PART I
1
in the summer of 54, after he had completed De oratore and while he was writing De re publica, Cicero wrote a long letter to Atticus including two paragraphs in which he discussed the settings of the two dialogues (Att. 4.16.2–3):
You write to me about Varro: I’ll put him somewhere, if there’s a place. But you know the way my dialogues work (nosti genus dialogorum meorum). Just as in the books about the orator, which you praise extravagantly, the speakers couldn’t mention anyone except people they knew or had heard of, so too in the dialogue about the commonwealth which I’ve begun I’ve assigned the roles to Africanus and Philus and Laelius and Manilius, adding as younger participants Q. Tubero, P. Rutilius, and Laelius’ two sons-in-law, Scaevola and Fannius. And so I thought that since I’m using prefaces for each book the way Aristotle did in the books he called “for outsiders” (ἐξωτερικούς), I might be able find some reason to mention him [Varro], and I know that would please you. If only I can finish what I’m trying to do! As you know, I’ve undertaken something big that demands a lot of free time—something I’m very short of.
As for the figure of Scaevola, whom you miss in the book you praise, I removed him deliberately, following what the divine Plato did in his Republic. After Socrates comes to Cephalus’ house in Piraeus, that rich and sociable old man takes part in the first part of the discussion. But then, after he has made some very appropriate comments, he says that he wants to go and sacrifice, and he doesn’t come back. I believe that Plato thought that it would hardly be appropriate to make a man of that age stay too long in such an extended conversation. I thought it was even more to be avoided by me in the case of Scaevola: you remember his age and state of health, and he was so eminent that it seemed somehow wrong for him to spend several days in Crassus’ villa at Tusculum. Besides, the conversation of the first book went well with Scaevola’s interests, but the other books have a technical discussion, as you know. I really didn’t want that whimsical old man (as you know he was) to take part in it.
Atticus had apparently asked Cicero to include Varro somewhere in his writings, and Cicero in explaining the difficulty of fulfilling that request offered a description of how his dialogues worked (nosti genus dialogorum meorum): both De oratore (in oratoriis) and De re publica (hanc . . . de re publica . . . disputationem) made use of historical settings from which Cicero strictly excluded any anachronism. At the same time, however, Cicero points out that he follows Aristotle’s practice in his dialogues of writing prefaces, and therefore he might find a place for Varro somewhere. In the event, Varro did not appear until nine years later, in the second version of the Academica, after Atticus had again urged Cicero to include Varro in one of his dialogues.1
The next paragraph of Cicero’s letter deals with a different comment about Cicero’s dialogues: Atticus seems to have expressed his regret that the old man Scaevola had appeared only in Book 1 of De oratore.2 Cicero’s response is striking: the role of Scaevola in De oratore, he says, was based on that of Cephalus in Plato’s Republic. That model explains both the presence and the subsequent departure of the old man, but equally, the departure of Scaevola at the end of Book 1 shows Cicero’s own concern for decorum: someone as old and distinguished as Scaevola would not have lingered at Crassus’ villa for another full day of conversation, nor would he have taken part in the technical exposition of rhetorical theory (τεχνολογία) which occupies much of Books 2 and 3. Nor, according to Cicero, would Plato himself have failed to recognize the implausibility of having Cephalus take part in the entire ten-book conversation that begins with him.
This passage is Cicero’s earliest and most precise description of his methods of composition, and it shows just how carefully he designed his first dialogues: his great concern with historical accuracy and the avoidance of anachronism, and, at least in the case of Scaevola, his attention to verisimilitude and decorum. The discussion of Scaevola shows how Cicero creates the setting for a dialogue that is at once historical and a literary fiction: he is aware not simply of how an old man should be portrayed, but of how Plato had portrayed a similar old man. Scaevola is positioned within two quite different frames of reference: on the one hand, Cicero describes the real Q. Mucius Q.f. Scaevola Augur, the consul of 117 bce,as he was in early September of 91 bce, and he draws on his own and Atticus’ memory to confirm his accuracy; on the other hand, Cicero’s Scaevola is an intertextual avatar of Plato’s Cephalus.
Scaevola is based on the Platonic Cephalus, and in fact within De oratore it is Scaevola who is reminded of Plato’s Phaedrus by the plane tree in Crassus’ estate (1.28). And given the obvious Platonic echoes not only in De oratore but in De re publica as well, one must ask why Cicero, both here and in the letter to Lentulus, describes the format of his dialogues as Aristotelian rather than Platonic. The generic comments in the first paragraph quoted above are just as striking as the equation between Scaevola and Cephalus in the second. Not only does he call the dialogue Aristotelian, but his use of the phrase genus dialogorum meorum suggests not only that there was a pattern that he intended to follow in other works but also that this was a pattern (one need not say genre) for which he knew that there were rules—or for which he had himself invented the rules. What kind of dialogue is this?3
Because we still possess both Plato’s and Cicero’s dialogues, and because Plato is such an obvious presence in De oratore, De re publica, and De legibus, it is natural to take the link between the two as direct and unproblematic, but that is not really true.4 In the first place, even though within these dialogues Cicero draws attention to his use of Plato, with one possible exception he never does so in his own voice: it is his Scaevola, for instance, who refers to the Phaedrus, not Cicero himself; it is Crassus, not Cicero, who talks about studying the Gorgias in Athens; it is Scipio, not Cicero, who translates a passage of the Republic in De re publica.5 And while Cicero obviously knew the Platonic dialogues very well and it is he who wrote these passages and put Plato’s words in his characters’ mouths, it is striking that he does not talk about Plato in his own voice; and when Cicero does talk about his dialogues in his letters, he calls them either Aristotelian or Heraclidean, not Platonic.6
Plato was not, in fact, the only game in town. Starting with Socrates’ contemporaries and continuing for several generations, dialogue was a common way to write about philosophy—or indeed about intellectual topics of any kind. Although we can read complete dialogues by only Plato and Xenophon, there were many others. Of the early Socratics, Antisthenes and Aeschines are important in this respect.7 Some of their dialogues, like Plato’s, included Socrates as protagonist: as early as De inventione (1.51) Cicero was familiar with a dialogue by Aeschines Socraticus in which Socrates converses with Xenophon and Xenophon’s wife. Other Socratics, notably Antisthenes, picked different historical and sometimes mythical contexts. In the next generation, Aristotle’s exoteric works certainly included dialogues (which have not survived, except in fragments) and Cicero knew them much better than he knew the Aristotelian works that we possess: he read the four-book dialogue On justice as well as On philosophy and the Protrepticus (which may or may not have been dialogues).8 Aristotle, moreover, had a lot of company: among his contemporaries, Heraclides of Pontus, Dicaearchus of Messana, and Aristo of Ceus all made significant use of dialogue. Cicero had read them all: he knew a great many works by Dicaearchus, including two dialogues of three books apiece, one set in Corinth and one in Mytilene, on the mortality or immortality of the soul;9 he knew a work of Aristo of Ceus that presented Tithonus discoursing on old age;10 and he certainly knew works, some of them dialogues, by Heraclides Ponticus.11
Heraclides and Aristotle, not Plato, are the only models Cicero explicitly invokes for his dialogues. They appear together in a letter to Cicero’s brother Quintus written in late October or early November 54, concerning the setting of De re publica (QFr. 3.5.1):
You ask what I’ve done with the books I began when I was at Cumae. I haven’t stopped working on them, and I still am, but I have changed the whole plan and structure of the book several times. I had already written two books, in which I set up a conversation on the nine-day holiday that took place in the consulate of Tuditanus and Aquilius [129 bce] a little before the death of Africanus, including also Laelius, Philus, Manilius, P. Rutilius, Q. Tubero, and Laelius’ two sons-in-law Fannius and Scaevola. The conversation, about the best organization of a state and the best citizen, was divided into nine days and nine books; it was proceeding very well, and the eminence of the men added weight to what they said. But when these books were read aloud at Tusculum in the presence of Sallustius, he reminded me that such a discussion would have much more authority if I were speaking about the commonwealth, especially since I am not some Heraclides Ponticus but a consular of great public experience. If I attributed it to such ancient men, it would seem fictional; there was charm in my distancing from myself what I wrote in my book about oratory, but I attributed it to people I had seen myself. And besides, he said, Aristotle himself says in his own voice what he writes about the commonwealth and outstanding men.
In the conversation Cicero reports, his friend Sallustius apparently said that Cicero’s words would have greater auctoritas if he were himself the speaker, and that attributing them to men of old would seem fictional: quae tam antiquis hominibus attribuerem, ea visum iri ficta esse. Sallustius contrasted the setting of De re publica with that of De oratore, which took place in Cicero’s own lifetime among people he had known and therefore had greater plausibility. After all, urged Sallustius, Cicero was a Roman statesman, not Heraclides Ponticus. And Sallustius also noted that Aristotle in his works on public affairs spoke in his own voice.
As far as the setting of De re publica is concerned, Sallustius’ comment is foolish: that a dialogue with a historical setting is obviously a literary fiction does not, for Cicero any more than for Plato, diminish the importance of the ideas expressed.12 Sallustius’ comment about De oratore, on the other hand, is helpful for understanding the composition of the earlier dialogue, in that it echoes what Cicero himself says in the prefaces to Books 2 and 3 (which Sallustius had clearly read) about its taking place in his lifetime among people he knew. There, Cicero means to lend credibility not to the substance of what is said about oratory and rhetoric but to his characterization of Crassus and Antonius: in the preface to Book 2 he talks about their knowledge of rhetoric and philosophy, and in the preface to Book 3 about their styles of speaking.
Sallustius’ suggestion that Cicero speak in De re publica in his own voice contrasts two models for dialogue, Aristotle and Heraclides Ponticus. Although Cicero himself mentions both models elsewhere, he never offers a coherent discussion of the nature or history of dialogue, and his descriptions of his own dialogues are—for the modern literary historian—annoyingly offhand. That is particularly true of his descriptions of Aristotle, the model he most frequently cites. In different contexts, he defines Aristotelian dialogue in three different ways. In the letter to Atticus quoted above (4.16.2), “Aristotelian” dialogue is one in which there is a preface in which the author speaks in his own voice; that applies to De oratore, and indeed to all Cicero’s dialogues with the exception of De legibus. Within De oratore, however, Cicero makes Crassus characterize an orator as “someone who can speak in Aristotle’s manner about every subject on either side, and after mastering his instructions can unfold two contradictory speeches” (3.80). That is a reasonable description of the last part of Book 1 of De oratore (and characterized as such by Antonius in Book 2), but certainly not of the dialogue as a whole, although it is clearly applicable to part of De re publica (the Carneadean debate about justice in Book 3) and to many of the dialogues Cicero wrote after the Civil War. Elsewhere, as in the letter to Quintus, what seemed notable to Cicero about Aristotle’s dialogues was that the author himself was the protagonist in the conversation. Too little of Aristotle’s dialogues remains, however, for us to know how accurate Cicero’s descriptions of them were. The three characteristics identified by Cicero (authorial preface; author as a principal speaker; use of opposing set speeches) are all found in Cicero’s later philosophical dialogues, particularly Academica, De finibus, De natura deorum, and De divinatione, but only the first (and in part the third) applies to the earlier De oratore and De re publica. Even though these two groups of dialogues are formally very different, however, Cicero at different times described both as “Aristotelian.”
Heraclides, on the other hand, plays a very small role in Cicero’s concept of dialogue: there is no reason to think that Cicero had thought about Heraclides as a potential model before Sallustius brought his name up.13 At the same time, however, although Cicero clearly thought of all his dialogues as in one respect or another Aristotelian, when he discussed the possible revisions to the setting of the Academica nine year later (Att. 13.19.4), he compared the use of antiquae personae in De oratore and De re publica to Heraclides, and dialogues in which the author plays the leading role to Aristotle.
Heraclides’ name was thus clearly a useful shorthand for talking about dialogues set in the past, as Sallustius’ reported comments also seem to indicate. What is not so clear is whether Heraclides was actually one of Cicero’s models. In some modern accounts all Cicero’s dialogues, both those of the 50s and those of the 40s, can be categorized as either Aristotelian (set in the present, with the author as protagonist) or Heraclidean (set in the past, with the author not a participant),14 but even if that division makes some sense, there is no evidence that it was in Cicero’s mind before Sallustius put it there—after Cicero had completed De oratore and written a substantial amount of De re publica. Sallustius’ remark may even have been a joke (or at least a tactless comment): apparently, he means to suggest that Heraclides is not an author with whom Cicero would wish to be associated. Nevertheless, Heraclides is worth discussing at least briefly, both because Cicero uses his name to describe his own dialogues and because his work illustrates very clearly the complexity of the Greek tradition of philosophical dialogue.
Heraclides was a pupil of Plato and a contemporary of Aristotle; he wrote a great deal on many philosophical and historical subjects, and the list of his writings in Diogenes Laertius is long.15 That list is also probably inaccurate: certainly some known works of his are not included, and the statement that all Heraclides’ works were dialogues is either an error (by Diogenes or by his source) or textually corrupt. Diogenes describes some of Heraclides’ dialogues as being comic and others tragic; he also says that he used a middle style when the speakers were serious people like philosophers and statesmen.16 Whatever Diogenes means by this, at least some of Heraclides’ works clearly included people of the same sort as the characters in Cicero’s dialogues. But there, so far as one can tell from the exiguous fragments, any similarity ends.
Four works by Heraclides or reasonably ascribed to him offer a useful sample; they seem to be dialogues and seem to concern historical figures, and hence they may be comparable to Cicero’s dialogues. The one best attested—although there is considerable uncertainty about attribution of fragments to specific works—is given two different titles in the sources, either Περὶ νόσων, “On diseases,” or (more probably correct) Περὶ τῆς ἄπνου, “Concerning the woman who was not breathing.”17 It appears to be a conversation taking place on the last day of Empedocles’ life, and part of it concerns his cure of the woman in question; but it also includes his friend Pausanias’ description of Empedocles’ fiery end, as well as a conversation between Pythagoras and Leon, the ruler of Phlious. Gottschalk (1980: 17) reasonably describes it as a blend of Symposium and Phaedo—the last day of the philosopher combined with a banquet including long speeches. And although the speakers were real people, the dialogue as a whole presents an unlikely combination of speakers and circumstances and clearly contained something of the miraculous. Similarly On the soul (Περὶ ψυχῆς), according to Proclus’ summary, told the story of Empedotimus (perhaps, but not necessarily, a genuine person) who, while hunting, had a vision of Pluto and Persephone and learned about the soul and the afterlife; but the same work, apparently, also included some mention of the sack of Rome by the “Hyperboreans.”18 An unnamed dialogue (fr. 139S) has a magus give a report about his circumnavigation of Libya to the court of the Sicilian tyrant Gelon. And a papyrus that some have attributed to Heraclides (fr. incert. 155S)—and which certainly seems like him—contains a conversation taking place in Pisistratus’ Athens, including an erotic tale, Solon, and a story about Periander and the Cypselids in Corinth.19
Heraclides’ works thus seem to have contained a healthy portion of fantastic and exciting tales along with a certain amount of eschatology, medicine, and other subjects; Cicero later (ND 1.34) described his works as stuffed puerilibus fabulis. They do contain historical figures, and the vision of Empedotimus belongs in the category of eschatological visions that includes Plato’s Myth of Er and the Somnium Scipionis—and Cicero certainly knew a great deal about that.20 But while Heraclides’ dialogues included real people, they do not seem to have emphasized real events and realistic combinations of interlocutors meeting in real historical circumstances. It seems fair to say that Cicero knew them, and perhaps enjoyed them, but Sallustius’ comparison of the draft of De re publica to Heraclides was probably meant to be dismissive, and one may suspect that Cicero took up the designation of “Heraclidean” in an ironic and self-disparaging way. There is no reason to think that by using the word Cicero meant any more than a dialogue in which the author took no role—and that is of course true of Plato just as much as it is of Heraclides.
There may well have existed some Hellenistic dialogue of which Cicero was aware that combined a personal preface with a dramatic setting in the past. Heraclides may even have written one, but his dramatic settings seem to have been highly artificial, and he is very unlikely to have tailored them so carefully to the known character and interests of the speakers as did Cicero in both De oratore and De re publica. No plausible Greek precedent exists or is known for the kind of dialogue Cicero created, combining prefaces in the author’s voice with a historical setting anchored in a precise moment of the past. What is more, there is no plausible Latin precedent either, and evidence for dialogue of any kind in Latin earlier than De oratore is very sparse.21 Some of Varro’s Menippean Satires were certainly dialogues, although they are too fragmentary to know much about their shape or date; his even more fragmentary Logistorici may well have been dialogues, and they appear to have been written at roughly the same time that Cicero was writing dialogues.22 Varro’s De re rustica, in three books, each with a different setting and date, was not written until after Cicero’s death.
In fact, we have definite knowledge of only one work written in dialogue form in Latin that must be earlier than De oratore, and it is, curiously, discussed in De oratore itself.23 In a trial in which the opposing speakers were Lucius Crassus, the protagonist of De oratore, and Marcus Brutus, the disreputable son of one of the great jurisconsults of the second century, Brutus seems to have quoted from two speeches of Crassus that offered contradictory views. Crassus responded, as described in Caesar Strabo’s discourse on humor (2.223–24), by quoting the opening sentences of the three books of the elder Brutus’ work De iure civili: it was cast as a dialogue between Brutus and his son, and each book began by describing the two of them talking together in a different property owned by the father, one at Privernum, one at Alba, the third at Tibur—all of which had been sold because of the younger Brutus’ profligate ways. The younger Brutus had similarly had to dispose of baths that he owned—and Crassus’ joke is that it was a good thing that there had not been a fourth book of his father’s work set in the baths to record yet one more piece of extravagance (2.224). That is all we know of Brutus’ work. It is unlikely to be the same as his collection of responsa; it was a dialogue, using real Roman settings, with real characters, with the author himself as the main speaker,24 and the fact that Cicero mentions it in De oratore suggests that he had it in mind, at least to a small degree, as a previous instance of a Roman author putting a dialogue in a Roman setting.
To call these Ciceronian dialogues “Platonic,” then, is misleading, although I will continue to do so to distinguish them from Cicero’s later dialogues. They are, from Cicero’s point of view, Aristotelian dialogues, by which, in all probability he means two things in particular. One is the feature he singles out in these letters, namely, that Aristotle wrote prefaces in his own voice; therefore Cicero thought it quite conceivable that he could somehow address, or speak about, Varro in such a preface without violating what he thought were the rules of Aristotelian dialogue. The other is that, unlike Platonic dialogues, Cicero’s dialogues are generally expository: genuine conversation largely serves to make transitions between parts of the argument, and instances of extended Socratic dialogue such as that between Scipio and Laelius at De re publica 1.54–63 stand out for their rarity. In De oratore Cicero clearly drew (either directly or through an intermediary) on Aristotle’s Rhetoric, and De re publica reflects Peripatetic and perhaps specifically Aristotelian works of political theory. But that is not why he thought of them as Aristotelian; it was a formal relationship, not necessarily a substantive one.
The role of Plato in Cicero’s dialogues is something rather different, and it is an issue that will be addressed a number of times in this book from different directions. For the moment, in looking for Cicero’s models, it is important to realize that the use of Plato is not at all obvious, and in fact is somewhat surprising. In what survives of Cicero’s writings earlier than De oratore and De re publica (in other words, before 55), I can find only three references to Plato: once in Pro Murena (63), when he is describing his own moderate Academic philosophy (compared to Cato’s rigid Stoicism) as derived from Plato and Aristotle; once in the famous letter to Atticus (2.1.8, from 60 bce) describing Cato as seeming to speak in Plato’s Republic rather than in Romuli faece; and once, in his letter to Quintus about good provincial governance (QFr. 1.1.29, also from 60), where there is a broad reference to the philosopher kings in Plato.25 But in the period Cicero was working on our dialogues, Plato appears more often, four times (outside the dialogues) in the year 54: a reference in Pro Scauro (4–5) to the Phaedo, in connection with the suicide of Cleombrotus of Ambracia described by Callimachus; a description in Pro Rabirio Postumo (23) of Plato’s unfortunate relationship to Dionysius of Syracuse; and the two references already discussed, in the letter to Lentulus and Att. 4.16, both from 54. Granted, one would not expect Plato’s name to turn up in speeches with great frequency, but it is at least curious that while Plato is mentioned in only one speech before 55 he appears in two in 54 alone. Plato is never mentioned in De inventione, but Socrates is: the anecdote referred to earlier was taken from a dialogue not by Plato, but by Aeschines Socraticus (1.51). And while Plato is absent, Aristotle is mentioned in De inventione seven times.
That does not mean that Cicero had not read Plato; he obviously had. One plausible estimate is that he knew most of the major dialogues, with the exception of Parmenides, Sophist, and Politicus. He translated the Protagoras (four fragments survive) and he at least began a Timaeus based on Plato’s (and with some close translation) sometime in or after 45 bce.26 Close translations or paraphrases of Plato are frequent in Cicero’s philosophical works, and he had clearly read and understood most of Plato. Although in his later works he discusses and uses some of Plato’s ideas—using the theory of forms to construct the idea of eloquence in the Orator, for instance—his more detailed and technical philosophical writings draw much more on Hellenistic sources—and that is true, to a high degree, of De oratore and De re publica as well.
What Cicero consistently admired about Plato was not his philosophy but his style: Plato and Demosthenes are the two Greek stylists whom he most hoped to emulate, and when he calls Plato divinus he is not talking about his philosophy.27 Aside from Cicero’s own lack of interest in Platonic philosophy, moreover, Plato, in the middle of the first century bce, was probably not an obvious model for doing philosophy in Rome or anywhere else. While other philosophical schools studied the texts of their founders, the skeptical Academy was devoted to refuting others’ arguments—in what they claimed was a Socratic manner—rather than expounding Plato’s beliefs.28 No commentary on Plato’s dialogues is known before the late first century; and it is only with the change from the skeptical “New” Academy of Carneades and Philo to the more doctrinally based “Old” Academy of Antiochus that a serious interest in explaining and maintaining Plato’s text begins. Cicero was familiar with both styles of Academic philosophy, and while he seems to have remained a consistent follower of Philo,29 he made heavy use of Antiochus’ interpretation of the history of philosophy, and in the final version of the Academica he had Varro present an Antiochean epistemology.
The discussion above of Greek forms of dialogue has been something of a dead end, but that is the point: the form of De oratore was something new. We can say that not only because there is no known precedent, but because of what Cicero himself says and does not say about it: he calls the dialogue Aristotelian in 54, he calls it Heraclidean in 45, and yet we know enough about the dialogues of Aristotle and Heraclides to be quite certain that it is neither. There were clearly precedents in Plato and Aristotle in particular for the individual major structural elements of Ciceronian dialogue—the use of personal prefaces, the use of formal expository speeches and debates, the use of historical characters, the careful chronological and physical setting—but no previous work had combined them in the way Cicero did. Given that De oratore is Cicero’s own creation, then, one needs to consider the obvious next question: why did he create this kind of dialogue? In that respect, what Cicero says about it in his prefaces and what Sallustius seems to have recognized as distinctive—the close connection that Cicero emphasizes between the author and his characters, between the present moment of writing and the historical setting of the dialogue—is of critical importance. Cicero shaped the dialogue in the way he did because its form has a direct bearing on the interpretation as well as the presentation of its content.
Cicero’s combination in De oratore of a personal voice with a historical setting establishes a strong relationship between his own present moment and the scene in September of 91 bce that he so vividly evokes. Even more important, it permits him to create a sense of historical depth, of development, and of direction. Aristotelian dialogue, with the author as its main speaker, exists in an eternal present: it is the ideas and arguments that matter, not the characters presenting them (except, of course, for Aristotle himself). Platonic dialogues pay far more attention to the relationship between the speakers and the arguments they are made to present—the choice of Gorgias to present the argument for sophistic rhetoric is not an accident—and they take place in the historical past;30 but they are like Aristotle’s dialogues insofar as they exist only within their own dramatic time: although the speakers are (mostly) historical characters, Plato himself never appears in them; on the rare occasions when there are narrative introductions, they are delivered by someone other than Plato; only very rarely and obliquely do they even point to events subsequent to the dramatic date of the dialogue.31 Although we know of the relationship between Socrates and Plato, Plato in fact says almost nothing about it and makes no attempt at all to tie himself to the speakers in the dialogues. But linking the (Aristotelian) present to the (Platonic) past creates a very different effect: Cicero explicitly links his present to the historical moment of the dialogue; he compels the reader to understand his personal ties to his characters—and within that frame he also allows those characters to explore their own relationship to a deeper past. Thus, in the opening scene, he presents a group of Roman senators self-consciously sitting under a Platonic platanus, talking about Roman oratory, and placing it in at least three frameworks at the same time: Platonic dialogue, the history of rhetoric, and the development and practice of oratory in Rome itself.32 De oratore as a whole, as we know to be true for Scaevola, is both analogical and archaeological: it is constructed on the analogy of Platonic dialogue, but it is also an archaeology of oratory (and indeed of the present condition in general) leading back to Platonic dialogue, to Socrates, and beyond. That archaeology and the concomitant sense of time, memory, change, and loss, pervade the dialogue from its first sentence to its conclusion; but nowhere is it more apparent than in Cicero’s choice of characters and setting.
1. For the correspondence, see Att. 13.12–16, 18–19.
2. On the interpretation of this passage, see also Stull 2011: 247–49.
3. On details of Cicero’s technique and methods in structuring the dialogue and setting of De oratore, see LPW 1: 76–84. The most valuable studies of ancient dialogue as a whole are Hirzel 1895 and Hösle 2012; Hirzel is more historical and more attentive to the history of dialogue between Plato and Cicero. Most recent scholarship on dialogue, and on Ciceronian dialogue in particular, provides broad analysis rather than detailed examination of individual works; the most valuable are Auvray-Assayas 2001, Schofield 2008, Gildenhard 2013b, and Auvray-Assayas 2015. Of these, the first two focus on the later dialogues; Gildenhard 2013b rightly emphasizes the importance of the dialogue form for Cicero as a (fictional) form of history and as an educational tool for performing the synthesis of Greek philosophy and Roman practice, while Auvray-Assayas 2015 has useful comments about De oratore in particular. Other important treatments are Becker 1938 and Zoll 1962: 25–37, and on De oratore in specific Steidle 1952. Rawson 1972: 39–42 = Rawson 1991: 70–75 discusses the importance of historical detail in Cicero’s dialogues, and Spahlinger 2005: 168–72 supplies useful bibliography.
4. So, e.g., Schofield 2017a: 53, who dismisses without discussion Cicero’s repeated references to Aristotelian dialogue. Hösle 2008 and 2012: 87–94 concentrates almost exclusively on direct comparison between Cicero and Plato. Auvray-Assayas 2001 rightly emphasizes the originality of Cicero’s treatment of Plato in a Roman context; on Romanization see also Becker 1938: 11–24. Atkins 2013 is the most substantial recent attempt to read Cicero’s dialogues (notably De re publica and De legibus) against Plato.
5. De orat. 1.28, 1.47, Rep. 1.66–67, respectively. A fragment of De re publica (1 fr. 4) in which (in Pliny’s paraphrase) Cicero says that he is Plato’s comes is generally assigned to the preface, but it could have been spoken by a character within the dialogue. De legibus has no preface, and it is the character “Marcus,” not the author Cicero, who refers to Plato’s Laws (Leg. 1.15, 2.14 and elsewhere).
6. Zoll 1962: 60–63 rightly observes the striking difference between the internal references in the dialogues to Plato as model and the external ones to Aristotle.
7. On pre-Platonic Socratic dialogues and their importance, see Hirzel 1895: 1.83–174; Kahn 1996: 1–35. On the popularity of Xenophon in Rome, see also Fantham 2004: 51.
8. On Aristotle’s dialogues, see Hirzel 1895: 1.272–300. For an attempt to identify traces of On justice in De re publica, see Moraux 1957: 65–79.
9. Cf. Tusc. 1.21 (fr. 19M), 1.77 (fr. 27M); fragments of Dicaearchus are cited from Mirhady 2001; on Dicaearchus, see also Hirzel 1895: 318–20. Cicero refers to at least nine works by him and used his writings in his discussions of the active and contemplative lives (see McConnell 2012 and 2014: 115–60); on Cicero’s use of Dicaearchus’ political writings, see also Smethurst 1952. He closely followed Dicaearchus in his account of the site of Rome: see below, Chapter 11.
10. Cf. Sen. 3, with Powell 1988 ad loc.
11. See further below. According to Diogenes Laertius (3.8), the Peripatetic Praxiphanes described a conversation between Plato and Isocrates about poetry at a country villa; cf. Hirzel 1895: 310. Hösle 2008: 162 is wrong to call Praxiphanes “the inventor of the location of the Ciceronian dialogues,” as there is no evidence at all that Cicero knew his work, and the villa location is far more likely to be drawn from Brutus’ dialogue (see below in this chapter), if it needs a source at all—and the jurist Brutus is not likely to have read Praxiphanes.
12. The next paragraph of the letter (to be discussed in connection with the composition of De re publica) shows that Cicero paid only brief attention to Sallustius’ criticism.
13. So, rightly, Zoll 1962: 69 n. 27; Schmidt 1969: 30–31.
14. So in particular Schmidt 1969.
15. The most important modern study of Heraclides is Gottschalk 1980, to which I am greatly indebted; there is still much of value in Hirzel 1895: 321–31. I cite the fragments from Schütrumpf et al. 2008. On Cicero’s use and opinion of Heraclides, see also Zoll 1962: 68–72.
16. D.L. 5.86–89 = fr. 1S; cf. Gottschalk 1980: 6–8.
17. Fr. 82–95S. See Gottschalk 1980: 13–36; also van der Eijk 2009.
18. Fr. 47–58S. See particularly Kupreeva 2009.
19. The papyrus is POxy. 664 + 3544; for recent discussion and bibliography, see particularly Dorandi 2009:15–19; Fox 2009: 60–61.
20. It is possible that Cicero took the Milky Way as the abode of souls from Heraclides: see fr. 50S. Varro also knew the work On the soul; frr. 51, 57S.
21. Cf. Hirzel 1895: 421–57; Zoll 1962: 37–41.
22. Hirzel 1895: 329–31 believed that the Logistorici were dialogues in the manner of Heraclides; the best modern account of them is Dahlmann and Heisterhagen 1957.
23. On Brutus’ dialogue, see Hirzel 1895: 428–32. It is also mentioned earlier by Cicero in Pro Cluentio 140–41. One other possible precedent (cf. Hirzel 1895: 455–57; Fantham 2004: 51) is a dialogue written by the elder Scribonius Curio with his son and Vibius Pansa as interlocutors with a dramatic date of 59; but Cicero’s criticism of it for anachronism in discussing Caesar’s later actions in Gaul (Brutus 218) shows that it could have been written after De oratore (but before Curio’s death in 53).
24. Our knowledge of his work is very slight; there are a few citations in the Digest and other sources, but none gives a title. Pomponius’ account of the early jurists (Dig. 1.1.2.39) ascribes seven books to Brutus and three to Manilius, but the reversal of those numbers (by Bertrand according to Mommsen’s apparatus; by Zimmern according to Lenel 1889: 1.77) is plausible, as we know that De iure civili had three books. That does not mean, however, that it was the same as the collection of his responsa, which according to Cicero (De or. 2.142) was simply a transcript, arranged by the name of the client.
25. In Commentariolum Petitionis 46 there is a reference to Cicero as homini Platonico, in that he is likely to balk at having to flatter or be less than honest to potential voters. But (a) this is clearly a joke, and Plato is here simply as a representative of philosophical uprightness; (b) it is very uncertain that this text was written by Quintus in the mid-60s; and (c) I suspect that the author (not Quintus) took the Platonic reference from Cicero’s annoyed reference to Cato in Att. 2.1.8.
26. The Protagoras translation has been considered both early and late; there is no evidence, and Cicero did translations as a rhetorical exercise throughout his life. On Cicero’s knowledge of Plato, see Long 1995: 43–44; see also DeGraff 1940 for a fuller account of Cicero’s references to Plato. Neither pays much attention to chronology. On the dating of Cicero’s Timaeus see Sedley 2013.
27. Thus at Att 4.16.3 (quoted earlier) deus ille noster Plato refers to his literary genius, not to his philosophy. So also, for instance, in the letter to Lentulus (Fam. 1.9.12), apud Platonem nostrum scripta divinitus (and compare De or. 1.49; Opt. Gen. 17, Sen. 44). At Tusc. 1.79, C. cites Panaetius as disagreeing with Plato even though he calls him divinus. Stull 2011 rightly points to the transfer in De oratore of deus/divinus from Plato to the Roman protagonists of De oratore as part of Cicero’s deliberate transformation of Plato (starting from the two uses of divinitus of the political conversation at 1.26 and of Socrates at 1.28), although I would place less emphasis than he on the substantive philosophical differences and more on matters of style. On the complexities of Cicero’s divinization of humans, see particularly Cole 2013.
28. On Plato in the first century, see particularly Sedley 1997, to which this paragraph is greatly indebted. Zoll 1962 assumes a wider familiarity with Plato, particularly as filtered through Panaetius and Posidonius; but that does not mean that many people (particularly Romans) actually read him. See further below Chapter 5 on the scenes in De oratore involving study of the Gorgias.
29. This has been doubted, but without good reason; cf. Görler 1995. For a good summary of the debate, see Gawlick and Görler 1994: 1085–89.
30. With, it should be noted, some stretching of historical plausibility, as in Protagoras, and no attention to geopolitical realities. As Hirzel 1895: 185–86 noted: “Das platonische Athen ist ein ganz neutraler Boden.”
31. For discussions of Platonic dialectic that help explain Plato’s absence from his own dialogues, see, for instance, Edelstein 1962, Frede 1992, Gill 1996 (with very useful analysis of approaches to Platonic dialectic). The literature is vast, and these are no more than a few articles that I have found helpful.
32. It terms of style and presentation, of course, Cicero’s dialogues—the setting, the choice of speakers, the structure of social interaction—are much more Roman than Platonic; for a good discussion and collection of examples, see Becker 1938: 11–25. What in fact may be the most Platonic aspect of De oratore is the urbane dissimulatio of its characters, a Roman version of Socratic irony; on this see particularly Zoll 1962: 105–24. Zoll also rightly points out (123) that in De oratore the Socrates-figure Crassus is actually an anti-Socrates in his belief in a kind of sophistic Allgemeinbildung; one might think of this as Ciceronian rather than Socratic irony.