Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

Transcription

1 The Dramatic Date of Plato's Republic Author(s): Debra Nails Source: The Classical Journal, Vol. 93, No. 4 (Apr. - May, 1998), pp Published by: The Classical Association of the Middle West and South, Inc. Stable URL: Accessed: 28/01/ :18 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org. The Classical Association of the Middle West and South, Inc. is collaborating with JSTOR to digitize, preserve and extend access to The Classical Journal.

2 THE DRAMATIC DATE OF PLATO'S REPUBLIC T he scholarly debate that followed Boeckh's 1874 argument that 411/410 is the dramatic date of Plato's Republic1 has led not to consensus but to a contemporary standstill. Recent translators and commentators who are not silent on the issue typically provide, and with little or no comment, one of two widely held dramatic dates, Boeckh's 411/410 or the earlier 422/421,2 but there have been scores of arguments bearing on several suggested dramatic dates for the Republic ranging from 424 to Guthrie declares the date uncertain,4 and Moors argues against both dates that Plato is deliberately producing a "timeless" dialogue.5 1 August Boeckh, Gesammelte Kleine Schriften (Leipzig 1874) For 411/410, cf. Lewis Campbell, Plato's "Republic" (Oxford 1894) 3.2; Paul Shorey, Plato The Republic (London 1930) introduction; Benjamin Jowett, The Dialogues of Plato translated into English with Analyses and Introductions (Oxford 1953); Eric Voegelin, Order and History (Baton Rouge 1957) 3.53 n. 4; and Allap Bloom, The Republic of Plato (New York and London 1968) 440 n. 3. For 422/21, cf. D. J. Allan, Plato Republic Book I (London 1940) 20; A. E. Taylor, Plato, the Man and his Work (Cleveland 1956) 264; Desmond Lee, Plato: The Republic (London and New York 1955) 60; Jacob Howland, The Republic: The Odyssey of Philosophy (New York 1993) xii-who specifies 421/420; and Robin Waterfield, Plato Republic (Oxford 1993) 380, with certain reservations. For convenience, I will refer to these dates as 411 and 421 below. 3 At the two extremes, H. D. Rankin, Plato and the Individual (London 1964) 120, supports 424; and Eduard Zeller, Plato and the Older Academy (London 1876), supports 409/408, as does J. Adam, The Republic of Plato (Cambridge 1926), who says "perhaps 409." Kenneth J. Dover, Lysias and the Corpus Lysiacum (Berkeley 1968) 53, identifies a window of on grounds that will be pivotal below. 4W. K. C. Guthrie, Plato the Man and His Dialogues (Cambridge 1975) , systematically considers dramatic date claims for each of Plato's dialogues. His word "uncertain," however, suggests that Plato intended a dramatic date, but that moderns cannot be sure what it was. I maintain rather that Plato never "edited" the dialogue from the standpoint of dramatic date at all. 5 Kent Moors, "The Argument Against a Dramatic Date for Plato's Republic," Polis 7.1 (1987) 6-31& 22; this amounts to a stronger claim about Plato's literary technique than that advanced by E. R. Dodds, Plato's Gorgias (Oxford 1959) 17-18, for the Gorgias (see below). Moors (24 n. 7) cites disputants not included by Guthrie. For general studies on dramatic order, see Diskin Clay, "Gaps in the Universe' of the Platonic Dialogues," Boston Area Colloquium on Ancient Philosophy 3 (1987) ; The Classical Journal 93.4 (1998)

3 384 One reason for current interest in the dramatic date of the dialogue is that the well-known "developmental hypothesis" in the philosophical interpretation of the dialogues has receded in recent years.6 According to this interpretive strategy, the dialogues' dates of composition are mapped against Plato's evolution from a mere follower of Socrates to an original philosopher in his own right: that is, early, middle and late dialogues map the rise and fall of the theory of forms. So long as this hypothesis held sway, the supposed order of composition dictated the order in which the dialogues were read and taught.7 Now, with developmentalism increasingly under attack,8 new and renewed approaches to the Platonic corpus are emerging. Increased interest in the dialogues as literature has made the attempt to establish a dramatic date for the Republic more urgent,9 particularly because its dramatic date impinges on those of several other dialogues.?10 What makes the and Charles L. Griswold, "Irony in the Platonic Dialogues," in The Sovereignty of Construction: Studies in the Thought of David Lachterman (Amsterdam 1996). 6 The position has a long history, beginning with Karl Friedrich Hermann, Geschichte und System der Platonischen Philosophie (Heidelberg 1839), and strongly influenced early on by the stylometric work of Lewis Campbell, The Sophistes and Politicus of Plato (Oxford 1867). It is closely associated nowadays with the writings of Gregory Viastos and those of his many followers. Vlastos's most complete defense of the view appears in Socrates: Ironist and Moral Philosopher (Cambridge 1991). 7 Cf. Jacob Howland, "Re-reading Plato: The Problem of Platonic Chronology," Phoenix 45 (1991) Direct and recent attacks have been made by Holger Thesleff, Studies in Platonic Chronology (Helsinki 1982) 8-17, 40-52; and Debra Nails, Agora, Academy, and the Conduct of Philosophy (Dordrecht 1995) both of which cite numerous precedents. Charles Kahn, who remarks in Plato and the Socratic Dialogue (Cambridge 1996) on his having thought innocently in 1991 that the developmental view of the earlier dialogues "had by then collapsed of its own weight" (xviii), remains remarkably developmentalist. John Cooper, Plato Complete Works (Indianapolis 1997), characterizes the developmental hypothesis charitably but concludes that it is "an unsuitable basis for bringing anyone to the reading of these works" (xiv). 9 In "The Origins of the Socratic Dialogue," in The Socratic Movement (Ithaca 1994), Diskin Clay writes, "More than any of the literary Socratics of the fourth century, Plato took care to provide some of his Socratic dialogues with a significant historical setting" (46). Kahn, Plato, ch. 1, echoes the compliment, crediting Plato with "invent[ing] the conversations of Socrates with the same freedom as other Socratic authors, but... more philosophically and more lifelike" (35). The paucity of extant fragments, much less whole texts, of the several writers of Sokratikoi logoi makes me reluctant to draw such conclusions myself, and still more reluctant to infer from them any consequence for reading the Platonic dialogues. 10 Implications for the Phaedrus and Symposium will be examined below. But there are others. The Timaeus, for example, has long been viewed as taking place the day after the Republic, and the Critias immediately after that; Guthrie, The Later Plato and the Academy (Cambridge 1978) 198, citing precedents, argues that the

4 THE DRAMATIC DATE OF PLATO'S REPUBLIC 385 issue even more critical is that the two dates most commonly defended suggest conspicuously dissimilar images of the conversation held in the Piraeus that summer day, and thus have different implications for how the dialogue as a whole ought to be understood. If the gathering took place in 421, Socrates ( ) was fortyeight, and Plato's brothers as well as the sons of Cephalus were all in their early to mid twenties; the optimistic springtime Peace of Nicias was a few months old in a war that had been going on sporadically for ten years, and which Athens still fully expected to win. If the date is a decade later, in 411, Socrates was fifty-eight, and because a different set of historical events and relationships comes into play, Cephalus's sons were in their mid forties, while Glaucon and Adeimantus were in their early twenties; more strikingly, twenty years into the Peloponnesian War, the expedition to Melos, Alcibiades's betrayal of Athens, and the humiliating Sicilian defeat had already been suffered, and Athens was living under the Four Hundred, then the Five Thousand, in the year before the democracy regained power-a gloomy, violence-torn, and pessimistic time. The earlier date has stimulated interpretations that point to more irony, more humor, in the dialogue than is found by those whose belief in 411 encourages them to find more allusions to Socrates's death and a general preoccupation with destruction and corruption. While a plausible case can be made for each date, as I will demonstrate below, there are problems with each as well. Ultimately, the strong evidence for two dates is better construed as evidence that Plato's great dialogue was cobbled together and revised over decades. I. INTRODUCTION The raw material for deciding between the two dates is chiefly historical and prosopographical. Besides conflicting details from various sources that will be considered below, two particulars from the first book of the Republic must fit the date: the peaceful summer during the Peloponnesian War years in which the conversation Philebus also takes place in the wake of the Republic. The dramatic date of the Symposium (416) puts it either soon after, or just before, the Republic, dependent on whether 421 or 411 is established; the Laches ( ), and both the Hippias Major and Hippias Minor ( ) are similarly affected. Cf. for Laches, Thesleff, Studies, 93-94; for Hippias Major (and thus Hippias Minor), Paul Woodruff, Plato Hippias Major (Indianapolis 1982) introduction.

5 386 explicitly takes place (350d), and the new festival honoring the Thracian goddess Bendis (327a, 354a), with which the Republic begins; and one item from the second book: a battle at Megara must have taken place recently (368a). Biographical details that we know independently-information about Plato's brothers, the family of Cephalus, Thrasymachus, and Socrates-must fit as well. Let us begin with a simple enumeration of the kinds of problems that beset us when we attempt to fix a dramatic date for the Republic. First, there are precedents of all sorts in the Platonic corpus: for exact dramatic dates, for deliberately indeterminate ones, and for impossible ones. Second, in part because he lacked our current categories of historical realism and poetic license, Plato's works are full of anachronisms. Third, it is the characters and events established in Books I and II-and their incompatibilities-that most strain efforts to set a dramatic date for the work as a whole. Fourth, insofar as Book I may have stood as a separate and aporetic dialogue, and insofar as a proto-republic may have been known in Athens in the 390s, a single dramatic date may recede from our grasp. The first two points can be treated by way of introduction, and the third will provide defenses of the two proposed dramatic dates; aspects of the fourth will then be sketched briefly. For a familiar example of an exact dramatic date in the Platonic corpus, without resorting to the dialogues set in 399: "the occasion portrayed in Symposium is Agathon's first theatrical victory, gained in l Yet there is precedent as well for carefully setting a dialogue "in no particular year," as Dodds argues is the case for the Gorgias. It is not that Plato gives no hints about dramatic date in that dialogue, it is that he gives concrete evidence for at least seven different dates stretching from 429 to The dramatic date of the Lysis is indeterminate in a less radical way. As Guthrie puts it: "There is nothing to indicate the dramatic date, nor is it important. At the end (223b), Socrates describes himself as an "old man," but since he is talking, not very seriously, to two schoolboys of twelve or thirteen, one cannot attach much weight to this."13 Finally, there is the baffling Menexenus with its impossible date. Guthrie again: "This is the shock. It is Socrates who recites the Kenneth J. Dover, "The Date of Plato's Symposium," Phronesis 10 (1965) 2-20, Dodds, Plato's Gorgias, Guthrie, Plato the Man, Guthrie, Plato the Man, 313. He adds, "It is also unlikely that Aspasia, the supposed author of the speech, was still alive. She bore a son to Pericles about 440 or earlier." Aspasia, if living, would have been in her late seventies or eighties.

6 THE DRAMATIC DATE OF PLATO'S REPUBLIC 387 speech, but the Peace of Antalcidas was concluded twelve or thirteen years after his death."14 The Republic may have a dramatic date, or several, or none; precedents cannot help us here. Anachronisms are a closely related problem. I have just presented information about the dramatic dates of a few dialogues as if Plato attended carefully to the notion of dramatic date and controlled it as skillfully as he did other literary devices. If that assumption is given free rein, as it sometimes has been,15 interpreters holding a 411 date may gasp and get goose flesh that Plato raises Cephalus from the dead to play a role in Republic I. Perhaps it is a stroke of literary genius to have a really dead man speak of dying conventions, but perhaps it is just an anachronism. Republic I includes at least two anachronisms that are no sparkling literary gems: Ismenias the Theban is used as an example of a corrupt rich man (336a) though it was not until after Socrates's death that his iniquities, taking bribes from the Persians, began. And the athlete Polydamas is hauled in (338c) as someone whose massive physique requires meat, but his fame for victory in the pancratium occurs only in On the face of it, Plato does not appear to have been especially concerned with historical accuracy, and that renders suspect facile claims about his deploying time for literary purposes. Such manipulation achieves its effect only if there is a nascent historical realism to play it against. Otherwise it goes unnoticed. II. THE EVIDENCE As Figure 1 shows, both dates meet the interval of peace criterion, albeit very differently: 421 because the Peace of Nicias was in effect following the Archidamean War; the appearance of Niceratus, son of Nicias, as a minor character in the Republic may be in honor of Nicias. But the year 411 also meets the criterion because Athens was under the rule of the Four Hundred, and then the Five Thousand-terror, political assassination, and oligarchic revolution, yes, but no foreign war to draw away Socrates's respondents Moors, "Argument Against," n. 4, argues from Philebus 30c that "the usage of time elements in the dialogues is intentional, as much a part of dialogical form as setting or characters." 16 Allan, Plato, 20 and 92. In antiquity, the Babylonian Herodicus, called the "Cratetean" (second century B.C.E.), was also much concerned about Plato's anachronisms; cf. I. During, Herodicus the Cratetean: A Study in Anti-Platonic Tradition (Stockholm 1941). 17 Thucydides ; his narrative breaks off in the middle of the year.

7 388 The inaugural festival of Bendis accommodates 421 marginally better than it does 411 but it does not accommodate either especially well. The religion of Bendis may have been known in Athens as early as 443, though Shorey was persuaded by "inscriptions to prove its establishment in Attica as early as 429/428.'18 Others, Dover for example, are less sanguine about the value of the inscriptions since the date itself rests on an emendation.19 It is plausible in any case that the first formal celebration in the Piraeus was held a few years later in 421. But it is likewise plausible, as Mommsen held, that Socrates was referring to special ceremonies introduced in 411/ 410. The depleted Athenian treasuries were in fact filling up again after so long at war, and religious festivals were early beneficiaries of the new revenue. So a torch-race on horseback fits both an inauguration in 421 and a splendid revival resembling an inauguration around 411. Moors has found in Thucydides a factor that adjusts the balance in favor of 421:20 Thracians were participants in the festival, and Bendis was their goddess. Whereas the period around the Peace of Nicias in 421 was a time of alliances between Athens and various Thracian cities, by 411 Sparta was pressing against the Hellespont and Thracian regions, making it less likely that a contingent of Thracians could afford to leave their homes vulnerable in summer, the war season. So far, both dates are possible, and 421 has a slight edge. Let us now consider evidence based on the ages of the participants, using precision in the service of clarity, however unwarranted by the available sources: Socrates was forty-eight in 421 and fifty-eight in 411. At either age he might engage the extremely old Cephalus (328b) in a discussion of what the last part of the road of life might hold, and there is not much in his further words or behavior to go on. One might examine whether his behavior closely resembles that in any of the dialogues for which a dramatic date is firmer: in the Protagoras he is about thirty-seven, in the Laches perhaps forty-five, in the Symposium, fifty-three, in the Ion sixty-four.21 Despite the firmer dramatic dates, the tracing of the character Socrates's subtle 18 T. Kock, Comicorum Atticorum Fragmenta (Leipzig 1880) 1.34, and August Mommsen, Feste der Stadt Athen im Altertum (Leipzig 1898) 490-both cited in Shorey, Plato, 8, nn. c and f, respectively. 19 Dover, Lysias, 31 n Moors, "Argument Against," 10 and nn Dramatic dates used for this gloss are 432 for the Protagoras, 424 for the Laches, 416 for the Symposium, and 405 for the Ion-but the literature on even such "firmer" dates is vast. Cf. note 10 above for Laches and Symposium; for Protagoras,

8 THE DRAMATIC DATE OF PLATO'S REPUBLIC 389 aging across those dialogues, particularly after the Protagoras, is a speculative matter yielding little if anything of substance. What is nonetheless important about Socrates's age is its relation to the ages of the respondents in the Republic, to whom I will now turn. Since Aristophanes lampoons Thrasymachus of Chalcedon in 427 in the Daiteles, the sophist's reputation was well in place by either date.22 Thrasymachus is assigned aflourit of , implying birth around 455, making him some fourteen years or so younger than Socrates. If the dramatic date is 421, Thrasymachus would be thirty-four, some ten years older than the rest of the participants; whereas if the date is 411, he would be forty-four, twenty years older than Adeimantus and Glaucon, but roughly the same age as Cephalus's sons, Lysias in particular, whom he may well have been visiting, perhaps out of their shared interest in rhetoric, for they are linked not only in the Republic, but in the Phaedrus and Clitophon as well.23 Clitophon, a minor character in the Republic, may be assumed to be the same Clitophon as the one linked to Thrasymachus and Lysias in the dialogue bearing his name. So Clitophon, Thrasymachus, and Lysias form a sort of unit in the Republic.24 In the remainder of our consideration of the evidence-which at this point becomes strikingly more controversial-the ages of two groups of brothers will be examined: Cephalus's sons, Polemarchus, Lysias, and Euthydemus; and Ariston's sons, Adeimantus, Glaucon, and Plato. To establish the facts about the life of Lysias, we are fortunate to have Dover's careful study. From the speech of Lysias known as "XII," the following biographical data emerge: Pericles persuaded Cephalus to settle in Attica where he lived for thirty years, dying before the Thirty Tyrants came to power. Lysias's brother, Polemarchus, was executed by the Thirty in 404. The remaining contemporary evidence is more difficult to evaluate, yielding wide open windows rather than exact dates. Alfred Edward Taylor, Plato: The Man and His Work (Cleveland 1956) 236, followed by Guthrie, Plato the Man, 214. The Ion is especially controversial: cf. e.g. W. R. M. Lamb, Plato Ion (London 1925) introduction, who argues for ; Charles Kahn, "Did Plato Write Socratic Dialogues," CQ (1981) , argues for internal contemporary allusions fixing the dramatic date at 394/393, and thus against any dramatic date, "since Socrates was unavailable" (308 n. 9). 22 H. Diels and W. Kranz, Die Fragmente der Vorsokratiker (7th edn. Berlin 1954) fr. A4. 23 Plato, Phaedrus 266c, 269d, and 271a; Clitophon 406a. 24 Cf. Dover, Lysias,

9 390 Dover accepts very little of the secondary evidence, though Plato's Phaedrus he finds useful, as we will see.25 From Dionysios of Halicarnassus, writing in the late first century B.C.E., Dover allows only that the family was originally from Syracuse, that Lysias was born in Athens, and that he was fifteen when he and his brothers went to the Athenian colony at Thurii.26 To allow Dionysios's more precise claims that Cephalus's sons settled in the colony when it was founded in 443, or that Lysias returned to Athens in 412/411, would have had the effect of rendering any dramatic date for the Phaedrus impossible since Phaedrus himself was in exile Dionysios's evidence yields a dramatic date of 411 for the Republic, when Lysias would have been about forty-seven (and his brothers, perhaps, in their mid forties to early fifties). As I mentioned, that would make Lysias a rough age-mate of Thrasymachus. Sifting bits of later evidence, Dover ultimately suggests that Cephalus settled in Athens after Pericles had become prominent, about , Lysias was born about 445, went to Thurii about 430, returned in the late 20s, and Cephalus died For Dover, thus becomes the range for the dramatic date of the Republic.27 By this calculation, Lysias (and his brothers, presumably) are in their mid twenties at the time of the action in the Republic, that is, roughly the ages of Glaucon and Adeimantus. The beginning of Book II provides new dramatic date issues. As Dover puts the problem: "Plato's brothers, Glaukon and Adeimantos, are old enough to have distinguished themselves in "the battle at Megara," but Glaukon, at least, was not too old at that time to have had a "lover" who composed elegiacs in his honour (368a)."28 So when was the battle? A. E. Taylor, preferring a dramatic date of 421, uses Thucydides's battle date of 424,29 whereas Shorey, preferring 411, uses Diodorus Siculus's 409-despite the 25 Dover uses evidence from the Republic to establish a prosopography for Lysias; I omit precisely that evidence below because it would make a very small circle of my argument. When Dover (Lysias, 32-33, & 43) considers the Phaedrus, while conceding that Plato may not have had a dramatic date or historical situation in mind, he goes on to argue from better known data about Isocrates's life that the Phaedrus puts the birth of Lysias "somewhat earlier than 440." When he reaches conclusions about Lysias's life, he assigns a dramatic date of to the Phaedrus-right between the two candidate Republic dates. 26 Dionysios, Lives of the Ten Orators Although he says " " (Lysias, 42) when assigning a plausible dramatic date, he later says " " (53) without further comment. 28 Dover, Lysias, Thucydides 4.72.

10 THE DRAMATIC DATE OF PLATO'S REPUBLIC 391 anachronism.30 Dover, reluctant to turn Plato into a historian, points out that there were several candidate battles, even limiting the search to ones mentioned by actual historians, but that the particular battle need not have been one that made the record books anyway.31 The Thucydidean date of 424, imagining Glaucon then about eighteen, would make Glaucon twenty-one at the time of the Republic, and Adeimantus older than that.32 That would cause a gap of at least fifteen years between the six-year-old Plato and his older brother Glaucon, more than that for Adeimantus.33 Such a gap, while not physiologically impossible, has concerned some authors enough that they have postulated that these men were actually Plato's uncles.34 And Moors has argued that Glaucon was in fact a younger brother of Plato on two grounds.35 First, Moors sees "similarities in character and dialogical structure" between the Republic and the Symposium that lead him to conclude that the Glaucon of the two dialogues is the same person. Second, Xenophon reports that when Glaucon was less than twenty, he made himself a laughing-stock by trying to gain political power, and that Socrates took him under his wing as a favor to Plato and Charmides, implying that Glaucon must have been Plato's younger brother.36 The conversation to be recited in the Symposium took place when Apollodorus and Glaucon were very young, long ago.37 But the date is firmly established as 416, so Moors suggests that Glaucon was born in 420 or later, making him younger than Plato, vindicating Xenophon, and utterly precluding any sort of participation by Glaucon in the Republic, no matter what its dramatic date. That suits Moors because he believes Plato took extraordinary steps to prevent the Republic from being assigned any dramatic date, and this is one of those steps. While evidence is not conclusive for any particular dramatic date, the Glaucon argument will not work. Not only, as Moors concedes, is Xenophon untrustworthy in some matters of 30 Diodorus Siculus, Dover, Lysias, 31. Assumed from Apology 34a. 33 Plato notoriously says at the end of Republic VII (540e-541a) that those over the age of ten are beyond the reach of the guardian-teachers and must be sent out of the polis. If the dramatic date of the dialogue is 421, Plato-born in 427-can stay, but if the dramatic date is 411, it is too late for Plato. 34 Cf. references in Shorey, Plato, n. e. 35 Moors, "Argument Against," Xenophon, Memorabilia, Plato, Symposium 173a.

11 392 detail, there is a problem with the text as well. Moors says Socrates does Plato a "favor," implying that the favor had been requested; but it seems rather that Socrates checked Glaucon out of feeling for (e?ivovu;) Charmides and Plato-an interpretation independent of hierarchies of age.38 Moors may still be right that the Glaucon of the Republic and the Symposium are the same, but he probably goes too far when he gives 420 or later as Glaucon's date of birth. Glaucon and Apollodorus were "children" (cai&cov OvTwcv giov ni, 173a) in 416, but they need not have been toddlers. To give Moors his due, since the framing conversation occurs in about 400,39 making Glaucon forty-two according to those who favor 421, sixteen years earlier he would have been twenty-six, and that is not a child, so some other calculation is indeed necessary if coherence is to be achieved. If one were to substitute the dramatic date of 411 and postulate an unrecorded Megarian battle the previous summer, Glaucon could have been fourteen. Or one might use Diodorus Siculus's alreadyanachronistic battle date of 409 to make Glaucon twelve or even ten (not eleven, because then he would be Plato's twin), assuring that he would fit the dramatic date of the Symposium. But the consequence of such a maneuver may not please: the upshot is either that the Symposium's Glaucon is not Plato's brother, or that the dramatic dates of the Symposium and the Republic are utterly incompatible. III. THE REMAINING UNCERTAINTY I have mentioned only the most formidable problems of assigning a dramatic date to the Republic, and I have streamlined-though I hope I have not oversimplified-even those. The dialogue is neither internally consistent, nor consistent with other dialogues thought to have been written in the same period,40 nor consistent with the historical record (such as it is). It does not help to fall back on Moors' 38 b?o)pa6tr E??)VOl) S (ov c)( 5iXX Te? Xapi5ri6v rov rxaukiccovoq Kicai ta ax&rova,uovoq E7cauo?v. E. C. Marchant, Xenophon Memorabilia (London 1923), translates the passage, "for the sake of Plato and... Charmides." 39W. R. M. Lamb, Plato Symposium (London 1925) Although there is controversy over the order of composition of the dialogues, Gerard R. Ledger, Re-Counting Plato: A Computer Analysis of Plato's Style (Oxford 1989) esp , uses a wide variety of contemporary statistical techniques and far surpasses his predecessors in establishing an order for segments in which one might have confidence.

12 THE DRAMATIC DATE OF PLATO'S REPUBLIC 393 claim, "Only the philosopher, it seems, can truly understand time, and thereby make a proper judgment on the propriety of the anachronism," because there are just too many anachronisms.41 I earlier raised the issue of whether the Republic was originally conceived as an aesthetic unity, or whether it may have had more than one origin, thus accounting for its lack of a univocal dramatic date. There are several arguments that raise doubt that the Republic was written all at once. The least controversial is that Book I stood as a separate dialogue, an idea with us since 1839 that seems to have gained some acceptance.42 While there are still disputes over whether Book I as it currently exists is the original Book I, or whether it was rewritten into its present form; and while there are disputes about the dating of Book I as a separate dialogue, stylometric evidence argues for its similarity to other Socratic or elenctic dialogues.43 But there is a good deal more diversity to be accounted for in the Republic. Else has argued that a considerable portion of Book X, though not the myth of Er, was an addition to the dialogue.44 And, apart from the so-called early style of Book I, there is cramped late style in parts of the dialogue as well, by which I mean passages that resemble the distinctive prose of such dialogues as the Sophist, Timaeus and Laws.45 Thesleff has identified several 41 Moors, "Argument Against," 24 n Cf. Hermann, Geschichte. Thesleff, Studies, esp. n. 19, details and provides references for seven succinct arguments that Book I was originally separate. Vlastos, Socrates: Ironist, 46-47, also takes Book I to have been separate, and is followed by Richard Kraut, ed., who separates the composition of Book I from that of the remainder of the dialogue in the prefatory material of the Cambridge Companion to Plato (Cambridge 1992) without comment as if the issue were no longer controversial. Charles Kahn mounts a defense of the opposing side in "Proleptic Composition in the Republic, or Why Book I Was Never a Separate Dialogue," CQ ns 43 (1993) Leonard Brandwood, The Chronology of Plato's Dialogues (Cambridge 1990) 251, citing Hermann Siebeck, Untersuchungen zur Philosophie der Griechen (Freiburg 1888), Constantin Ritter, Untersuchungen uber Platon (Stuttgart 1888), and H. von Arnim, De Platonis Dialogis Quaestiones Chronologicae (Rostock 1896). Kahn, "Proleptic," counters that, in his view, stylometry demonstrates only that Book I was written earlier than the other books, but "cannot possibly show how much earlier it was written" (134, his emphasis); Kahn thinks "scholars like Wilamowitz and Friedlander were simply taken in" (133) by von Arnim, who himself was "selfdeceived" (134). 44 G. F. Else, "The Structure and Date of Book 10 of Plato's Republic," Abhandlungen der Heidelberger Akademie der Wissenschaften, Philosophisch-historische Klasse 3 (1972). 45 None of this material is touched by Kahn, "Proleptic"-nor can it be. His proposal of prolepsis is a hypothesis about composition, but it fails as a hypothesis about style.

13 394 late-style passages in Books II-V and VII, very likely indicating late revision of the Republic.46 Thesleff has suggested, however, that when Plato became old enough to require the modern prosthetics now taken for granted (eyeglasses, hearing aid), an assistant (avaypca(pe6;), possibly Philip of Opus, is likely to have helped in the writing of the dialogues we now call "late." If so, alas, we see the secretary's hand in those late-style passages of the Republic.47 Finally and most controversially, efforts to date not only the composition of the Republic but its dramatic date as well will be altogether feckless if there was in fact a proto-republic; not Book I, and not a sketch of the whole, but a subsection of the dialogue that was known in Athens before 392, and which corresponds roughly to the late-style passages in Books II-V and VII that I just mentioned. And if the result is what it appears to be, i.e. the very late editing of a very early segment, it is a perfect example of what cons a computer engaged in stylometric studies into a label of "middle": the two extremes cancel one another out.48 While this is not the place to rehearse the numerous arguments for the existence of a proto- Republic,49 I1 would be remiss not to name a few of the more persuasive among them: (a) there are provocative parallels between Republic V and Aristophanes' Ecclesiazusae in ideology, action, and languageyet the play was produced in 392, long before the date normally assigned for the writing of the dialogue;50 (b) Aulus Gellius writes that a "two-roll" Republic came to light initially; 1 (c) Timaeus explicitly 46 Thesleff, Studies, 137. If he is right about the original boundaries of the proto- Republic, beginning at 369b, then the biographically problematic passages about Plato's brothers, like the notorious, appended end of Book I (from 353e), would seem to have been written to fasten the two previously independent works into one. 47 Thesleff, Studies, Ledger, Re-Counting, There are many, beginning with Hermann, Geschichte. Cf. J. Hirmer, "Entstehung und Komposition der Platonischen Politeia," Jahrbiicher fur Classische Philologie, Supplementband 23 (1897) , esp ; Thesleff, Studies, , and his "Platonic Chronology," Phronesis 34.1 (1989) 1-26; and Nails, Agora, Kahn, "Proleptic," 131, having introduced what he calls the "separatist proposals" of several early scholars, dismisses their proposals as having "largely collapsed under their own weight," but he does not address the contemporary arguments of Thesleff, Studies, which he notes only as a "revival" of the earlier enterprise (131 n. 5). 50 Gilbert Murray, Aristophanes (Oxford 1933) Thesleff's discussion, Studies, , includes extensive citations from the literature. Remarking on linguistic parallels are two of the play's translators: Douglass Parker, [Aristophanes's] The Congresswomen (Ann Arbor 1969) 90 n. 43; and B. B. Rogers, Aristophanes Ecclesiazusae (London 1924) introduction. 51 Aulus Gellius, Noctes Atticae,

14 THE DRAMATIC DATE OF PLATO'S REPUBLIC 395 purports to summarize the whole of the preceding day's discussion irpi not'roeia;, but in fact summarizes only parts of the argument of Republic II, III, and V.52 If the Republic sprang whole from the head of Plato, no univocal dramatic date would thereby be entailed, but the expectation would gain rather in credibility. If not, if Plato wrote a proto-republic before the Academy was established, then its existence militates against the possibility of determining a single dramatic date for the dialogue. And if the Republic was stitched together from a separate Book I, a proto-republic, and new material, and revised late into an almost seamless whole,53 then a single firm dramatic date seems ever more implausible.54 Mary Washington College 52 Plato, Timaeus 17b-19b. F. M. Cornford, Plato's Cosmology (Indianapolis 1975) 4-5, argues that Socrates may have described his ideal state more than once, and that the partial summary that twice calls itself complete is thus not a summary of the Republic but of another conversation. Clay, "Gaps," , argues to the same end, noting that the sets of characters in the two dialogues are different. Were it not for the presence and density of other arguments, I would find their explanation adequate. 53 Diogenes Laertius (3.37) mentions the claim of earlier commentators, Euphorion and Penaetius, that the beginning of the Republic was found after Plato's death in a variety of versions. 54 This paper develops ideas originally presented at the Hunter College conference on "The Uses of the Republic" in It is a pleasure to recall the long and spirited conversations with Holger Thesleff in the days before the conference that heightened my awareness of the complexity of the issues I was raising. I am also grateful to David Ambuel, William Levitan, Jerry Press, and The Classical Journal's anonymous referee for their comments and helpful insights.

15 396 SIGNIFICANT EVENTS FOR A 421 DATE Cephalus settles in Athens Thurii founded b. Lysias b. Glaucon Peloponn. War, Archidamean War I Lysias to Thurii religion of Bendis? Aristophanes' Dait1les tcephalus b. Plato Megara battle (Thucydides) Lysias(23) returns I Peace of Nicias Agathon's Spartans in Thrace GURE b. Lysias SIGNIFICANT EVENTS FOR A 411 DATE Lysias to Thurii Thurii founded Peloponnesian War b. Glaucon b. Plato Aristophanes' Daiteles victory, Melos expedition Phaedrus exiled (returns 404) Alcibiades joins Spartans Sicilian defeat Lysias (46) returns the 400, then the 5, "democracy," festivals renewed Megara battle (Diodorus) pancratium victory of Polydamus