coherent structure (1.6).21 His work is therefore closer to a different sort of Roman poetry: Ennius’ epics, or Afranius’ comoediae togatae, which make more sparing and judicious use of Greek material (1.7).

Yet no matter which of these two models Cicero prefers, it is still poetry that provides the standard by which his readers should judge his philosophy, and the connection has a certain logic to it. Not only did Roman poetry provide the most worked-out programme for adapting Greek literature, it was also the genre that was currently being subjected to canonization and scholarly study among Roman intellectuals. By arguing that his philosophical works should be judged in the same way Romans judge the poetry of the second century, Cicero both provides a well-trod hermeneutic for his project and also implies—all but explicitly—that he hopes his own works will be canonized and studied in the same way. To underscore his fitness as a classic, he proceeds to quote his own poetry within these dialogues alongside the greatest poets of earlier Roman literature, helping to create and affirm the evolving poetic canon while also placing himself inside it.

Hellenistic Philosophy and Roman Poetry in the Philosophical Dialogues

Cicero’s decision to use his philosophical works as a venue in which to recuperate and canonize his poetry has its roots in the Greek sources for his project. Greek philosophical treatises were, of course, the sine qua non of this project: its very nature, which aimed to give thorough accounts of the tenets of each philosophical school on a variety of topics, required Cicero to be a compiler of multiple Greek sources, although he was not as mechanical a copyist as he has sometimes been portrayed.22 Indeed, the venture—as Cicero himself notes in the passage from De Finibus in which he links the project to (p.268) the poetry of Ennius, Pacuvius, and Afranius—was indisputably a creative one: it entailed Cicero collecting representative samples of each school’s teaching, connecting them to one another by means of an appealing dialogic frame, and making them suitable for a Roman audience by assigning each section to a Roman speaker who incorporated Roman exempla and poetry into his argument. Cicero considered this sort of project to be sanctioned by his affiliation with Philo’s Academy; as I discussed in Chapter 2 (section ‘Plato in Antiochus’ Academy’), with no positive doctrine of their own, Academics collected the beliefs of other schools either to pronounce the need for suspension of judgement, or, in the weaker strain of scepticism represented by Philo, to determine which tenets were most probable.23 At several points in these dialogues Cicero notes that, like the philosophical school to which he claimed allegiance, his aim is simply to provide evidence and then allow others to determine its truth or falsity for themselves (e.g. Luc. 7–9, Tusc. 5.11 and 5.83, Nat. D. 1.10–12, Div. 2.150).

But it is also clear how this project, which involved Romanizing the tenets of Hellenistic philosophy on a grand scale over a very short period of time (eleven treatises were composed between early 45 and late 44) could be misunderstood as the work of a copyist or translator. Certainly Cicero recognized the potential for this critique, which is why he went to such lengths in the prologue to De Finibus to distance his work from ‘mere’ translation. And there is no doubt that he saw the project as something else entirely. In one letter to Atticus, he boasts that there is nothing quite like the Academica, even in Greek (Att. 13.13–14 [321]), while in another he describes his compositional practice, also with regard to the Academica, as not so much a translation of his Greek source as a paraphrase written with a distinctly Ciceronian sheen (Att. 13.19 [326]).

With this turn of phrase, Cicero is not just referring to the prose style whose idiosyncrasies he hoped would become the norm for classical Latin, but also to the dramatic details that he crafted for these dialogues: he took great care with their settings, choosing interlocutors and dramatic dates in order to achieve as much verisimilitude as possible.24 His preoccupation with this aspect of the dialogues is well illustrated by the Academica, whose speaking parts were first given to Catulus, Lucullus, and Hortensius and then transferred to Atticus, Varro, and himself once he decided that its subtle philosophical arguments were unrealistic in the mouths of aristocratic politicians from a (p.269) previous generation.25 But in other cases, he believed that respected Roman interlocutors could add authority to an argument, as in De Senectute, which Cicero argues in De Amicitia was enhanced by Cato’s standing: ‘this style of conversation seems somehow to have more weight when it is placed under the authority of those distinguished men of old’.26 Of course, this was because the more general nature of a work like De Senectute made it a suitable topic for a less philosophically advanced Roman of an earlier generation.

Verisimilitude also extended to the portrayal of each philosophical school within the dialogues, and Cicero is careful to give as realistic and even-handed an account of each school’s doctrine and practices as possible. It must have been in this aspect of his composition that the idea of quoting his own poetry first came to him, for, as Jocelyn has shown, Cicero’s verisimilitude towards the Greek philosophical schools extends to their attitudes towards poetry.27 Far from being simply decorative or included on a whim, the poetic quotations in Cicero’s philosophical works are designed to accurately reflect each school’s use of poetry in arguing for its respective tenets. A brief survey of the major schools’ attitudes towards poetry will suffice to illustrate this fact.

Stoics, Epicureans, Academics, and Peripatetics all held different opinions on the propriety of quoting poetry in a philosophical treatise. On one end of the spectrum were the Epicureans, who were mistrustful of poetry and disinclined to use it, at least in their philosophical arguments.28 At the other end were the Stoics, who believed that pre-philosophical myths contained elements of truth about nature and the universe and that many poets provided valid and authoritative expressions of this truth; this was their justification for quoting frequently and extensively from poetry, which they argued foreshadowed and supported their doctrines.29 Academics and Peripatetics occupied a middle space, recognizing that sparing poetic quotation could add a flourish to their arguments that educated men would find attractive.30

(p.270) These differing attitudes are represented quite faithfully in Cicero’s treatises. Epicurean speakers in Cicero use no poetic quotations, and their references to poetry are sarcastic, mocking those who take it seriously.31 Academic and Peripatetic speakers quote poetry sparingly, and then only as decoration.32 Cicero’s Stoics, on the other hand, are great lovers of poetic quotation. Two of his Stoic arguments, Balbus’ in De Natura Deorum 2 and Quintus’ in the first book of De Divinatione, are filled with poetry, to an extent that feels excessive when compared to poetic quotation at other points in the dialogues. It should thus come as no surprise that quotations from Cicero’s own poetry are clustered in these two books.

If Cicero’s use of poetic quotation in his dialogues is carefully crafted to reflect the use each school made of poetry, then it would be strange if the quotations themselves did not also mirror what he found in his sources for each school. In fact, several parallel passages between Cicero’s dialogues and later philosophical compilations (which clearly draw on a common tradition) show that the poetry Cicero found in his sources influenced the poetry that he quoted in his dialogues.33 But rather than simply translating the poetry he found in his sources, Cicero often made creative substitutions, replacing a Greek passage with what he considered to be its Roman equivalent. For example, where a Greek source discussing Stoic views on causes quotes from Euripides’ Medea (Clem. Alex. Strom. 8.9.26–7 Koltz), Cicero, describing the same topic, draws on Ennius’ Medea instead (Fat. 34–5; cf. Top. 61).34 When Ps-Plutarch, while relaying Chrysippus’ views on the nature of visions, quotes Orestes and Electra describing the Furies in Euripides’ Orestes (Plac. 4.12, cf. Sext. Emp. Math. 7.170, 249), Cicero uses another character afflicted by the Furies to discuss the visions suffered in madness: Alcmaeon in Ennius’ eponymous play (Luc. 52, 89). Likewise, where the consolation to Apollonius attributed to Plutarch draws on Priam’s detailed account of his imminent death from Iliad 22.56–78 (Mor. 113f–14c), Cicero, borrowing from the same tradition, substitutes the description Ennius has Andromache give of Priam’s (p.271) death in his Andromacha (Tusc. 1.85). On other occasions, Cicero himself translates a passage from Euripides (Fin. 2.105 ~ Plut. Mor. 630e; Nat. D. 2.65 ~ Clem. Alex. Strom. 5.14.114 Koltz and Heracl. Alleg. 23; Div. 2.12 ~ Plut. Mor. 399a, 432c) or from Homer (Fin. 5.49 ~ Sext. Emp. Math. 1.42) that he clearly found in his source, perhaps because a Latin version of the poem did not already exist or because he did not have it readily to hand.

But these in situ translations are far less significant than the substitutions that Cicero makes, because it is clear that those substitutions reflect in microcosm the same motivation he claimed for the project as a whole: the wholescale Romanizing of Greek material. And in the case of his poetic quotations, the Romanization has a further significance beyond the mere transfer of cultural material. The quotation of Homer, Hesiod, and the Athenian tragedians in Greek philosophical treatises both reflected and helped reinforce their canonical status. Cicero’s substitution of Roman poets for these classical Greek poets is a sign of his active participation in creating a parallel Roman canon of poetry. The inclusion of his own poetry in this canon is one way that he benefits from its creation, but there are also broader benefits. As I have noted before, the assumption of a basic equivalence between Greek and Latin literature—and the idea that Cicero can himself create Latin literature that is equivalent to Greek works—lies at the heart of his career as an author from beginning to end. The replacement of philosophically appropriate passages from Greek poets with philosophically appropriate passages from Roman poets is perhaps the most simple and straightforward way for him to show that the two languages are fully equivalent in accurately illustrating philosophical truth.35 And the inclusion of his own poetry among this material provides the concrete realization of his position as a linguistic equal to the best poets of Greece.

The motives that drove Cicero to quote his own poetry are most clear in the extensive quotations by Balbus in De Natura Deorum and Quintus in De Divinatione. While Cicero’s poetry is quoted several other times in his philosophical works, all other instances involve the quotation of no more than two lines at a time, and all are placed in the mouth of Marcus Cicero himself, where—in good Academic fashion—they function as short glosses to illustrate a point.36 What Quintus and Balbus do with Cicero’s poetry is something different.

As I have already suggested, Quintus’ and Balbus’ unbridled use of Cicero’s poetry is at its root connected to their philosophical orientation: as (p.272) representatives of Stoic views, they naturally turn to poetry as a source of wisdom that can add authority to their points. Indeed, Cicero is far from the only poet used in the course of their arguments: Ennius, Accius, and Pacuvius are frequently quoted, with Ennius’ tragedies and Annales in the most prominent position.37 Roman comic writers also appear, though with less frequency.38 Both interlocutors also provide some translations from Greek poetry, including Euripides (Nat. D. 2.65 = Frag. 941 Nauck), Homer (Div. 1.52 = Il. 9.363), and a Pythian oracle (Div. 1.81 = App. Proverb. 2.55).39 Some of these quotations are quite lengthy: Quintus quotes passages from Ennius (Div. 1.107–8 = Fr. Ann. 72–91 Goldberg and Manuwald) and Accius (Div. 1.44–5 = Fr. Praetext. 17–38 Warmington) that each total at least twenty lines, while Balbus’ longest citation from another poet is a quotation from Accius comprising sixteen lines in total (Nat. D. 2.89 = Fr. Trag. 381–95 Warmington). But no poet is quoted more prolifically than Cicero himself. Balbus quotes over ninety lines of the Aratea, while Quintus reels off twenty-three from Prognostica, thirteen from the Marius, and an astounding seventy-eight-line continuous quotation from De Consulatu. These are the longest quotations of poetry in Cicero’s philosophical treatises by far, topping the next longest—the translations of Homer and the Greek tragedians in the Tusculan Disputations and De Divinatione—by a comfortable margin.40

Balbus’ and Quintus’ philosophical orientation does not just explain the volume of their poetic quotation, it also explains their choice of source material: Aratus’ Phaenomena was, as I showed in Chapter 1 (section ‘The Phaenomena in Hellenistic Greece’), frequently read and interpreted as a work of Stoic philosophy by later Greek authors. The Stoicizing nature of Aratus’ poem and of subsequent commentary on it clearly did not escape Cicero in his translation of the poem as a young man, and its prominent use in Stoic philosophical treatises must have been equally notable to him as he sought material from which to adapt his philosophical works forty years later.41

Two philosophers in particular must have illustrated for Cicero the way that Aratus could be used to support a Stoic argument: Boethus of Sidon and Posidonius of Apamea. I have already discussed Boethus’ four-book commentary (p.273) on Aratus in Chapter 1 (section ‘The Phaenomena in Hellenistic Greece’), and its likely influence on Cicero’s production of an Aratea that was even more Stoic than the original poem. But Boethus also composed more straightforward philosophical treatises, including an On Nature and On Fate (Diog. Laert. 7.148–9), the latter of which may well have been a resource for De Divinatione. Certainly Boethus was on Cicero’s mind in the composition of that work: he is mentioned twice in it (Div. 1.13, 2.47), both times in reference to his research into weather signs.

Posidonius is, of course, an even more central figure for Cicero’s philosophical thought, so much so that he was once considered the source of nearly every piece of Ciceronian philosophical writing.42 While few would still subscribe to the image of Cicero feverishly copying Posidonius’ thoughts into Latin, it is certainly the case that Cicero flaunts his familiarity with Posidonius in these treatises as a way of burnishing his philosophical credentials; he is mentioned by name in them more often than any other philosopher, often in terms of great familiarity.43 Astronomy was among Posidonius’ many topics of study, and this included a concomitant interest in Aratus; in one of his treatises, for example, Posidonius discussed the causes of weather signs, likely with reference to Aratus (Div. 2.47).44 And in what was probably a separate treatise that is quoted in the scholia to Aratus, On the Comparison of Aratus and Homer Concerning Astronomy, he advanced the oft-repeated idea that Aratus was not really a scientist, since his knowledge of astronomy was based entirely on Eudoxus’ treatise (In Arat. 17.14–21 = F 48b E–K.).45

Nonetheless, Posidonius saw the value of drawing on Aratus’ popularity to illustrate his more technical accounts on astronomy. In his calculation of the size of the earth, preserved in the De Motu Circulari Corporum Caelestium of the astronomer Cleomedes (c.370 CE), Posidonius builds his argument around (p.274) the bright star Canopus (α‎ Carinae), a star that, as he reminds the reader, is not described in the Phaenomena (F 202 E–K). This is, he explains, because the star is completely invisible in Greece, where the poem was composed: it only becomes visible first on the edge of the horizon at Rhodes and then completely in Alexandria. The difference in the star’s visibility across different regions can be used to determine the size of the meridian on which these geographical areas lay, which in turn allows for calculation of the size of the earth. Posidonius’ use of a concrete example from a well-known poet saves his explanation from becoming too obscure—the very charge that Cleomedes levels against Eratosthenes’ similar geographical measurements.

There can thus be little doubt that Cicero encountered the quotation of Aratus in Posidonius’ works, and this may well have inspired him to use his own translation in his dialogues. It is unlikely, however, that Posidonius ever quoted from Aratus as extensively as Cicero does with his own Aratea; in fact, according to Galen, Posidonius criticized the overuse of poetry in philosophical arguments (Hipp. et Plat. 4.5.34 = T 104 E–K), and when Galen himself provides a heuristic for the appropriate amount of quotation, he claims Posidonius would agree with his views:

ἀνεπίφθονον ἤδη καὶ τοὺς πρεσβυτέρους ἐπικαλεῖσθαι μαρτυρήσοντας οὔτε περὶ πραγμάτων ἀδήλων παντάπασιν ἀλλὰ ἤτοι περὶ φαινομένων‎…ὡς καὶ Ποσειδώνιος εἶπεν‎.

Hipp. et Plat. 5.7.84–5 = T 87 E–K

Quotation should not be used at the beginning of an argument, but only when you have adequately proved what you set out to prove, then it is unobjectionable to call in an older generation as witness; and not on subjects utterly obscure but for evident facts…as Posidonius too said.46

Galen’s comments here suggest that while Posidonius likely did quote from the Phaenomena in his treatises to illustrate otherwise abstract principles, he did not do so at length.

Cicero’s familiarity with the works of both Boethus and Posidonius suggests that he was no stranger to the interpretation of Aratus’ poem as a piece of Stoic philosophy, as well as the quotation of it to illustrate points of Stoic philosophy. The nature of the Stoic arguments in De Natura Deorum and De Divinatione similarly invites the quotation of Aratus as an authoritative support. There is no reason not to assume, then, that Cicero found Aratus quoted in the sources for his Stoic arguments in these dialogues.47 But even if this was the case, it is highly unlikely that Aratus was quoted as extensively in these authors as Cicero quotes from his own translation in these two dialogues. Much like Cicero innovated on a large scale in adapting his dialogues from Greek sources, he also must have innovated in the use of his own poetry within these works. Examining the (p.275) poetry’s appearance in both dialogues will give us a sense for Cicero’s novel approach to adapting Greek sources for his own benefit.

The Aratea in De Natura Deorum

De Natura Deorum is the earlier of the two works both in composition and in dramatic date. Set sometime between 77 and 75, it gives an account in three books of the teachings of three philosophical schools on the existence and nature of the gods: in the first book, Gaius Velleius speaks for Epicureanism, which Gaius Cotta, the Academic figure (and host of the gathering), then rebuts. In the second, Quintus Balbus gives an account of Stoic theology, which Cotta again rebuts in the third (and now fragmentary) book.48 These three are the main interlocutors, but as the dialogue opens, the reader learns that Marcus Cicero (who was between twenty-nine and thirty-two at this point in time) was also present for the discussion (1.15). Marcus’ role is restricted mostly to brief statements at the beginning of the dialogue and at the end, when he votes on which of the views he found most persuasive. Nonetheless, his presence does intrude at several points, and it is important to keep him in mind.

Though Cicero’s Aratea is not quoted until rather late in the second book, much of the preceding material prepares the reader for its appearance, and it is clear that Cicero crafted the second book as a whole—and even parts of the first book—to naturalize its quotation. It will be useful, then, to examine the various ways in which the use of the poem is foreshadowed throughout the first and second books.

The first clues can be found in Velleius’ critique of Stoicism in the first book. In order to explain what Epicurean theology is, Velleius begins by explaining what it is not, and ridicules and refutes the theological systems of other philosophical schools in the process. In describing the views of Chrysippus, the third head of the Stoa, he singles out two features of his method for particular censure: first, Chrysippus’ creation of dubious etymological explanations for the gods’ names, and second, his wilful misuse of poetry (1.39–41). Each of these methods will, in fact, also be used by Balbus, and in both cases Cicero’s poetry will play a major role in their use.

Velleius attacks Chrysippus’ use of etymology first, saying that Chrysippus argued ‘that the one men call Jupiter is the aether [the upper air], and Neptune (p.276) is the air that is diffused throughout the sea, and that the one called Ceres is the earth; he explains the names of the rest of the gods in a similar way’.49 Though polemical, Velleius’ account here more or less accurately reflects the importance of etymology for Stoic theology.50 As I have already shown in Chapter 1 (section ‘Aratus’ Phaenomena’), Stoics believed that words were by nature connected to and representative of the thing they signified; they grounded this belief in their conception of pneuma, the divine breath that animates all inert matter. Because both gods and language contain a measure of this divine breath, it stands to reason that a god’s name would be an appropriate linguistic reflection of that god’s essential nature and function, and Stoics therefore believed that the etymologies of the names of the gods could be used to determine each god’s unique manifestation of pneuma.

Velleius clearly finds this sort of etymological evidence for divinity—which he says Chrysippus used in the first book of his work on the gods—dubious at best,51 but he is even more exercised by Chrysippus’ approach in the second book, where he used poetry to the same end:

In secundo autem volt Orphei, Musaei, Hesiodi, Homerique fabellas accommodare ad ea quae ipse primo libro de deis inmortalibus dixerit, ut etiam veterrimi poetae, qui haec ne suspicati quidem sint, Stoici fuisse videantur.

Nat. D. 1.41

[Chrysippus] attempts in the second book [of his Nature of the Gods] to reconcile the fables of Orpheus, Musaeus, Hesiod, and Homer to the things he himself said in his first book on the immortal gods, so that even the hoariest poets appear to have been Stoics, though they never had a single thought of such things.

According to Velleius, Chrysippus’ desire to use the authority of the most revered poets in support of his own views led him to twist the words of classics like Homer and Hesiod so that they seemed to agree with his doctrine. Chrysippus was, in fact, frequently rebuked in antiquity for his abuse of poetry, and although Velleius does not mention the length of his poetic quotations, other sources claim they were excessive: indeed, Diogenes Laertius reports that he once quoted almost the entirety of Euripides’ Medea in a single treatise (7.180).52

(p.277) By having Velleius single out these two elements of Chrysippus’ approach for mockery in the first book of his dialogue, Cicero prepares his reader for Balbus’ approach in the second. Over the course of his argument Balbus, who is described at the outset of the dialogue as a Roman especially well educated in Stoic doctrine, shows that he is also educated in Stoic means of proving this doctrine.53 First he makes extensive use of etymology in order to explain the hidden meanings of the gods’ names, and later, he quotes copiously from a poem that he represents as supporting his argument—not least because the poem itself (Cicero’s Aratea) appears to make use of significant etymologies. It is noteworthy that Balbus draws not on Greek etymologies and poetry, but on Latin ones: he is updating Stoicism for a Roman audience, and he includes Cicero among a group of revered poets who recognized the truth of the cosmos.

Like Chrysippus had done in his On the Nature of the Gods, Balbus uses etymology to support his argument first, then poetry. At 2.60–9, he launches into a discussion about the true sources from which the gods’ names derive, and argues that these names were chosen to reflect the benevolent actions each god confers upon humanity.54 Corn and wine, for example, are called Ceres and Liber after the gods from whom they originated, while Fides, Mens, and other abstract but useful forces are divinities whose names are a direct reflection of the benefits they provide (2.61–2). Similarly, particularly renowned and benevolent human beings are now worshipped as gods; their names simply point to the human beings who once carried them (2.62). Last of all, Balbus considers the traditional Olympian pantheon, the gods who ‘supplied the poets with fairy tales and crammed human life full of every superstition’ (2.63: fabulas poetis suppeditaverunt, hominum autem vitam superstitione omni referserunt): what Balbus will show is that, in fact, the etymologies of their names reveal that these deities have a more sober and providential role than popular tales might suggest.

At this point Balbus begins to supply the sort of etymological puns alluded to by Velleius—with the one important distinction being that his etymologies largely focus on Roman, not Greek, names for the gods.55 Indeed, the few (p.278) Greek names that are mentioned are those that can be shown to have parallel meanings with their Roman counterparts: thus, both the Greek Kronos (derived from χρόνος‎, ‘time’) and the Roman Saturn (derived from saturari annis, ‘to be saturated with years’) have to do with time, and Dis and Pluto are both related to each language’s words for wealth.56 But for the most part Balbus’ goal is to demonstrate that the names of Roman deities reflect the provident component of the natural world that they represent: Jupiter is the iuvans pater (‘helping father’), Ceres ‘bears’ the earth’s crops (gerere), and so on (2.64–9).57 By showing that the Roman names for these gods have such significant meanings, Balbus demonstrates that the Latin language is fully equal to Greek in its accurate representation of the providence immanent in the world, and the linguistic equality that these etymologies prove helps lay the foundation for his quotation of a Roman poem in place of a Greek one later in his argument. Furthermore, if both languages reveal divine providence, then Balbus has twice as much proof for the existence of divinity and for its provident and benevolent nature than Chrysippus had had.

Equally strong proof for divine providence can be found in nature itself, particularly in the stars, whose orderly motion throughout the sky marks the change of the seasons.58 Indeed, from the very beginning of his argument, Balbus foregrounds the importance of the stars to the case he is making for intelligent design, and the recursion of astronomy throughout the early parts of the book paves the way for his later use of a poem on the topic. At the very outset, he asks ‘when we look up at the sky and contemplate the celestial bodies, what could be more obvious and transparent than that there is some divine power of outstanding intelligence by which they are governed?’59 As he explains shortly after this, astronomy is not just evidence for intelligent design, it is the very best evidence:

Quartam causam esse eamque vel maximam aequabilitatem motus <constantissimamque> conversionem caeli, solis, lunae, siderumque omnium distinctionem, (p.279) utilitatem, pulchritudinem, ordinem, quarum rerum aspectus ipse satis indicaret non esse ea fortuita.

Nat. D. 2.15

The fourth cause [of the innate belief that gods exist] and the greatest is the uniform motion and exceedingly regular rotation of the heavens: observing the individuality, usefulness, beauty, and arrangement of the sun, the moon, and all the stars is enough to indicate that they are not the products of chance.

Later, Balbus demonstrates that the stars are themselves divine; shaped out of the purest and most mobile part of fire and aether, they obviously contain a large share of pneuma and thus possess keen sensation and intelligence (2.39–42). A clear sign of their divinity is found in the ‘order and regularity’ (ordo et constantia) of their movement across the sky, a movement that would not be possible without divine design (2.43–4).

Part of what makes the stars’ movement such a striking witness to the existence of the gods is the fact that this motion is circular, the shape that excels all others for its harmony and perfection (2.47); furthermore, the sun, moon, planets, and stars complete this circular course in perfect regularity, bringing benefit to mankind with their revolutions (2.49–56). Each heavenly body is providential in its own way: the sun and moon provide the means for growth and nourishment of all plant and animal life, the stars reveal divine power and intelligence with the ‘marvellous and incredible constancy’ of their motion (cum admirabili incredibilique constantia, 2.55), and even the planets, which many have falsely believed to wander throughout the sky, actually complete a fixed orbit over a long period of time.60 All evince the provision of the gods for humanity (2.58).

Balbus proceeds to give several proofs of the universe’s intelligent design, but in the end decides that a description of nature’s beauty and adornment will prove more persuasive to his audience. As he moves from philosophical argument to aesthetic catalogue, he once again highlights astronomy in particular, saying that the ‘very invariable’ (2.97: constantissime) motion of the heavens leaves us in no doubt that they were ordered by divine reason (2.98). He will save such sublimities for the end of his description of the universe’s beauty, he decides, and begin with the first and lowest elements instead. He then begins to describe, in heightened, expressive language, the many beautiful features of the earth and sea, animals and birds and humankind, before finally turning to the celestial realm of the sun, moon, and planets (p.280) (2.98–104).61 But when he reaches the outermost sphere of the universe, the sphere of the fixed stars, he decides to take up the expressive language of another, and announces his intention to use the Aratea for this purpose.

As I showed in Chapter 1, the poem is especially well suited to the sort of argument Balbus wants to make with it. Indeed, any ancient thinker who wished to use the regular movement of the heavens as proof of the existence of god would undoubtedly have thought of the Phaenomena, whose proem announces that Zeus set the stars in the sky to be orderly, well-defined signs for mankind (Phaen. 10–14), and which memorably evokes the sense of wonder at the movement of heavenly bodies (e.g. Phaen. 469–79) that Balbus claims as strong proof of intelligent design (Nat. D. 2.75).

Balbus, therefore, was probably not the first Stoic to use Aratus’ poem as an illustration for this sort of argument. Indeed, evidence for the Phaenomena being used in this way prior to Cicero can be found in a chapter from the first-century CE Placita of the doxographer Aetius entitled ‘how men acquire knowledge of the gods’ (Plac. 1.6, Πόθεν ἔννοιαν ἔσχον θεῶν ἄνθρωποι‎).62 After describing how the Stoics define the nature of god, Aetius explains that knowledge of the divine was first acquired from the beauty of the universe, since everyone can recognize that nothing so beautiful could have been formed by chance. This beauty is typified by the world’s circular shape, colour, and size, and by the stars. The last of these items—the beauty of the stars—is then illustrated poetically, first by Aratus’ description of the zodiac (Phaen. 545–9), then by two tragic verses purportedly from Euripides.63

The similarities between Aetius’ and Balbus’ accounts suggest a common source, and it is likely that both derive from Posidonius.64 In all likelihood, then, Posidonius had used the Phaenomena in the same sort of argument and for the same reasons that Cicero has Balbus use the Aratea. This does not, of course, mean that Cicero went through Posidonius’ work systematically replacing Greek quotations of Aratus with the Aratea; after all, the lines of his translation that correspond to the passage in Aetius (Arat. 320–31) appear (p.281) nowhere in Balbus’ argument, and his quotation of the poem is significantly longer than what is found in Aetius—indeed, it is longer than the amount of quotation of which Posidonius approved, as I have already shown. In other words, while Posidonius may have given Cicero the inspiration for Balbus to use the Aratea to make his case, what Cicero did with the poem was entirely his own.

Indeed, when Balbus first begins to quote from the poem as part of his description of the universe’s beauty, the uniqueness of the situation is made clear:

‘Sequitur stellarum inerrantium maxima multitudo, quarum ita descripta distinctio est ut ex notarum figurarum similitudine nomina invenerint.’ atque hoc loco me intuens, ‘utar,’ inquit, ‘carminibus Arateis, quae a te admodum adulescentulo conversa ita me delectant, quia Latina sunt, ut multa ex iis memoria teneam.’

Nat. D. 2.104

‘There is also a great multitude of fixed stars that are so clearly and distinctly arranged that they have acquired their names from their similarity to recognizable figures.’ And at this point looking at me he said, ‘I will use the verses of Aratus translated by you when you were quite a young man, which, because they are in Latin, delight me so much that I retain many of them in my memory.’

Cicero’s aside here—me intuens, ‘looking at me’—is one of the few times in the dialogue that the reader is reminded of Marcus’ presence. While to some extent the intrusion is practical (since it identifies the author of the poetry), Cicero is also making his reader fully aware of the oddness of what is to come: a young poet who will silently listen to his own poetry conscripted into a Stoic argument.

‘Many’ is indeed the correct term for the number of verses Balbus can supposedly recall extempore; while he has quoted from other Roman poets throughout his argument (including Ennius, Pacuvius, Accius, and Terence), it is only here that, like his forebear Chrysippus, his poetic quotation becomes extensive.65 Between 2.104 and 2.114 he quotes or paraphrases over ninety lines of the Aratea that describe forty-five constellations. His selections from the poem are wide-ranging, beginning with a quotation from the very opening of the work (fr. III ~ Phaen. 19–20) and ending with a passage from the close of the poem’s astronomical section (Arat. 222 ~ Phaen. 450), right before the topic moves to a discussion of the simultaneous risings and settings of the constellations.66

(p.282) Elsewhere in Cicero’s philosophical works, such lengthy quotations tend to be given in one large chunk.67 Balbus, on the other hand, quotes in lines and half-lines that are linked together and contextualized by what is essentially a running commentary. For the most part this is a matter of practicality: since Balbus draws on many different sections of the poem, he finds it necessary to describe each constellation’s location and/or its spatial relationship to the next constellation he will mention. Nonetheless, on a few occasions he does insert comments of a scholarly or philosophical nature, and these are worth noting. The remarks with which Balbus begins and ends his first quotation are particularly revealing of his agenda:

Ergo, ut oculis adsidue videmus, sine ulla mutatione aut varietate

  • ‘Cetera labuntur celeri caelestia motu
  • cum caeloque simul noctesque diesque feruntur,’

quorum contemplatione nullius expleri potest animus naturae constantiam videre cupientis.

Nat. D. 2.104–5 (= Arat. fr. III)

Therefore, as we see continuously with our own eyes, without any alteration or variety

  • The other celestial bodies glide in a swift motion
  • carried all day and night together with the sky,

and no one who desires to see the invariability of nature can get his fill of the contemplation of them.68

A similar point is made in his introductory remarks to fr. XXII on the Twins, the Crab, and Leo: ‘See how the constellations are marked out in such a way that divine ingenuity is apparent in their arrangement.’69 All three comments evince an explicitly Stoic way of thinking about the stars, emphasizing that their regularity is proof of intelligent design. Balbus’ use of the term constantia (‘invariability’) in conjunction with the poem is especially notable, since this is the term he has used throughout his argument to claim that the orderly movement of the stars is strong evidence for the existence of god. By claiming (p.283) that the poem depicts this constantia as well, Balbus turns it into a key piece of evidence in favour of intelligent design and, hence, Stoic cosmology.

Balbus also lingers over moments of etymological significance in the poem, which, as we saw in Chapter 1 (section ‘Cicero’s Aratea’), was one of the themes of the translation upon which Cicero lavished the most care. At 2.108–9, he quotes two passages from the Aratea that include examples of fairly straightforward etymological wordplay: the constellation Engonasin has acquired his name ‘because he is carried along resting on his knees’ (genibus quia nixa feratur), while Ophiuchus’ name is calqued as ‘the one who holds the Snake’ (continent Anguem).70 This is followed by a far more subtle wordplay at 2.109 that glosses the name of the constellation Bootes (‘Oxen-Driver’) with the phrase ‘because it is as if he drives the Bear yoked to a waggon pole’ (quod quasi temoni adiunctam prae se quatit Arctum): the use of Temo, an alternate Latin name for the constellation the Waggon, makes it possible to allude to the etymology of both names for the figure (literally, Bear-Guarder and Oxen-Driver).71

This interest in the poem’s use of etymology culminates at 2.111, where Balbus discusses the description of the Hyades from the Aratea. But this time, instead of quoting an etymology embedded within the Aratea itself, he provides one of his own:

Eius caput stellis conspersum est frequentibus;

  • ‘has Graeci stellas Hyadas vocitare suerunt’

a pluendo (ὕειν‎ enim est pluere), nostri imperite Suculas, quasi a subus essent non ab imbribus nominatae.

(Nat. D. 2.111 = Aratea fr. XXVIII)

The head [of Taurus] is sprinkled with numerous stars: ‘these stars the Greeks are accustomed to call the Hyades’. [Their name comes] from rain—for ὕειν‎ means ‘to rain’—though our countrymen ignorantly call them the Piglets, as if they were named from pigs and not from rain.

This constellation is, in fact, especially useful for Balbus’ argument. By emphasizing that its name is derived from the Greek verb for rain (ὕειν‎), he can show that constellations as well as gods get their names from the benefits they provide for mankind: the Hyades mark the beginning of the rainy season, which nourishes the crops that have been planted and allows them to grow.72 (p.284) Understanding the origin of the constellation’s name thus means recognizing the providence of the universe and its divine creator.

Here, however, Balbus has extrapolated from what Cicero wrote in the poem to an external resource. All Cicero had done in the verse Balbus quotes was draw attention to the constellation’s Greek name: it is Balbus who provides the meaning of that name. He does so, clearly, by quoting from a commentary on the Phaenomena; this etymology is found in the S scholia to Aratus among other possible options for the derivation of the Hyades’ name, they are said to have been named ‘either because they lie in the shape of the letter Upsilon, or because they are able to make it rain (ὑετόν‎), or because the theta has fallen off of Thyades.’73 The scholion’s insistence that the name reflects the constellation’s value as a weather sign is itself Stoicizing, making it the perfect complement to Balbus’ commentary. By inserting into Balbus’ mouth a quotation from (presumably) the same commentary he used when composing the Aratea, Cicero is essentially writing his own Stoic commentary for his translation.

But is Balbus correct to interpret the Aratea in this way? After all, in the first book Velleius warned of the Stoic proclivity for co-opting the authority of non-Stoic poets in order to prop up their own teachings. And Cicero himself had opened his authorial preface to the dialogue with a gesture of allegiance to the Academy (1.1), particularly Carneades’ method of refutation (1.4), for which he says he has been subjected to criticism by curious readers who want to know what opinions he himself holds on these philosophical topics (1.6). He chastises these readers, decrying the fact that in many philosophical schools, authority takes the place of reasoned argument (1.10), and reaffirming his commitment to the use of in utramque partem debate in order to determine the most probable doctrine (1.11–12).74 The preface to De Natura Deorum, then, is explicitly designed to turn the reader away from the question of Cicero’s own beliefs and from any reliance on his authority to approve or disprove a certain theological system.

But the fact remains that the Aratea, just like the poem that it translates, does show a clear affinity with Stoicism.75 The universe of the Aratea is, after all, a stable one, where, for example, none of the Pleiades have ever disappeared from the sky, but ‘for no purpose, thoughtlessly, without any reason, (p.285) are there commonly said to be seven’ (Arat. 32–3: sed frustra, temere a vulgo, ratione sine ulla/septem dicier). Anyone who looks up at night will see the constellations ‘lighting up the sky with their regular orbit’ (Arat. 224: legitimo cernes caelum lustrantia cursu) and the poet’s task is to relate ‘the [heavenly bodies] that always revolve fixed in a certain orbit’ (Arat. 235–6: quae semper certo [e]volvuntur in orbe/fixa). The names of these constellations reflect their provident design, because ‘that guardian of the stars marked them according to reason, deeming these celestial signs worthy of a true name’.76 And indeed, Cicero provides several of these true names in the poem, including ones for the zodiac (‘sign-bearer’, signifer, which he has Balbus repeat at 2.53) and for the Pleiades (Vergiliae). Aratus’ poem was no doubt the most famous representation in antiquity of an intelligently designed universe, and for Roman readers of the late republic, Cicero’s translation of it had also acquired this status well before the composition of De Natura Deorum, at least judging by Lucretius’ in-depth polemic with Cicero’s translation.77 Thus, while the length of Balbus’ quotation from the Aratea may be on the model of Chrysippus, it is less likely that he is reading Stoic meaning into Cicero’s poem that was not there before.78 It is far more likely, in fact, that Cicero saw the poem—which itself puts Latin on a par with Greek in accurately depicting the universe—as the perfect complement not just to Balbus’ argument, which rests on a similar foundation, but to his own philosophical project of Latinizing Greek wisdom more broadly.

Moreover, Cicero seems to betray a personal preference for Stoic cosmology in the surprising end to the dialogue, where Marcus declines to agree with Cotta’s Academic scepticism, stating that ‘to me [the argument] of Balbus seemed nearer to a resemblance of the truth’.79 It is an unexpected decision, given Cicero’s emphasis in the preface on his continued allegiance to the Academy. Yet Cotta lays the groundwork for it only a few paragraphs earlier, when he likens Stoic theology to the vengeful feuds of myth delighted in by poets: ‘whether poets have corrupted the Stoics, or whether the Stoics have given authority to the poets I cannot easily say,’ he remarks, ‘since both speak of monsters and shameful deeds’.80 Balbus is no doubt a Stoic that has given a great deal of authority to a poet, though his poetry was on loftier topics than (p.286) these. Nonetheless, the close link Cotta makes here between Stoics and poets foreshadows the final remarks of the dialogue and offers a potential explanation for Marcus’ agreement with Balbus. This agreement, coming from an avowed Academic, also serves to lend even greater weight to the role Cicero’s poetry plays in the dialogue, suggesting that Balbus was not wrong to read the Aratea as in sympathy with Stoicism. In fact, the suggestion is that its description of the cosmos was so convincing that it even convinced a sceptical Academic of its verisimilitude. Like a true classical text, then, the Aratea possesses the authority to make or break an argument, even if Cicero, as the Academic author of the work, must openly deny such authority.

Cicero’s Poetry in De Divinatione

De Divinatione is explicitly marked as a continuation of the topic first raised in De Natura Deorum, and it is likely that Cicero had in mind from the beginning a trilogy of theological works: a De Natura Deorum followed by De Divinatione and De Fato.81 In fact, De Divinatione is, in many ways, a work about reading and responding to De Natura Deorum. In the authorial preface, after listing the preponderance of philosophical arguments in favour of divination (mostly Stoic in nature), Cicero says that since he is himself in doubt as to whether divination exists, he has decided to ‘compare argument with argument, just as I did in those three books I wrote on the nature of the gods’.82 The dialogue proper, meanwhile, opens with Quintus Cicero informing his brother that he has recently completed a close reading of Book 3 of De Natura Deorum, which sparked his interest in discussing an aspect of theology left out of that work: namely, divination.83

In structure, too, the dialogue resembles De Natura Deorum, with Quintus giving an essentially Stoic argument in favour of divination in the first book, which Marcus rebuts with Academic scepticism in the second. Though Quintus is not an official proponent of Stoicism like Balbus was, his argument is clearly indebted to Stoic defences of divination, and Posidonius’ Περὶ‎ μαντικῆς‎ (mentioned in the prologue at Div. 1.6) has long been thought to be a source (p.287) for the dialogue.84 Cicero draws attention to the philosophical affiliation of Quintus’ argument several times, most pointedly at its conclusion, when Marcus says, ‘Truly, Quintus, you have defended the opinion of the Stoics carefully and in the way a Stoic would’ (Div. 2.8: adcurate tu quidem…Quinte, et Stoice Stoicorum sententiam defendisti).85

Marcus’ compliment that Quintus has argued carefully is perhaps overly flattering, since at times it seems as if there is nothing careful about Quintus’ argument. Though an underlying order has been detected to it, for the most part Quintus’ strategy in this book is, as he admits at its outset, to accumulate anecdotes that prove the existence of divination without delving into its underlying causes (Div. 1.12).86 In fact, this is probably what Marcus means when he says Quintus has made his argument in the manner of a Stoic: Diogenes Laertius reports that many Stoics, including Posidonius, attempted to prove that divination was a techne ‘on the basis of certain results’ (7.149: διά τινας ἐκβάσεις‎), which suggests that it was an accepted Stoic practice to provide examples that focused on the successful outcome of various acts of divination without grappling with what had caused them.87

Much like Aratus’ star catalogue would have been a natural choice for an author who wished to illustrate Stoic cosmology, his Weather Signs would no doubt have come to the mind of someone looking for examples of portents whose results were certain even if their causes might be unclear. Though Aratus’ catalogue of events foretelling the weather does not constitute divination per se, it does serve as an illustration of the main argument used to justify its existence, namely the interconnectedness of seemingly random events. It is quite possible, therefore, that Cicero found the latter part of Aratus’ poem quoted in Stoic treatises on divination—perhaps, for example, in the On Fate of Boethus—much as he had found the first part in (p.288) Stoic treatises on cosmology.88 Even if this were not the case, Cicero surely would have liked the neatness of quoting the first part of his translation in his first work on theology and the second part in his second one.89

The dramatic circumstances of De Divinatione, however, differ significantly from those of De Natura Deorum, giving the appearance of the poem quite a different valence in this dialogue. The dramatic date of De Divinatione is roughly contemporaneous with its composition in 44, and Marcus Cicero is no longer a young senator relegated to an unspeaking role amongst his elders, but an old man discussing philosophy with his younger brother in a private moment on his estate.90 Quintus’ argument is in fact designed to take advantage of these circumstances: it is tailored specifically to appeal to Marcus, focusing only on those examples of divination deemed most persuasive to him. It is clear that Quintus believes this will be an easy task, and that part of his reason for thinking so is his familiarity with his brother’s literary output. At the outset of the discussion, he makes much of Marcus’ agreement with Balbus’ argument at the end of De Natura Deorum, noting that if the gods exist (as Marcus’ acceptance of Balbus’ position suggests), then divination does, and vice versa (1.9), and he caps his argument by once again stating that agreement with Balbus’ views requires assent to the existence of divination (1.117). And because Marcus’ writings (including De Natura Deorum) and his career seem to evince a personal belief in divination, Quintus draws extensively on both.91 He quotes from the Prognostica (1.13–15), De Consulatu (1.17–22), and Marius (1.106), as well as from De Natura Deorum (1.33, 1.93). He reminds his brother of the trustworthiness of auguries he has taken at various points in his career (1.25), and says that it is his duty as an augur to defend the art (1.105). He recalls several occasions on which Marcus himself observed the success of divination, whether through dreams (1.59), prophecy (1.68), or omens (1.72). The thrust of his approach is summed up in the conclusion of his quotation of De Consulatu: ‘Can you, then, bring yourself to contradict these arguments I make about divination when you have both done (p.289) what you have done and when you have so carefully written the words that I proclaim?’92 Quintus’ use here of the word pronuntiavi to describe the act of quotation gives Marcus’ words (and actions) the weight of an official ruling by a magistrate, or indeed of an augur.93

The Prognostica is the first piece of evidence Quintus uses to make the case for divination, and the poem in fact serves as his justification for enumerating the effects of divination without discussing their causes.94 He opens his argument by saying that just as we may not know why certain herbs work medicinally, but can still observe their efficacy, likewise we can clearly see the effects of divination without needing to know what caused it (1.12–13). Then he turns to a related type of event:

Age ea quae, quamquam ex alio genere sunt, tamen divinationi sunt similiora, videamus:

  • Atque etiam ventos praemonstrat saepe futuros
  • inflatum mare, cum subito penitusque tumescit,
  • saxaque cana salis niveo spumata liquore
  • tristificas certant Neptuno reddere voces,
  • aut densus stridor cum celso e vertice montis
  • ortus adaugescit scopulorum saepe repulsus.

atque his rerum praesensionibus Prognostica tua referta sunt. Quis igitur elicere causas praesensionum potest?

Div. 1.13 (= Progn. fr. III)

Come, let us examine things which, although from another category, are nevertheless very similar to divination:

  • The puffed-up sea also often gives forewarning of coming wind
  • when it suddenly begins to swell from within,
  • and the foamy rocks, frothy with snowy brine
  • vie to lift up their gloomy voices to Neptune,
  • or when arising from the high mountain peak a continuous whistle
  • grows stronger as it reverberates off the cliffs again and again.

Your Prognostica is full of these sorts of forewarnings. Yet who can ascertain the causes of such forewarnings?95

(p.290) The implication of Quintus’ tag is clear: if Marcus himself was willing in the Prognostica to record so many examples of weather signs—which, like divination, warn of events to come but have unclear causes—then he should be congenial to Quintus doing the same. Having signalled his belief that the poem justifies his approach, Quintus quotes several further passages from a small subsection of the Weather Signs concerned with signs of wind and rain (~ Phaen. 909–87), each time providing a similar piece of commentary that emphasizes human inability to know the causes of these signs.96

Though not as tightly organized as Balbus’ quotation from the poem, the verses Quintus quotes do have an internal logic. His first quotation, on signs of wind from the sea, serves as a prologue of sorts for the quotations that follow. In the next chapter of the dialogue (Div. 1.14), he combines two different passages, both concerned with bird signs for wind and rain, into one continuous quote: signs of wind from herons (Progn. fr. III) and signs of rain from an acredula and crows (Progn. fr. IV).97 He bookends this continuous quotation with further disavowal of the causes that produce these effects: ‘But who can reliably say why these things occur?’ (Div. 1.14: illa vero cur eveniat quis probabiliter dixerit?) and ‘hardly ever do we see these signs to be false, but we still do not see why this is so’ (Div. 1.15: videmus haec signa numquam fere mentientia nec tamen cur ita fiat videmus).

Quintus then gives further signs of rain from animals, first quoting from a passage concerned with the ability of frogs to offer a premonition of a storm (Progn. fr. IV) and noting that while no one would suspect frogs could have such foresight, yet they have ‘by nature some power of foretelling something, clear enough in itself, but too dark for human understanding’ (Div. 1.15: vis et natura quaedam significans aliquid per se ipsa satis certa, cognitioni autem hominum obscurior). His next quotation also focuses on an animal’s foreknowledge of rain, oxen in this case (Progn. fr. IV), and once again, ‘I do not ask why, since I understand what happens’ (Div. 1.15: non quaero cur, quoniam quid eveniat intellego).

Quintus’ final quotation from the poem is drawn from a later section concerned with signs that presage the changing of the seasons, and takes as its example the mastic tree, whose blooming three times throughout the year signifies the three seasons in which one can plough (Progn. fr. V). While this is unrelated to the signs of rain and wind he has drawn on up to this point, it is thematically connected to what follows at 1.16 (a discussion of the healing properties of various plants), and serves as a bridge between the two sections. (p.291) He concludes his quotation of this final passage from the translation by reaffirming the importance of the poem to his larger argument:

Ne hoc quidem quaero, cur haec arbor una ter floreat aut cur arandi maturitatem ad signum floris accommodet; hoc sum contentus, quod, etiamsi cur quidque fiat, ignorem, quid fiat intellego. pro omni igitur divinatione idem quod pro rebus iis quas commemoravi respondebo.

Div. 1.16

I do not even ask why this tree alone blooms three times or why it makes the sign of its flowering conform to the proper time for ploughing. I am content with this: that even if I am ignorant as to why each thing occurs, I know that it does occur. Therefore I will give the same answer for every kind of divination as I did for these things that I just mentioned.

This final remark underlines what Quintus has been suggesting throughout his quotation from the Prognostica: namely, that the poem—which was translated by Marcus, who must therefore approve of its content—provides the justification for the heuristic he will apply throughout the first book of De Divinatione. His comments here make it very clear that he considers the poem to have successfully authorized his decision not to probe the causes of divination.

In fact, these particular sections of the poem were clearly chosen because they best reinforce Quintus’ point: it is no accident that all of the passages he quotes are concerned with terrestrial phenomena rather than celestial ones. As Quintus himself admits immediately after his first quotation from the poem, the weather signs in Aratus that pertain to the sea and sky have recently been explained by Boethus of Sidon (1.13). But the sorts of weather signs whose causes are known are of no help in authorizing Quintus’ larger argument, so he focuses only on those signs whose causes have not yet been scientifically determined.

Quintus’ interest in Cicero’s translation of Aratus unites him with Balbus, Cicero’s other proponent of Stoic theology. Both, for example, quote the poem alongside extensive quotations from other Roman poetry (in Quintus’ case, the total is over two hundred and fifty lines).98 And though Quintus’ quotation from the translation is not as lengthy as Balbus’ (only twenty-three lines are quoted in all), he too has appropriated the poem to reinforce an important aspect of Stoicizing theology, and he too indicates this fact with linguistic echoes that link the poetry to his broader argument: much like Balbus had insisted that the Aratea shows the constantia (‘invariability’) of the heavens, Quintus claims that the poem is a record of praesensiones (‘foretellings’) without clear causes. But where Balbus’ appeal to Marcus’ Stoic tendencies was implicit, Quintus makes much of his brother’s potential implication in a (p.292) Stoic worldview. The poem is thus perfectly suited to the larger thrust of Quintus’ argument, which is designed to appeal to his brother’s sensibilities. While it is therefore possible that Cicero came across Aratus’ Weather Signs quoted in his Stoic sources, his use of the poem in De Divinatione—much like its use in De Natura Deorum—is perfectly tailored to suit the structure he has created for the dialogue.

Quintus’ quotations from Cicero’s other works of poetry are equally suited to his strategy of targeted persuasion. Just like he uses his brother’s translation of Aratus to authorize his larger argument, so too he introduces his lengthy quotations from De Consulatu and Marius by appealing to Marcus’ authority. ‘But what authority or witness can I better employ than you?’ (Div. 1.17: sed quo potius utar aut auctore aut teste quam te?) he asks right before quoting a long section of the De Consulatu that lists the many prodigies that portended Catiline’s conspiracy and their causes (fr. II); the passage he quotes shares many common themes with the Aratea and generally evinces a Stoic outlook.99 When he borrows a description from Marcus’ Marius of Marius’ reading of a bird sign that foretold his return to Rome, he says he does so ‘to employ you above all others as my authority’ (Div. 1.106: ut utar potissumum auctore te).100

Poetry is not the only type of Marcus’ writing that Quintus takes up. Twice in the course of his argument, he refers to examples of divination included in the second book of De Natura Deorum. The first is an augury received by Tiberius Gracchus that Balbus relates (Nat. D. 2.10–11), which Quintus introduces by saying ‘doesn’t that account you gave about Tiberius Gracchus prove that augury and haruspicy are arts?’ (Div. 1.33: quod scriptum apud te est de Ti. Graccho, nonne et augurum et haruspicum conprobat disciplinam?). Later he reproduces one of Balbus’ etymological accounts—this one designed to show that the names given to divination by early Romans show that they recognized the art’s accuracy (Nat. D. 2.7)—by saying ‘the words themselves skilfully chosen by our ancestors make clear the force of these [types of divination], as you are accustomed to say’ (Div. 1.93: quorum quidem vim, ut tu soles dicere, verba ipsa prudenter a maioribus posita declarant).

It is significant that both of the examples that Quintus takes from De Natura Deorum are pieces of evidence given not by ‘Marcus’ or Cicero in propria persona, but by his interlocutor Balbus, who is explicitly making a case for Stoic theology. It is also significant that Quintus apparently neglected to read the work’s preface, in which Cicero specifically warned his readers against (p.293) reading his philosophical works in the hope of uncovering what he himself believes. Quintus’ effacement of the distinction between a dialogue’s interlocutors and its author in De Divinatione is similar to the way he acts in De Legibus, where he appears to misunderstand an allusion by Marcus to Plato (Leg. 2.17), and it is possible that in both cases, the reader is meant to infer that he simply does not fully appreciate how a dialogue works.101 He would have been wiser, indeed, to choose the Marcus of De Legibus as his exemplum, since there Marcus’ approval of divination is about as full-throated as it can be (though of course this option was not available to him, since the catalogue of Cicero’s philosophical works at Div. 2.1–4 makes it clear that that dialogue was unpublished at this point).102 But it is equally likely that Quintus is meant to be seen as wilfully ignoring Cicero’s careful dialogic structures in order to continue to base his argument on his brother as an authority for the existence of divination, because he thinks this strategy has the best chance of convincing Marcus to concede.

Either way, Quintus has apparently misread his brother’s commitment to the Academy and its disavowal of authority, and Cicero underscores this in the work’s second book, where Marcus vigorously rebuts the existence of divination. In his rebuttal, he directly addresses Quintus’ attempts to persuade him through his own former actions and writings, focusing particularly on the quotation of De Consulatu. As part of his argument that the striking of a thunderbolt has no prophetic meaning, he quotes back at Quintus three lines from the poem (Div. 2.45 = Cons. fr. II.36–8). He then adds that by quoting the poem, Quintus was ‘press[ing] upon me my own verses’ (2.45: sed urges me meis versibus), and summarizes his brother’s approach accordingly:

‘Tu igitur animum induces’ (sic enim mecum agebas) ‘causam istam et contra facta tua et contra scripta defendere?’ frater es; eo vereor. verum quid tibi hic tandem nocet? resne, quae talis est, an ego, qui verum explicari volo? itaque nihil contra dico, a te rationem totius haruspicinae peto.

Div. 2.46

‘Will you therefore persuade yourself’ (this is how you pleaded with me) ‘to defend that cause of yours against your own deeds and writings?’ You are my brother; on that account I am afraid to. But what is it that really offends you in this situation? Is it the sort of subject that it is, or is it me, because I want the truth (p.294) to be set forth? So I will say nothing in response to your charge: rather I will seek from you an account of the entirety of divination.

Just as Quintus had been speaking in his brother’s voice at many points in the first book, here Marcus speaks briefly in Quintus’ voice, closely echoing the comments he had made when closing his quotation of De Consulatu at 1.22. Yet instead of using this fraternal impersonation in order to convince his brother of his argument, Marcus uses it to reject this persuasive tactic—even if their relationship colours Marcus’ response too, making him less willing to refute his interlocutor. Rather than explain Quintus’ apparent misinterpretation of his works, he attempts to turn Quintus back to the real subject at hand, which is, in a manner reminiscent of Socrates, providing a logical explanation and account for divination rather than simply listing examples of its success.

Marcus then mentions Quintus’ use of the Prognostica, saying that to his mind it did not help his argument (2.46–7), since even Quintus admitted that the poem’s weather signs do not count as divination (2.14). But there is another reason its quotation was inappropriate, and this is that both Boethus and Posidonius have made real inroads into understanding the causes of weather signs. Even those signs whose causes have not yet been discovered are clearly still capable of being scientifically studied; therefore the poem is useless in authorizing Quintus’ accumulation of results without explaining their causes. The same would have been true of the portents that accompanied Catiline’s conspiracy recorded in De Consulatu, if only they recurred regularly, allowing for further observation in the way that weather signs do.103

By constructing De Divinatione in such a way that compels Marcus to argue against the use of his own poetry, Cicero gives himself the opportunity to show off his bona fides as an Academic sceptic who rejects all arguments from authority, even when the authority is his own.104 At the same time, he willingly exploits the ambiguities in authority that come from his dual position as interlocutor and author. And in fact, earlier in his argument, he himself drew on his own writings, in this case recalling examples of painful deaths that he recounted in his Consolatio (2.22). Does this mean, then, that Quintus’ methods were not wrong, merely the conclusions he drew from them? Marcus’ (p.295) position is further muddled at the very end of De Divinatione, when, as he wraps up his argument against divination, he speaks eloquently of tearing its sort of superstition out by the root (2.148). This was also the goal, he says, of De Natura Deorum. Here, Marcus must be referring to the arguments of Cotta, who played the role of the Academic sceptic in that work.105 Quintus, in other words, might have believed that the comments made in that work by Balbus reflected Marcus’ beliefs, but Marcus himself, as the author, authoritatively corrects this belief, in the process seemingly confirming the personal views he had chastised others for wanting to know.

But not so fast. Rooting out superstition, Marcus continues, is not the same as dispensing with religion. Furthermore, ‘the beauty of the universe and the order of the heavens compel me to confess that there is some outstanding and eternal nature, and that it must be esteemed and admired by the human race’.106 Once again we are reminded of De Natura Deorum, but this time the interlocutor being recalled is not Cotta, but Marcus himself. Here, at the end of De Divinatione, Marcus explains the reasoning that led him as a young man to agree with Balbus at the end of De Natura Deorum. And this, in turn, suggests that sometimes Cicero’s poetry can be used to correctly intuit his views: Balbus recognized that the Aratea was the work of a young man who accepted the tenets of Stoic cosmology, and his quotation of it compelled Marcus to assent to Stoic theology as the most probable of systems.

Marcus is not done yet, however. First he suggested that his position is closer to Cotta’s in De Natura Deorum, then to Balbus’ (and by extension, to the viewpoint found in his own poetry). But a few lines later, he ends the dialogue with an appeal to the methods of the Academy:

Cum autem proprium sit Academiae iudicium suum nullum interponere, ea probare quae simillima veri videantur, conferre causas et quid in quamque sententiam dici possit expromere, nulla adhibita sua auctoritate iudicium audientium relinquere integrum ac liberum, tenebimus hanc consuetudinem a Socrate traditam eaque inter nos, si tibi, Quinte frater, placebit, quam saepissime utemur.

Div. 2.150

But because it is characteristic of the Academy to put forward no judgement of its own, to approve what seems closest to the truth, to compare arguments, to draw out whatever can be said for each opinion, and—without using any authority of its own—to leave the judgement of its listeners untouched and free, I will stick to this custom, handed down from Socrates, and use it in our conversations as often as possible, if you are amenable, Quintus.

(p.296) Quintus agrees, and the two brothers depart, with Cicero having suggested in the final paragraphs of the work a dizzying array of options for what he himself believes, and in turn having provided a multitude of ways for interpreting the use of his earlier writings within the dialogue.107

It is clearly no accident that Cicero bookended the two dialogues that make the most extensive use of his poetry with explicit appeals to the anti-authoritative Academy. Indeed, as early as De Legibus, he was already examining the tension between invoking his poetry as authoritative and disavowing such claims to authority. That conversation features Quintus as well, but there it is Atticus, the most historically minded of the dialogue’s three interlocutors, who insists on using Cicero’s poetry (in this case, the Marius) in a way of which Cicero does not approve (Leg. 1.1–5). Atticus wishes to know the real-world referents of the poem, and in correcting this impulse, Marcus reminds him that ‘those who demand the truth in this sort of matter not as if from a poet, but as if from a witness in court, act foolishly’ (Leg. 1.4: faciunt imperite, qui in isto periculo non ut a poeta sed ut a teste veritatem exigant). In history, he says, the aim may be truth, but in poetry, it is pleasure (1.5). Much like Atticus desires truth and proof from Cicero’s Marius in De Legibus, Quintus in De Divinatione openly refers to his brother’s De Consulatu as a ‘witness’, testis (1.17).108 In each case, Marcus rejects this use of his poetry: it is inappropriate to bring poetry as an authoritative witness into the prose genres within which his various interlocutors hope to benefit from its use (historiography in De Legibus, philosophy in De Divinatione).

As De Legibus shows, Cicero had been preoccupied with the role of his own authority within his literary works for some time. The use of his poetry within these works was only one of the things that vexed him about the correct application of literary authority: another was how, and whether, he should represent himself in those works in which he appears as an interlocutor. As early as De Re Publica, he was wondering if he should make himself an interlocutor in one of his dialogues, as his friend Sallustius had recommended, since doing so would lend ‘greater authority’ (Q. fr. 3.5.1 [25]: maiore auctoritate) to the proceedings; his decision not to include himself in that dialogue shows that he had Academic concerns about authority in mind already.109 Certainly he must have agreed with Sallustius that the choice of interlocutors (p.297) played an important role in authorizing the views of any given work, since he acknowledges this fact when he claims that De Senectute was lent greater authority through the choice of Cato as its speaker (Amic. 4). But the author’s inclusion of himself in one of his own dialogues was not necessarily as straightforwardly authoritative as Sallustius thought it to be. In his famous letter to Lucceius, Cicero argues the opposite, after a fashion, pointing out that writing about oneself gives one inherently ‘less authority’ (Fam. 5.12.8 [22]: minor auctoritas) since the strictures of modesty and propriety mean an author must necessarily both praise and condemn himself too little. But less authority is exactly what Cicero was looking for in his later philosophical works, where, in fact, he explicitly takes the Academic position that his own authority must be avoided altogether.

The difficulties Cicero believed were attendant on maintaining an appropriate level of authority (or lack thereof) when writing about oneself are illustrated well by a curious practice that can be found in all nine dialogues in which he is featured as an interlocutor: nowhere in any of them does he refer to himself by name.110 Though he addresses the other interlocutors many times in the vocative, he is not frequently addressed in return, and even on the rare occasion when a vocative is directed his way, it never includes his name: his son once refers to him as ‘my father’ (Part. or. 140: mi pater) and Quintus occasionally calls him ‘brother’, frater, in De Legibus (e.g. 1.5). If a situation calls for third-person address rather than a vocative, Marcus is still not named; in multi-person dialogues, this is achieved by strange circumlocutions such as the one we have already seen at De Natura Deorum 2.104 (me intuens, ‘looking at me’), or, for example, ‘and he called me by my name’ (Luc. 13: me autem nomine appellabat). The primary effect, very much in keeping with Academic principles, is to make ‘Marcus’ into merely another seeker of knowledge, rather than an authoritative arbiter of correct judgement. Cicero’s philosophical works are, after all, premised on the idea of dispassionately presenting in utramque partem debate, and he would be guilty of gross hypocrisy if he were to jeopardize this foundation by disclosing his own beliefs or seeking to influence his readers with his own authority.

It is this concern that lies behind Marcus’ condemnation of Quintus’ use of his poetry in De Divinatione. It is no accident that Quintus twice introduces his quotations with the word auctor: he is taking the poetry not only as a repository of Cicero’s positive beliefs, he is also using Cicero as an authority for his own beliefs. Marcus, the Academic sceptic wary of anyone’s authority, is required to condemn this use of his writings and his life in no uncertain terms.

(p.298) But much like the ending of De Natura Deorum suggests that Balbus was largely right to read the Aratea as an expression of Marcus’ (Stoicizing) beliefs, so too does the dizzying end of De Divinatione raise the question of where Cicero’s real allegiance lies. And as much as it was mistake for Quintus to assume that Marcus believed in the truth of divination because a character in one of his dialogues did, it is equally a mistake to assume that Cicero himself agrees with the arguments he makes in the guise of Marcus on the manner in which his poetry should be used. Condemnation of the authority implied by the word auctor can come easily to Marcus; he is, after all, just a dedicated seeker of knowledge in a philosophical dialogue. But it no doubt comes less easily to Cicero himself, who is the auctor in the more common sense of the word—the work’s author.

It is in recognizing this fact that the gravity of the problem Cicero identified in his letter to Lucceius becomes clear. How could a person truly write about themselves with authority? How could Cicero—whose hope for remembrance by posterity after Caesar’s accession lay chiefly if not solely in his textual output rather than his political deeds—authoritatively promote this textual self? For Cicero, the answer lay in the dialogism of his late philosophical and rhetorical works. Refracting the praise of his literary self through a figure like Balbus (and through Atticus in De Legibus) is ultimately Cicero’s way of avoiding, or at least mitigating, the problems that he flagged to Lucceius as endemic to any writing about oneself: by assigning praise of his poetry to these interlocutors while having his own avatar graciously demur, Cicero could exhibit a lack of modesty in practice while respecting it in theory, and at the same time denounce his own authority in theory while making sure it was carefully in place in practice.111 De Divinatione represents his full mastery of this approach. What better way, after all, to transform himself into a fully textual presence—a classical author and authority—than by having his own brother, the person who is in many ways Cicero’s second self, interact with him primarily as an author, a textual self?


The motivation that lies behind Cicero’s quotation of his own poetry in De Natura Deorum and De Divinatione is illustrated nicely by an anecdote about (p.299) the poet Lucilius that Cicero repeats three times in his philosophical treatises.112 The most telling version can be found in the authorial preface to De Finibus:

Nec vero, ut noster Lucilius, recusabo, quominus omnes mea legant. utinam esset ille Persius! Scipio vero et Rutilius multo etiam magis; quorum ille iudicium reformidans Tarentinis ait se et Consentinis et Siculis scribere. facete is quidem, sicut alia; sed neque tam docti tum erant ad quorum iudicium elaboraret…ego autem quem timeam lectorem, cum ad te, ne Graecis quidem cedentem in philosophia, audeam scribere?

Fin. 1.7–8

Unlike Lucilius, I will not object to everyone reading my works: I only wish that that Persius of his was still alive! And even more so Scipio and Rutilius, in fear of whose judgement he said that he wrote for the people of Tarentum, Consentia, and Sicily. He is of course being facetious, just as he is elsewhere; indeed, there were in his time no such learned men for whose judgement he had to exert himself…But what reader need I fear, if I dare to write to you [Brutus], who does not yield even to Greeks in philosophy?

The witticism of Lucilius to which Cicero refers here is related more fully by Crassus at De Oratore 2.25. There Lucilius is recounted as saying that he did not want his writings read by Persius, but by Decimus Laelius. Persius, Crassus explains, was the most learned Roman scholar of the period, while Laelius was a typical Roman aristocrat of moderate learning. In other words, Lucilius ‘wanted the things that he wrote to be read neither by the most unlearned nor by the most learned: the latter would understand nothing, while the former would understand perhaps more than he himself did’.113

Lucilius is not the only Roman author whom Cicero imagines as motivated by such concerns; in fact, Varro in the Academica expresses similar anxieties. After Marcus asks him why he has not tried his hand at philosophy, Varro responds that he has thought long and hard about doing just that, but that philosophy in Latin appears to be an impossible task: Romans interested in the subject would read the Greek originals, and if someone were disinterested in Greek subjects, they would certainly have no interest in philosophy, no matter what language it was in. ‘And so,’ he concludes, ‘I did not want to write things either that unlearned people would be unable to understand or that learned people would not care to read.’114 Varro believes that the middle ground Lucilius could find for his satires does not exist for philosophy: all potential readers are either too learned or not learned enough.

Cicero, in his philosophical works, sets out to prove Varro wrong; when he boasts in the preface to De Natura Deorum that so much progress has been made (p.300) in philosophy that the Greeks are no longer superior to Romans in richness of philosophical expression (1.8), he clearly considers the bulk of this work to have been done by himself.115 Nor is he just writing for the middle ground; as his comments in De Finibus make clear, he is so confident in his philosophical self-worth that he wishes the learned critic Persius were still alive. The frequent appearance of his poetry in these treatises shows that this boast is not solely reserved for his philosophy. Putting his Aratea in the mouth of Balbus, a modern Persius if there ever was one, is one of the many ways in which he demonstrates this fact. Balbus is described at the opening of De Natura Deorum as a man who has devoted his life to learning, someone who has progressed so far in his study of philosophy that he is able to compete with even the most learned Greeks (Nat. D. 1.15). His learning is put on display in his quotation of the Aratea when he recognizes and identifies a reference in the poem to an arcane bit of scholarship on the etymology of the Hyades. If the Aratea can withstand the scrutiny of Balbus, surely it can withstand anyone’s scrutiny. It can therefore confidently be used in the manner of any classical text, much as Quintus uses it in De Divinatione: as a piece of evidence with tremendous authority that can be used to support specific views and arguments.

Of course, Cicero’s apparent self-confidence about his poetry—and about his philosophical works—is a fairly transparent fiction. His anxiety about who would read these works, and how, is particularly clear in these moments of self-quotation, by virtue of the mere fact that he has engineered scenarios in which he can control its quotation. This anxiety is likely a sign that Cicero recognized that he could only sow the seeds for his posthumous legacy, not control the process completely, and this fact is borne out nowhere better than in the fate of his poetry. Because while the quotation of his own poetry in De Natura Deorum and De Divinatione may have been designed to elevate it into the Roman canon, the meagre fragments of the poetry that survive are almost solely indebted for their survival to Cicero’s success in making himself a classic in other genres.


(1) He had also already dabbled in a type of self-interpretation as early as the De Legibus, which opens with a discussion of his poem Marius, and in which, in the guise of the character Marcus, he provides commentary on the laws he proposes (e.g. Leg. 2.24–31).

(2) Frede 1999, Sedley 2003, and Hatzimichali 2013 all consider how classicism affected philosophy in the late second and early first centuries.

(3) Both Varro and Cicero speak of second-century poets in triads (e.g. Ennius, Pacuvius, Accius), suggesting that a consensus on canonization had begun to form in the late republic: on this, see Goldberg 2005. Kaster 2010 notes the numerous commentaries produced in the late republic on Naevius, Ennius, Plautus, and Lucilius.

(4) Citroni 2006: 215–17 and Habinek 1998: 64 discuss Cicero’s role in canonizing second-century poets through his philosophical treatises; Shackleton Bailey 1983 collects his judgements on early Latin poetry.

(5) Schofield 2008: 77 and Baraz 2012: 6–7 both note that the composition of the prefaces as a group shows that Cicero’s preoccupation with critical readers was an important aspect of these late philosophical works.

(6) See especially Baraz 2012 and Habinek 1994 and 1998: 66–8. The relationship of the philosophical works to the changed nature of the Roman state under Caesar is the subject of Bringmann 1971, Strasburger 1990, Wassmann 1996, and Gildenhard 2007.

(7) For the nature of the work, see Plasberg 1892 and Bringmann 1971: 111–23. The Hortensius drew heavily, as I noted in Chapter 3 (section ‘The Features of Cicero’s Aristotle’), on Aristotle’s Protrepticus (SHA Gall. 20.1).

(8) Luc. 5: sunt enim multi qui omnino Graecas non ament litteras, plures qui philosophiam, reliqui qui etiam si haec non inprobent tamen earum rerum disputationem principibus civitatis non ita decoram putent.

(9) These three works were written between June and December 45. Baraz 2012: 13–43 shows that in the intellectual climate of late republican Rome, Cicero was right to be concerned about such criticisms.

(10) On Cicero’s careful choice of interlocutors with auctoritas, see Steel 2005: 106–14, Linderski 1988: 108–10. As Baraz 2012: 204–12 notes, the dedicatees of the treatise are in possession of the auctoritas Cicero wants to claim for the arguments contained within them.

(11) Habinek 1998: 103–21 rightly identifies the gulf between writing and speech as one of the principal anxieties for Roman authors, while Dugan 2005: 285–8 considers how Cicero obviates this anxiety in Orator, the first work in which he seeks to make himself a textual entity. Baraz 2012: 196–8, however, reads Hirtius’ statement as an acknowledgement that De Fato will be more of a text, and less of a conversation: less multi-vocal, more straightforwardly didactic.

(12) Schofield 2008: 82–3 considers this passage from the Tusculans as part of a broader reflection on the ways that Cicero exploits authorial presence in his dialogues; Gildenhard 2007: 269–75 takes it as evidence for the increased philosophical sophistication of the interlocutor in the final book of the dialogue.

(13) Fox 2007: 217–18 considers this passage in light of the problems with authorial voice that are raised in De Divinatione.

(14) See Steel 2005: 55–9, Gibson and Steel 2010: 122–3. While Pliny the Younger does claim Cicero as a model for the propriety of a senator composing poetry, he does so solely in terms of nugatory poetry (Ep. 5.3), suggesting that it was Cicero’s epic poems that were the most unusual (on this, see also Volk 2013).

(15) The line cedant arma togae (‘let arms yield to the toga’) was particularly vulnerable, since it suggested disregard for Pompey, and Cicero mentions criticism of it at Pis. 72–5, Off. 1.77–8, and Phil. 2.20; cf. Dom. 92–3, where Cicero’s attendance at councils of the gods in De Consulatu may be the target. Allen 1956 notes that contemporary parodies of Cicero’s poetry were part of a pre-existing culture of poetic parody, and Fox 2007: 237–8 reads the quotation of De Consulatu in De Divinatione as an ironic appropriation of this parody.

(16) On Cicero’s response to Piso, see Dugan 2005: 62–5, Volk and Zetzel 2015: 208–11, Steel 2005: 68–9.

(17) Baraz 2012: 150–86 argues that similarly, the dedicatees of the philosophical works stand in for an ideal anonymous reader in whom Cicero incites feelings of amicitia and therefore a more charitable mode of reading.

(18) Cicero first developed this defence in the introduction to his unfinished translations of Aeschines and Demosthenes, suggesting that he considered it a useful analogue for any type of prose translation (Opt. Gen. 18).

(19) Puelma 1980: 148–9 discusses how in his philosophical works Cicero sets himself up as the ‘rival of Plato’, Platonis aemulus, by way of Ennius as ‘another Homer’, alter Homerus. Baraz 2012: 113–27 considers this passage from De Finibus in light of Rome’s fraught relationship with Greece as it cut across both literary genres and historical periods.

(20) As Baraz 2012: 116 notes, Cicero creates here a straw man that will preclude any member of his audience from being comfortable with anything less than his own point of view.

(21) McElduff 2013: 108–9 points out that by denigrating word-for-word translation, Cicero aggrandizes his own role in the process: access to Greek literature must be obtained for his readers through his own distinctive style and preferences. See also Copeland 1991: 29–30.

(22) As Douglas 1968: 27–34 describes in some detail, late nineteenth- and early twentieth-century work on the philosophica tended to view it as a patchwork quilt of Hellenistic philosophers mechanically copied by Cicero, but recent scholarship (following on the groundwork laid by, e.g. Douglas 1965 and Boyancé 1970) instead emphasizes Cicero’s command over this material, and his judicious adaptation of it for a Roman audience. MacKendrick 1989: 3–7 lays out the basic lines of the debate.

(23) See Schofield 2008, Gersh 1986: 60, Schmidt 1978: 124–5. Fox 2007 examines all of Cicero’s philosophical works from the perspective of his Academic scepticism.

(24) For Ciceronian verisimilitude see, e.g. Ruch 1948a, Levine 1957, Dyck 1998, and Dugan 2005: 96–9; Rawson 1972: 40–1 notes that Cicero’s care with historical accuracy was unusual in philosophical dialogues, and part of a larger interest in antiquarianism. But Kronenberg 2009: 76–93 argues that the dialogues still came across as artificial, and that Varro satirizes their attribution of Greek knowledge to traditional Roman figures.

(25) The change is recounted to Atticus over the course of several letters (Att. 13.13–14 [321], 13.16 [323], and 13.19 [326]).

(26) Amic. 4: Genus autem hoc sermonum positum in hominum veterum auctoritate et eorum illustrium, plus nescio quo pacto videtur habere gravitatis.

(27) Jocelyn 1973: 67–71.

(28) For Epicurean attitudes towards poetry, see Plutarch Mor. 15d and 1087a; Heracl. Alleg. 4, and Sextus Empiricus Math. 1.272–3. The poetic works of Lucretius and Philodemus suggest, though, that there was more room within the Epicurean school for differing opinions on this topic than has often been thought.

(29) For the Stoic attitude towards poetry, see Tusc. 2.26, Nat. D. 1.41, NA 6.16.6, and Galen Hipp. et Plat. 3.2–3 (on Chrysippus). Modern analyses can be found in De Lacy 1948, Pfeiffer 1968: 234–51, Long 1992, and Struck 2004: 111–41.

(30) For Academic views on poetry, see Diog. Laert. 4.26 (on Crantor); Tusc. 2.26 (on Philo), and, in general, Plut. Mor. 16. The Peripatetics, who are rarely represented as a separate entity from the Academics in Cicero’s dialogues, were of a similar mind-set: Aristotle seems to have mistrusted poetry, but believed that since it expressed true observations, it was permissible to use it to support a proposition (see, e.g. Rh. 3.1.1404a20–39).

(31) The Epicurean arguments in Cicero are those of Torquatus in Fin. 1 and Velleius in Nat. D. 1. For mockery of the more serious attitude see, e.g. Velleius’ remarks on Chrysippus at Nat. D. 1.41.

(32) Since Cicero associated himself with the Academy, he frequently assigns the Academic position to himself, including the arguments in Fin., Div., and Luc.; at Tusc. 2.26, he claims to have modelled his use of poetry on Philo. Cotta represents the Academic viewpoint in Nat. D. 3 and Varro takes an Antiochean position in Ac.; although he represents Antiochus’ ‘Old Academy’, his attitude towards poetry is the same.

(33) Jocelyn 1973: 71–9 also notes this fact, although he assumes Cicero was uncritically copying his sources’ quotations, rather than using them to fashion something new in Latin. The shared tradition was likely doxographical in nature, as Mansfeld 1990 and 1992 shows.

(34) Reinhardt 2003: 329 notes that the proem to Euripides’ Medea had an established place in Greek philosophical discussions of causality, and Cicero’s substitution is clearly meant to found a parallel tradition in Roman philosophy. Sharples 1995 uses the shared quotation to consider the relationship between the respective discussions of causality in Topica and De Fato.

(35) I make a similar argument at greater length in Bishop 2016.

(36) These include a hemistich of the first line of the Aratea (Leg. 2.7), two lines of fr. VII from the same poem (Luc. 66), a single line from fr. XXIII (Orat. 152), and the famous cedant arma togae (‘let arms yield to the toga’) of De Consulatu at Off. 1.77. A single line from Marius is quoted at Leg. 1.2, where it sparks an extended discussion on the nature of poetry.

(37) On Cicero’s use of Ennius, see Elliott 2013: 152–95 on the Annales and Zetzel 2007 more generally.

(38) For a list of authors quoted by Cicero in his treatises, see Shackleton Bailey 1996: 140–2. Cicero’s Academic speakers use slightly more comedic quotations than his Stoics.

(39) It is also worth noting that Quintus quotes at length (supposedly translating extempore, it would appear) from Plato Republic 9 (Div. 1.60–1) and from the Peripatetic philosopher Cratippus (Div. 1.70–1). In Book 2, Marcus answers with a significant translation from the Iliad (Div. 2.63–4) and short quotations from Homer (Div. 2.82), Euripides (Div. 2.12), and the Delphic oracle (Div. 2.115).

(40) A twenty-nine-line quote from the Iliad is found at Div. 2.63–4, while forty-five lines from Sophocles appear at Tusc. 2.20–2, and thirty-two from Aeschylus at Tusc. 2.23.

(41) On this, see also Gee 2001: 530.

(42) This was such a pervasive idea in Ciceronian Quellenforschung that an entire article, Dobson 1918, was written to rebut it.

(43) Note, e.g. Fin. 1.6 (‘our friend Posidonius’: familiarem nostrum Posidonium); Nat. D. 1.6 (as one of ‘those leading men’, principes illi, with whom Cicero trained in his youth), 1.123 (‘Posidonius, a friend to all of us’: familiaris omnium nostrum Posidonius), 2.88 (‘our friend Posidonius’: familiaris noster…Posidonius); Div. 1.6 (‘our Posidonius’: noster Posidonius). At Att. 16.11.4 [420] Cicero mentions that he has requested a copy of Posidonius’ treatise on duties for his own De Officiis.

(44) Citations of Posidonius in the scholia to Aratus are all found in the Weather Signs section of the poem, suggesting this was his primary field of interest in the poet: see In Arat. 429.5–430.10 (on Phaen. 881), 509.21–10.14 (on Phaen. 1091), and 511.20–12.12 (on Phaen. 1093). It is possible that these derive from a work by Posidonius on meteorology; possible title(s) for this work include Μετεωρολογική‎ (Στοιχείωσις‎) (F 14–15 E–K), Περὶ‎ Μετεω‎́ρων‎ (F 16–17 E–K), and Μετεωρολογικά‎ (F 18 E–K).

(45) As Martin 1956: 22–3 notes, Rhodes was a major centre for astronomy in this period, and it is likely that Posidonius was familiar with Attalus’ and Hipparchus’ commentaries on Aratus, the latter of which is intent on demolishing Aratus’ scientific credentials and showing his dependence on Eudoxus. On this see also van den Bruwaene 1973.

(46) Translation from Kidd 1999.

(47) As both Pease 1920–3: 28 and Jocelyn 1973: 75 do.

(48) For the range of possible dramatic dates, see Pease 1955: 25. Walsh 1997: xxxviii suggests a date of 76; since the dialogue takes place during the Feriae Latinae, he pinpoints the dialogue as taking place somewhere between April and July of that year.

(49) Nat. D. 1.40: idemque disputat aethera esse eum quem homines Iovem appellarent, quique aer per maria manaret eum esse Neptunum, terramque eam esse quae Ceres diceretur, similique ratione persequitur vocabula reliquorum deorum.

(50) Further examples of Stoic etymologies can be found at Diog. Laert. 7.147 (which shares a common tradition with Cicero’s source) and SVF 2.1061–1100. According to Diogenes Laertius (7.200), Chrysippus wrote two works on etymology: περὶ‎ τῶν ἐτυμολογικῶν πρὸς Διοκλέα‎, in seven books, and ἐτυμολογικῶν πρὸς Διοκλέα‎ in four.

(51) Cotta likewise considers the use of etymology a ‘dangerous habit’ (3.62: periculosa consuetudo), and in other treatises from this period Cicero shows a mistrust for it; see further Rawson 1972: 37.

(52) Cf. Plutarch Mor. 34b; this was a standard criticism of Stoics in general in antiquity, as Struck 2004: 15 notes, and at Tusc. 2.26 Cicero also criticizes the Stoic Dionysius of Cyrene for overuse of poetry. Philodemus, De piet. col. 13 (= SVF 2.1078) is an almost verbatim echo of Velleius’ criticism, and it is generally agreed that Cicero used Philodemus as his source for this passage, while adding his own stylistic flourishes: see Obbink 2001 and Dyck 2003: 8.

(53) Cicero gives Balbus Stoic credentials that mirror his own—though unlike Cicero, Balbus has progressed so far in Stoicism that he can even rival the Greeks (1.15). He is like Cicero in his personal friendships with both Posidonius (2.88) and Antiochus of Ascalon, who has, at the time of the dialogue, recently dedicated a treatise to him (1.16).

(54) Cicero’s Academic speaker, Cotta, will criticize this argument at 3.62–3.

(55) Dietrich 1911: 32–41 argues that Balbus’ Roman etymologies derive from the scholar and antiquarian Aelius Stilo, who taught both Cicero and Varro, and who had an affinity for Stoicism (Brut. 206, Leg. 2.59). At Top. 10 Cicero attributes an etymology to him, so Aelius was clearly one of his sources for Latin etymologies.

(56) As Dietrich 1911: 33–5 points out, many Latin etymologies (particularly the earlier ones) seem to be imitations or translations of Greek etymologies, and these examples certainly bear this out. See also Pease 1955: 710–11, 720.

(57) Cicero is the first extant author to connect Jupiter to iuvans pater, though at Ling. 5.67 Varro alludes to the etymology in deriving Juno (which Cicero also derives from iuvans) from una iuvat (‘she helps together’). The tradition of deriving Ceres from gero, however, can already be found in Ennius’ Epicharmus (ap. Ling. 5.64): quae ‘quod gerit fruges, Ceres’ (‘she who [is called] Ceres because she bears fruits’).

(58) Balbus considers the intelligent design of the universe (which is best illustrated by the regular movement of the heavens) to be the best argument for the existence of god, as was common in Stoic philosophy. For Stoic doctrine on astronomy, see Diog. Laert. 7.144–6, 155–6; for Stoic physics writ large, see L–S 43–55.

(59) Nat. D. 2.4: quid enim potest esse tam apertum tamque perspicuum, cum caelum suspeximus caelestiaque contemplati sumus, quam esse aliquod numen praestantissimae mentis quo haec regantur? A similar argument is attributed to Chrysippus at 2.16.

(60) As I discussed in Chapter 1 (n. 127), the planets represent a rare moment of potential instability in Aratus’ orderly cosmos, and it is thus significant that Balbus calls the description of them as ‘wanderers’ (errantes)—a description that Cicero had in fact used in the Aratea (230: malunt errare vagae, ‘they prefer to wander aimlessly’)—false (2.51). While he takes Cicero’s Aratea as the best example for an orderly cosmos, he must correct even the few moments of disorder within it.

(61) Balbus’ movement from low to high reverses Aratus’ order in the Weather Signs, perhaps intentionally, and interestingly, Vergil proceeds along similar lines at the end of Georgics 1 (though he culminates with the sun): see Farrell 1991: 79–83. Setaioli 2005: 241–3 argues that Vergil alludes to the section of De Consulatu embedded in De Divinatione; it is likely that Balbus’ catalogue in De Natura Deorum had a similar influence.

(62) Diels’ famous reconstruction of the doxography of Aetius is explained well by Mansfeld 2016. Mansfeld 1990 and 1992 discuss the relationship between De Natura Deorum and doxography.

(63) These are verses 33–4 of the so-called Sisyphus Fragment, whose authorship is attributed variously to Euripides and to Critias.

(64) Aetius names Posidonius as a source at several points (see, e.g. Plac. 1.28.5, 2.9.3, 3.1.8), and a comparison of the two texts that make up Diels’s Placita (Stobaeus’ Eclogae 1 and Ps-Plutarch’s epitome of Aetius) suggests he is also the source for 1.6 specifically. He likewise provided at least part of the argument for De Natura Deorum 2: see Pease 1955: 45–8 and MacKendrick 1989: 182–4.

(65) Ennius: 2.4 (a suggestive passage on the divine intelligence that orders the heavens), 2.49, 2.65; Accius: 2.89; Pacuvius: 2.91; Terence: 2.60.

(66) In fact, Balbus stops just before the famous recusatio on the planets (Phaen. 454–61 = Arat. 223–36), and Cicero has him give a recusatio of his own at 2.119: ‘I do not wish to seem excessive to you on the subject of the stars, especially those which are said to wander’ (nolo in stellarum ratione multus vobis videri, maximeque earum quae errare dicuntur). He does, however, briefly quote from the Aratea again at 2.159 (fr. XVIII) as part of his argument that everything in the world (here, oxen) is suited to human use.

(67) So, for example, the eighty-line quotation from De Consulatu in the first book of De Divinatione (1.17–22) and the twenty-nine-line passage from the Iliad in the second (2.63–4). The same holds for the two long quotations in the Tusculan Disputations.

(68) When discussing those verses of the poem that appear in De Divinatione and De Natura Deorum, I follow the text of the Aratea printed by Pease 1920–3 and 1955, respectively, while providing cross-references to Soubiran 1972. There are minor variants between the two, but this is not surprising, given Cicero’s interlocutors are in theory (and Cicero may be in fact) quoting from memory.

(69) Nat. D. 2.110: atque ita dimetata signa sunt ut in tantis descriptionibus divina sollertia appareat. Other Stoicizing comments can be found in Balbus’ quotation of fr. VII at 2.106 and his conclusion of the quotation at 2.115. Gee 2001: 529–30 notes well the way these passages echo Cicero’s poetic vocabulary in the Aratea.

(70) These are frs. XII and XIV respectively; while in the Aratea Cicero transliterates the name Ophiuchus (fr. XIV), Balbus refers to the constellation with the calque Anguitenens (Nat. D. 2.108). Ophiuchus’ name is treated similarly by Aratus, on which see Pendergraft 1995: 55.

(71) This is fr. XVI. Temo (‘waggon pole’) was used to describe this constellation at least as early as Ennius; see Ling. 7.73–5. Aratus’ use of etymology for this constellation is discussed by Pendergraft 1995: 56–7.

(72) On this etymology for the Hyades, see O’Hara 1992: 55, Kidd 1997: 246, and Pease 1955: 822–3.

(73) In Arat. 167.1–3: ἢ ὅτι παραπλησίως τῷ Υ στοιχείῳ κεῖνται, ἢ ὅτι δύνουσαι ποιοῦσιν ὑετόν, ἢ κατ‎’ ἔλλειψιν τοῦ θ θυάδας‎. Thyades was the Delphic name for the female followers of Dionysus.

(74) As Fox 2007: 46–7 suggests, there is already some tension here in ascribing anti-authoritative sentiments to authoritative figures like Carneades, a tension that clearly extends to the use of Cicero himself as a literary authority in the work. Woolf 2015: 51 also notes the tension between Balbus’ flattering play for the young Cicero’s assent and the elder Cicero’s disavowal of appeals to authority.

(75) On this see Gee 2001 and 2013b passim.

(76) Arat. 162–3: haec ille astrorum custos ratione notavit/signaque dignavit caelestia nomine vero.

(77) Gee 2013b: 57–109.

(78) In fact, Marcus confirms his attraction to Stoic cosmology at Div. 2.148, a passage I will discuss at the end of this chapter.

(79) Nat. D. 3.95: mihi Balbi [sc. disputatio] ad veritatis similitudinem videretur esse propensior. Marcus’ choice has provoked much scholarly discussion: see, e.g. Pease 1913 and 1955: 33–6 and Levine 1957. Schofield 2008: 73–4 notes that regardless of Marcus’ choice, the dialogue still ends aporetically, with opinions divided.

(80) Nat. D. 3.91: utrum poetae Stoicos depravarint an Stoici poetis dederint auctoritatem non facile dixerim; portenta enim ab utrisque et flagitia dicuntur.

(81) He first mentions the idea at Nat. D. 3.19.

(82) Div. 1.7: argumenta cum argumentis comparemus, ut fecimus in iis tribus libris quos de natura deorum scripsimus. Similar comments are made in the authorial preface to the second book (Div. 2.3).

(83) Div. 1.8. Quintus assumes that the omission of divination from De Natura Deorum was purposeful: ‘But there is something that was omitted in those books (I assume because you thought it more convenient that it be investigated and discussed separately), namely divination’ (sed, quod praetermissum est in illis libris (credo quia commodius arbitratus es separatim id quaeri deque eo disseri), id est de divinatione).

(84) Pease 1920–3: 18–29, MacKendrick 1989: 197–8, and Wardle 2006: 28–36 give an overview of the Quellenforschung: the communis opinio is that Posidonius was the primary source for the first book, and that the arguments of the Peripatetic Cratippus were incorporated to a lesser extent. For the Περὶ‎ μαντικῆς‎ see F 26–7 E–K.

(85) It is unclear why Cicero chose Quintus to represent Stoic views in this dialogue; the real Quintus was more a Peripatetic than Stoic, as Cicero allows him to admit at 2.100 (and cf. Fin. 5.96). But I suspect Quintus was chosen for an argument that hinges so much on his brother’s public writings and careers precisely because he was his brother, and thus, as Cicero points out in Fam. 2.15 [96], in many ways his alter ego.

(86) Krostenko 2000 provides an outline of Quintus’ argument, which Wardle 2006: 20–6 elaborates upon. But Schofield 1986: 52 suggests that the messiness of Quintus’ argument is part of its point, because it allows Cicero to indulge his storytelling skills and showcase his poetic talent.

(87) Diogenes mentions several Stoic philosophers by name who took this approach to divination: Zeno, Chrysippus, Athenodorus, and Posidonius. Cicero provides a similar doxography at Div. 1.6, and says at 1.5 that the practice of divination long ago was sanctioned more by results than by reason.

(88) Pease 1920–3: 28 suggests Posidonius used the Weather Signs in his Περὶ‎ Μαντικῆς‎, and Jocelyn 1973: 75 also argues that Cicero came across its quotation in his source(s) for De Divinatione.

(89) Thus too van den Bruwaene 1973: 436. This is not unlike how Cicero created complementary moments of self-display at the end of the paired works Brutus and Orator, but as Fox 2007: 233–4 rightly notes, De Divinatione counters the silent, defeated Cicero of Brut. 330 with an ‘elaborately multi-vocal self portrait’.

(90) Wardle 2006: 37–8 suggests the dialogue is set in April 44, when Cicero made a brief visit to his Tusculan villa, but Cicero probably did not expect his readers to date the work so precisely; as Linderski 1988: 110–11 notes, after the accession of Caesar, Cicero became less precise about his dialogues’ dramatic dates. Ones set before then had to occur in times of otium; under Caesar, all time was otium.

(91) Fox 2007: 209–40 considers how the nature of Quintus’ highly intertextual argument—which, as he notes, makes mention of nearly all of Cicero’s non-forensic writings—allows Cicero to explore his own relationship with now-vanished republican institutions.

(92) Div. 1.22: tu igitur animum poteris inducere contra ea quae a me disputantur de divinatione dicere, qui et gesseris ea quae gessisti, et ea quae pronuntiavi accuratissime scripseris? Krostenko 2000: 365–70 notes the importance of authority, both poetic and otherwise, for Quintus’ argument.

(93) Fox 2007: 212 n. 3 notes the connection of the word with divination, but it could also be used more broadly for a legal verdict or official ruling; both senses can be found in the TLL (s.v. I).

(94) The twenty-three lines Quintus quotes from the Prognostica are now almost all that is extant of it. Though Quintus’ quotation is split into five blocks, Soubiran 1972 groups the lines into three fragments (III ~ Phaen. 909–15, IV ~ Phaen. 946–55, V ~ Phaen. 1051–3); three other short fragments (I, II, and VI) are preserved in Priscian.

(95) Just as Cicero used a commentary to Aratus to translate the first section of the Phaenomena—possibly the one in four books by the Stoic Boethus of Sidon—so too did he draw on exegetical resources here: compare his translation with In Arat. 441.4–8.

(96) Quintus notes the internal organization of these passages at Div. 1.16 (‘thus the signs of wind and rain’, sic ventorum et imbrium signa), as does Marcus at Div. 2.14.

(97) Opinions remain divided on the identity of Cicero’s acredula, as well as the animal described by Aratus in the corresponding passage, an ὀλολυγω‎́ν‎. Cicero’s description is reminiscent of a turtle dove, and likely borrows from the scholia, which also argue that Aratus’ animal is a bird (cf. In Arat. 458.14–59.4).

(98) In addition to several short tags from Ennius (1.88, 1.114), Pacuvius (1.24, 1.29, 1.80, 1.131), and Plautus (1.65), Quintus quotes twenty-two lines from Accius’ Brutus (1.44–5), thirty-seven from Ennius’ Annales (1.40–1, 1.107–8), fifteen from his Alexander (1.66–7), and another twelve from an unnamed tragedy, likely by Ennius (1.42).

(99) Volk 2013 discusses the poetic significance of this fragment from De Consulatu; Fox 2007: 235–6 notes that Quintus has chosen it well, since Urania’s speech corresponds nicely to the arguments under debate. Both Kubiak 1994 and Setaioli 2005 point to the passage’s connections with the Aratea and with Stoicism.

(100) Quintus also names Marcus an authority one further time, at 1.68, when he relates a prophecy his brother heard about the battle of Pharsalus.

(101) As Schofield 2008 notes, there is evidence that Cicero thought his readers would expect thematic unity for interlocutors who appear across dialogues: at Tusc. 5.32, the respondent says that Marcus’ views are inconsistent with his thoughts in De Finibus, and he replies that as an Academic he has the right to change his mind about issues. Quintus too shares characteristics across dialogues: he claims allegiance to the Lyceum at Div. 2.100 and Fin. 5.96, and shows a marked preference for Roman exempla over Greek argumentation in both De Legibus (on which see Atkins 2013b: 163–4) and De Divinatione (on which Schofield 1986: 51–3, Lehoux 2012: 22–5).

(102) Linderski 1982, Beard 1986, Schofield 1986, Krostenko 2000, and Lehoux 2012: 30–7 all reckon with the inconsistencies between the Marcus of De Legibus and the Marcus of De Divinatione.

(103) Marcus says at 2.54 that he will disprove the portents he recorded in De Consulatu, but fails to do so; he does, however, rebut the several personal instances of divination that Quintus claimed he had experienced (2.65, 2.114, 2.136–42) and at 2.70–1 says he has no duty as an augur to defend augury. On Marcus’ handling of Quintus’ ad hominem tactics, see further Schofield 1986: 56–8.

(104) I am sympathetic to the arguments of Linderski 1982 and Krostenko 2000, both of whom suggest that Marcus’ rejection of divination here (particularly as evinced in the De Consulatu and Marius) is a way for him to distance himself from elite use of religious symbols as a mode of authority, which Caesar’s rise had led him to re-evaluate. But this argument does not sufficiently explain the role of the Prognostica—with its ties to the use of the Aratea in De Natura Deorum—in the dialogue.

(105) Marcus’ similarity to Cotta is also highlighted at 2.104, where he uses the same passage of Ennius that Cotta had at Nat. D. 3.79 to argue that there are gods, but they do not care about humanity.

(106) Div. 2.148: et esse praestantem aliquam aeternamque naturam et eam suspiciendam admirandamque hominum generi pulchritudo mundi ordoque rerum caelestium cogit confiteri.

(107) Fox 2007: 232 also notes the variety of Ciceros on display in this closing passage. The play of voices was foreshadowed in the work’s opening, since Quintus’ first piece of proof for divination (1.12) echoes something Cicero himself says in the authorial preface (1.2).

(108) In both cases, it is Cicero’s historical poetry that is being used, and each time the person using it wishes it to have the same truth value as historiography (or indeed, given the motif of witnesses, oratory). Marcus’ refusal to approve this approach is a tacit admission that the generic trappings of epic poetry require embellishment that limits their historical value.

(109) As Fox 2007: 85–8 and 2009: 55–60 argues. Schmidt 2001 discusses the various changes Cicero considered to the structure of the dialogue.

(110) Dickey 1997 compiled evidence for this phenomenon in the dialogues, but could ‘offer no explanation for this reluctance’ (586).

(111) Cicero’s propensity for inserting self-praise into the mouths of his interlocutors was recognized already by Quintilian (Inst. 11.1.21); see too Douglas 1966: xviii and Dugan 2005: 173. Fox 2007: 47, speaking of the preface of Nat. D., likewise notes that ‘there is an unsolved tension here between a desire for the preservation of an individual’s opinions and a recognition of an appealing purity in the non-doctrinal processes of philosophy’.

(112) De or. 2.25, Fin. 1.7–8, and the lost preface to the De Re Publica (Pliny, HN pr.7).

(113) De or. 2.25: quae scriberet neque se ab indoctissimis neque a doctissimis legi velle, quod alteri nihil intellegerent, alteri plus fortasse quam ipse…

(114) Acad. 1.4: itaque ea nolui scribere quae nec indocti intellegere possent nec docti legere curarent.

(115) As Baraz 2012: 110–11 notes, Cicero’s philosophical expertise may have been questioned, but prose style was an area in which he knew he was without peer.