Clouds, Eupolis and Reperformance

5 Clouds, Eupolis and Reperformance Hallie Rebecca Marshall The extant text of Aristophanes’ Clouds is generally held to be an unperformed revision of the play that had been staged in competition at the Dionysia in 424/3 BC.1 The reasons for the opinion that this version of the play was not staged are threefold: (1) An ancient hypothesis to the play (Hyp. I Dover = Hyp. 6 Wilson; hereafter Hyp. I) claims: τοῦτο ταὐτόν ἐστι τῷ προτέρῳ. διεσκεύασται δὲ ἐπὶ μέρους ὡς ἂν δὴ ἀναδιδάξαι μὲν αὐτὸ τοῦ ποιητοῦ προθυμηθέντος, οὐκέτι δὲ τοῦτο δι’ ἥν ποτε αἰτίαν ποιήσαντος. ‘This [the extant text] is the same as the previous one, but it has been revised in details, as it would be if the poet wanted to produce it again but for some reason or other did not after all do so.’2 (2) A scholion on Clouds 553: Ἐρατοσθένης δέ φησι Καλλίμαχον ἐγκαλεῖν ταῖς διδασκαλίαις, ὅτιφέρουσιν ὕστερον τρίτῳ ἔτει τὸν Μαρικᾶν τῶν Νεφελῶν, σαφῶς ἐνταῦθα εἰρημένου, ὅτι πρῶτος καθεῖται. λανθάνει δὲ αὐτόν, φησίν, ὅτι ἐν μὲν ταῖς διδαχθείσαις οὐδὲν τοιοῦτον εἴρηκεν· ἐν δὲ ταῖς ὕστερον διασκευασθείσαις εἰ λέγεται, οὐδὲν ἄτοπον· αἱ διδασκαλίαι δὲ δηλονότι τὰς διδαχθείσας φέρουσιν. ‘Eratosthenes says that Callimachus found fault with the Didaskaliai because they show Marikas being produced two years later than Clouds, though here it clearly stated that it (Marikas) was produced first. Eratosthenes says that Callimachus was ignorant of the fact that no such thing was said in the version that was produced, and there is nothing strange about it being said in the later revision. Obviously the Didaskaliai list the plays that were produced.’3 (3) the text as it has been transmitted to us seems not to have been revised in full, nor is it complete.4 In this paper I argue that in light of the evidence of plays with apparently similar textual traditions, and considering the scholarship done in recent years on the reperformance of plays in Attica and beyond, we ought to reconsider what the extant Clouds might represent in terms of ancient performance. I contend that it is more reasonable to assume that this text is fundamentally similar to other dramatic texts from the fifth century, and therefore almost certainly a text rooted in performance, than to argue that it remains uniquely unperformed .5 The first piece of evidence brought to bear on the argument for or against performance are the words of Hypothesis I, cited in the previous paragraph. 63 No Laughing Matter layout.indd 63 24/03/2012 16:02 Hallie Rebecca Marshall Hypotheses II Dover (= Hyp. 5 Wilson; hereafter Hyp. II) similarly reports that the revision was not staged: αἱ πρῶται Νεφέλαι ἐδιδάχθησαν ἐν ἄστει ἐπὶ ἄρχοντος Ἰσάρχου, ὅτε Κρατῖνος μὲν ἐνίκα Πυτίνῃ, Ἀμειψίας δὲ Κόννῳ. διόπερ Ἀριστοφάνης ἀπορριφεὶς παραλόγως ᾠήθη δεῖν ἀναδιδάξας τὰς Νεφέλας τὰς δευτέρας καταμέμφεσθαι τὸ θέατρον. ἀτυχῶν δὲ πολὺ μᾶλλον καὶ ἐν τοῖς ἔπειτα οὐκέτι τὴν διασκευὴν εἰσήγαγεν. αἱ δὲ δεύτεραι Νεφέλαι ἐπὶ Ἀμεινίου ἄρχοντος. The first Clouds was produced at the City Dionysia during the archonship of Isarchis (424/3 BC), when Cratinus won with Wineflask, Ameipsias <came second> with Konnos. So Aristophanes, tossed out unreasonably, thought it necessary to produce the second Clouds and chastise the audience. But his luck was even worse with this and he did not thereafter produce the revised version. The second Clouds dates to the archonship of Ameinias (i.e. 423/2 BC).6 In each hypothesis, the author starts from what he could presumably establish as fact, either through his own recourse to the texts and didascalic records, or by reference to information provided by commentaries. Hypothesis I tells us that the second Clouds was the same as the first, but revised in details, and continues on to give specifics about the nature of the revision, saying that the parabasis, the contest between Right and Wrong, and the burning of Socrates’ school belong in their entirety to the second version.7 This has the appearance of information derived from a direct comparison of the two texts in question. Similarly, the statement in Hypotheses II as to the year that the play was first produced, the festival at which it was entered in competition, and how it placed represent the sort of detail which would have its origin in the didascalic records and which could have, and at some point would have, been checked against those records. In both hypotheses, however, the information turns from the evidence provided by the text itself or by the historical records to an attempt to explain why two differing scripts existed in the textual tradition but only the first had a corresponding entry in the Didascaliae. The author of Hypothesis I is more restrained in his speculation than the author of Hyposthesis II: his evidence demonstrates clearly that Aristophanes revised the text, and he assumes that this must have been with the intent of reperformance, but since there is no didascalic record of a second production, he deduces that for some reason the revised text must not in the end have been performed.8 The author of Hypothesis II seems less restrained, but he does not provided any further evidence to support his claims. He too is clearly aware that there were two divergent texts of Clouds, and he seems to make the assumption that the second version of Clouds was Aristophanes’ response to the first Clouds having placed third in competition. One suspects that the author has relied on the parabasis of the revised Clouds in his reasoning here (520-5): οὕτω νικήσαιμί τ’ ἐγὼ καὶ νομιζοίμην σοφός, ὡς ὑμᾶς ἡγούμενος εἶναι θεατὰς δεξιοὺς καὶ ταύτην σοφώτατ’ ἔχειν τῶν ἐμῶν κωμῳδιῶν, †πρώτους ἠξίωσ’ ἀναγεῦσ’ ὑμᾶς†, ἣ παρέσχε μοι 64 No Laughing Matter layout.indd 64 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance ἔργον πλεῖστον· εἶτ’ ἀνεχώρουν ὑπ’ ἀνδρῶν φορτικῶν ἡττηθεὶς οὐκ ἄξιος ὤν· So I may win the prize and be thought sage, I took you for intelligent theatregoers and this for the most sophisticated of my comedies; that is why I though you deserved to be the first to savor it, a play that cost me very hard work. Then I lost the contest, defeated by vulgar men, though I didn’t deserve it.9 The author of the hypothesis dates the revision to the year following its first performance, to the archonship of Ameinias (423/2 BC), but claims that it was even more of a failure and therefore was not produced (by which he presumably means that Aristophanes was not granted a chorus for the play at the Dionysia in 422).10 The date given by the second hypothesis for the revised text is demonstrably wrong. This is seen on the basis of exactly the evidence that had caused confusion for Callimachus, as attested by the scholion to Clouds 553. The revised Clouds makes specific mention of Eupolis’ Maricas, which was produced two years after the first Clouds. The extant play must therefore postdate the production of Maricas.11 More relevant for my argument is the fact that Hypothesis II provides no details for the supposed greater failure of the second Clouds, though the reasoning is likely the same as that of the author of the first hypothesis. The evidence that survive in a variety of sources indicates two relatively secure facts: (1) two texts of Clouds were in circulation in antiquity, with one clearly being a revised version of the earlier play; and (2) there were didascalic records indicating that Clouds had been produced at the Dionysia in 424/3, where it placed third, but there was no corresponding entry for a revised version of the play in subsequent years. There is no indication of evidence, however, to support anything in the hypotheses beyond these two facts. The remainder of the information provided by them seems to be a series of deductions to explain these details: that a revised text surely indicated that Aristophanes wanted to remount his play; that the new parabasis suggested that this was in part to defend his reputation as a playwright; that since there was no didascalic record of a second production, the revision must have been even less successful than the first; and finally that the only thing less successful than coming in third place, is not being granted a chorus in the first place. In both cases the logic of the progression from evidence to speculation is comprehensible, and similar, if not identical, to the steps of logic followed by modern scholars when faced with the same evidence. Indeed, Rosen describes the scholion to Clouds 553 as ‘the most compelling evidence to suggest that the revised Clouds was not, in fact, performed’.12 I will argue, however, that the assumption that our texts were only performed at the Lenaia and City Dionysia is incorrect, and that recent scholarship on performance outside these venues means that this assumption about the relationship between text and competitive performance ought be re-evaluated. We may even evaluate to what extent modern scholars have been led to conclusions about the non-performance of the second Clouds by the hypotheses and the scholion, 65 No Laughing Matter layout.indd 65 24/03/2012 16:02 Hallie Rebecca Marshall by comparing the standard view of the performance history of Clouds to other Athenian comedies. Other plays of Old Comedy apparently circulated in two separate versions (Aristophanes’ Aeolosicon, Thesmophoriazusae, Peace, and Wealth, Diocles’ Thyestes, Magnes’ Dionysus, and Eupolis’ Autolycus).13 Also relevant is Aristophanes’ Frogs, which survives as a single text, but preserves variants which represent revision for reperformance. The number of titles known to us suggest that it may in fact have been if not common, at least not unusual for ancient plays to be revised in some fashion.14 We must be cautious, however, to separate our knowledge of ‘double’ texts of tragedies from ‘double’ texts of Old Comedy, as the available evidence suggests that these may represent two separate phenomena. When Butrica wants to argue for the two attested versions of Thesmophoriazusae being two distinct plays under a single title, the ancient evidence that he brings to bear for this possibility derives exclusively from tragic examples.15 The examples from Old Comedy are fundamentally different from the tragic examples, however. It seems that all the double plays in Old Comedy bear some narrative relationship to one another. We can see this difference in the practice of the Hellenistic librarians: where they could give adjectival titles to distinguish two similarly named tragedies, comedies are only ever distinguished by number.16 The only two examples where the evidence extends meaningfully beyond very limited evidence that the same title was used for two separate Old Comedies is with the text of the revised Clouds and its first hypothesis and testimonia surrounding Eupolis’ Autolycus. While no text of Autolycus has survived there is sound evidence, as laid out by Storey, that two versions of this play were in circulation in antiquity.17 In addition to the five citations which seem to distinguish between two separate Autolycus plays, there is Galen’s very helpful definition of epidieskeuasthai (ἐπιδιεσκευάσθαι): λέγεται βιβλίον ἐπὶ τῷ προτέρῳ γεγραμμένῳ τὸ δεύτερον γραφέν, ὅταν τὴν ὑπόθεσιν ἔχον τὴν αὐτὴν καὶ τὰς πλείστας τῶν ῥήσεων τὰς αὐτὰς τινὰ μὲν ἀφῃρημένα τῶν ἐκ τοῦ προτέρου συγγράμματος ἔχῃ, τινὰ δὲ προσκείμενα, τινὰ δ’ ὑπηλλαγμένα· παράδειγμα δ’ εἰ βούλει τούτου σαφηνείας ἕνεκα, τὸν δεύτερον Αὐτόλυκον Εὐπόλιδος ἔχεις ἐκ τοῦ προτέρου διεσκευασμένον. This term is used of a work that is rewritten from the original version. It has the same plot and most of the same text, but will have some things removed from the original, some things added, and some revised. If for the sake of clarity you need an example, you have the second Autolykos of Eupolis ‘revised’ from its previous version.18 As Storey notes, Galen’s definition also fits well with the Clouds revisions described by Hypothesis I. Within three years of Clouds, then, Autolycus was apparently revised in a similar manner, even though, like Clouds, we have only a single date for performance attested.19 But Storey’s immediate assumption, which would I think be generally accepted, is that Eupolis’ play was performed twice and that the divergent texts represent the two separate performances. Indeed Storey says, ‘Autolycus is the only play by Eupolis that we know was performed a second time’.20 But the reality is 66 No Laughing Matter layout.indd 66 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance that not a single piece of the evidence for a second Autolycus says anything about a second performance. Storey assumes that two texts mean two performances, and I contend that this situation is exactly parallel to the evidence we have for Clouds. The strong correspondence between the testimonia regarding the revised nature of both Clouds and Autolycus as ‘double’ plays should encourage us to question whether other Old Comedies that circulated under a single title were in fact completely new plays. As Butrica noted, it may well have been that tragedies with the same title were able to be distinguished from each other in the fifth century by the fact that they did not stand alone as plays, but were part of a tetralogy – which Ajax was being described could be defined by its companion plays.21 This is not true of Old Comedy, which could potentially mean that what at first glance looks like a parallel habit in both tragedy and Old Comedy of circulating new plays under previously used titles is nothing of the sort. While it may have caused confusion for later scholars who were often reading tragedies in the absence of their accompanying plays, fifth-century Athenians may never have conceived of these as ‘double’ plays. Clouds and Autolycus suggest that Old Comedies circulating under a single title were seen as revisions of previously existing plays. Similarly, there is no strong evidence to assume that Clouds alone of these ‘double’ plays went unperformed in its second incarnation. The lynchpin for the non-performance for the extant Clouds is the apparent lack of a second entry for Clouds in the didascalic records: as Rosen puts it, ‘Eratosthenes’ brief comment, then, certainly holds that a second, revised performance of Clouds never appeared in the didascalic records at Athens, and so was probably never performed’.22 The absence of Clouds from the didascalic records a second time can only be taken as evidence that the revised text was not performed in either of the two central Athenian competitions following its initial performance in 424/3.23 It should not be taken as evidence that the revised text was never performed.24 This has implications for Autolycus as well: a second script does not necessarily mean a second performance in competition at the Lenaia or Dionysia. Indeed, several other possible performance contexts present themselves: (a) the revised script may have been performed at another dramatic festival in Athens. (b) the revised script may have been performed under specific non-competitive conditions in Athens (as was the case with Frogs, perhaps25). (c) the revised script may have been intended for production in a non-Athenian context, which may or may not have been competitive. (d) the revised script, though unperformed (perhaps having failed to secure a chorus), remained in circulation and was seen as a text intended only for reading purposes. Clouds has generally been thought to fall into this final category, uniquely. The other three bear consideration, however, though I believe we may also exclude the first possibility. Even though we know that major tragedians did not think it beneath them to compete at the lesser festivals,26 there is no comparable evidence for comedy. Further, plays at the Lenaia and Dionysia apparently shared dramatic conventions: a play written for one festival was dramaturgically coherent with a play written for the other,27 and I presume that the same would be true of other Athenian dramatic competitions. Since the extant text of Clouds does not cohere with established 67 No Laughing Matter layout.indd 67 24/03/2012 16:02 Hallie Rebecca Marshall performance conventions, this seems considerably less probable.28 Nevertheless, two likely performance circumstances remain: either non-competitive performance within Athens or performance in another community. Clouds and Autolycus are not our only evidence for the practice of revision of Old Comedy. There were two versions of Aristophanes’ Peace, Thesmophoriazusae, and Wealth, and Frogs seems to preserve small-scale revisions alongside the original performance text of 405 BC. While we do not know whether plays listed under the same title are in fact completely distinct plays rather than one being a revision, where two versions are accepted, the assumption is always that these alternate texts represent a double performance tradition. There is some evidence to support this. The Hypothesis to Frogs, for example, cites Aristotle’s student Dicaearchus as the source for the fact that ‘the play was so admired because of its parabasis that it was actually produced a second time’.29 No one doubts the play was reperformed, but when and where are not certain, and it is not clear if this honorary re-staging was within the competitive context or not. As in the case of all the other double texts of Aristophanes, we have only one secure date for production, on the basis of which, coupled with dateable internal references in the text or fragments, scholars guesstimate when it might have been staged. There are two points to be made here. First, the evidence leans towards certain plays being revised by Aristophanes and Eupolis, making it therefore likely that other poets of Old Comedy did so as well. Not every play was revised it would seem, but enough were that it was not seen to be unusual or peculiar. Secondly, in every instance (except for Plutus), there is only one date that has come down to us from antiquity for performance in competition at the Dionysia or Lenaia. This pattern seems to suggest that Clouds fits into a larger pattern, and that it did not uniquely remain unperformed. Whatever happened is best seen as part of a larger practice among fifth-century poets that resulted in plays circulating in two scripts with a single production date. If Clouds is not unique there are a number of consequences. First and foremost we need to be much more circumspect about taking the parabasis (esp. Clouds 520-5) to heart. As Major has argued, it shares features and tone with a number of other Aristophanic parabases.30 As a general rule, again, the parabasis of Clouds has been treated as a unique passage, with the author directly addressing his failure in competition, and this has been read as a reflection of Aristophanes’ genuine feelings. Not only does the parabasis of Clouds share features with the parabases of other plays, but there is even evidence that the parabasis could be somewhat formulaic. Indeed, while Aristophanes claims not to bring things out two or three times,31 Peace 751-60 repeats Wasps 1029-37 almost verbatim. Indeed, it is on the evidence of this repetition that Storey dismisses the relevance of Eupolis fr. 89 K-A to Clouds, suggesting that the same accusation must also have been present in an earlier play.32 And as much as I would like to take that fragment as evidence that Eupolis expected the revised Clouds parabasis to be widely known, Storey is right to allow the possibility that Aristophanes made the same charges in another play. Parabases by their nature are both formulaic and topical, referring to events in the very recent past, and so are likely targets for largescale revision in reperformance. Within these formulaic conventions of chastising the audience, Aristophanes got good mileage out of the result of the dramatic competition of 423, and thus returned to it in at least two parabases.33 68 No Laughing Matter layout.indd 68 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance We must also reconsider the assumption that the first Clouds failed as miserably as Aristophanes later describes. Taking the parabasis as autobiographical fact, many believe Aristophanes revised the play in order to redeem his reputation following what he felt was an unjust third-place finish. If this were the case, the extant text of Clouds might represent one of two things: either, following Hypothesis II, Aristophanes revised it with a view to reperformance but was not given the opportunity to perform his play (and to this modern scholars add the suggestion that Aristophanes then, still believing in his play when no one else apparently did, put the text into circulation to redeem himself with a reading audience34), or the revision was never intended for reperformance, and was simply a literary revision for an exclusively reading audience.35 I find neither of these plausible. If the play was such a failure, in either scenario, we must ask who do we imagine would be interested in reading the play, and who do we think would be willing to invest the time and money on producing copies of such a text? In every single competition someone had to come in last place (just ask Euripides). Surely many poets thought that they had deserved to place better. The sort of revision and distribution that we are envisioning for Aristophanes, however, did not apparently lead to a flood of revised texts seeking redemption for their authors. Furthermore other examples from comedy provides a counter-example. Peace placed second at the Dionysia in 421, and even Frogs, which placed first, was partially revised for reperformance. This suggests that a play may have been revised regardless of how it had placed, and so, whatever is behind these doublets, it is unlikely to be poetic redemption before the eyes of the public. We may even question ‘losing’ might really mean. Because of the nature of the voting process, a third-place finisher might still have received support from several judges (and not even necessarily the fewest votes),36 and placement in competition need not correspond to a play’s popularity with the audience. We even have a (late) anecdote from Aelian, VH 2.13, that suggests Clouds was in fact a crowd favourite: ἐκρότουν τὸν ποιητὴν ὡς οὔ ποτε ἄλλοτε, καὶ ἐβόων νικᾶν, καὶ προσέταττον τοῖς κριταῖς ἄνωθεν Ἀριστοφάνην ἀλλὰ μὴ ἄλλον γράφειν They (the audience) applauded the poet as never before and shouted that he should win and commanded the judges from above to write no other name but Aristophanes.37 So while Clouds placed third when it was first performed in 424/3, there is evidence to support the idea that despite Aristophanes’ apparently indignant response, his play was in fact very well received in this year. And despite ‘losing’ at the Dionysia 424/3, Aristophanes was back at the next major dramatic festival, the Lenaea of 422 with not one but two plays: Wasps, which he directed and which placed second, and Proagon, directed by Philonides, which placed first. It seems a remarkable coup for any poet to have two plays in the same competition38 and this suggests that despite appearances Aristophanes had somehow won substantial favour with those who granted the honour of producing plays at the festivals. Two choruses in the same festival would require substantial expenditure of Athenian resources, financial and otherwise, and 69 No Laughing Matter layout.indd 69 24/03/2012 16:02 Hallie Rebecca Marshall can only be read as a clear marker of faith, by everyone involved, in Aristophanes’ talent as a poet of Old Comedy. Considered in this light, it becomes increasingly improbable that Clouds was the failure that it is so often made out to be. If the revised text is not a response to a failed production, we need to reconsider what it might represent. In light of what Storey has called ‘the curious matter of the Lenaia festival of 422’, we should give serious consideration to the possibility that, despite placing third at the Dionysia, this was among Aristophanes’ most warmly received plays. This fits with the anecdote in Aelian (which, though from the second century AD, may have some historical basis), and it raises the possibility that Clouds was revised because it was still being performed several years after its debut at the Dionysia. This possibility also finds some potential support in Eupolis fr. 89 K-A, which if it does refer to the revised parabasis of Clouds would mean that the play was still expected to be familiar to audiences when Eupolis’ Baptai was performed c. 417-415 BC.39 This theory of comic reperformance explains why a limited number of scripts (perhaps those featuring notorious caricatures of figures such as Euripides and Socrates, who continued to be prominent in Athenian life) also circulated in doublets, but large numbers of plays apparently did not.40 References to Clouds by Plato and Xenophon, both followers of Socrates, provide two final pieces of evidence which ought to be considered. At Apology 19a-c, Plato has Socrates say: Ἀναλάβωμεν οὖν ἐξ ἀρχῆς τίς ἡ κατηγορία ἐστὶν ἐξ ἧς ἡ ἐμὴ διαβολὴ γέγονεν, ᾗ δὴ καὶ πιστεύων Μέλητός με ἐγράψατο τὴν γραφὴν ταύτην. εἶεν· τί δὴ λέγοντες διέβαλλον οἱ διαβάλλοντες; ὥσπερ οὖν κατηγόρων τὴν ἀντωμοσίαν δεῖ ἀναγνῶναι αὐτῶν ‘Σωκράτης ἀδικεῖ καὶ περιεργάζεται ζητῶν τά τε ὑπὸ γῆς καὶ οὐράνια καὶ τὸν ἥττω λόγον κρείττω ποιῶν καὶ ἄλλους ταὐτὰ ταῦτα διδάσκων’. τοιαύτη τίς ἐστιν· ταῦτα γὰρ ἑωρᾶτε καὶ αὐτοὶ ἐν τῇ Ἀριστοφάνους κωμῳδίᾳ, Σωκράτη τινὰ ἐκεῖ περιφερόμενον, φάσκοντά τε ἀεροβατεῖν καὶ ἄλλην πολλὴν φλυαρίαν φλυαροῦντα, ὧν ἐγὼ οὐδὲν οὔτε μέγα οὔτε μικρὸν πέρι ἐπαΐω. Let us then take up the case from its beginning. What is the accusation from which arose the slander in which Meletus trusted when he wrote out the charge against me? What did they say when they slandered me? I must, as if they were my actual prosecutors, read the affidavit they would have sworn. It goes something like this: Socrates is guilty of wrongdoing in that he busies himslef studying things in the sky and below the earth; and he makes the worse into the stronger argument, and he teaches these same things to others. You have seen this yourself in the comedy of Aristophanes, a Socrates swinging about there, saying he was walking on air and talking a lot of nonsense about things which I know nothing at all.41 Plato’s Apology takes place in 399 BC, and was written sometime thereafter. Socrates was a frequent butt of comedic jests in the plays of Old Comedy between 423 and 399. But despite numerous jokes about Socrates in multiple plays, Plato assumes not only that the jurors have seen Clouds in performance but that they remember the play, 70 No Laughing Matter layout.indd 70 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance and that their opinion of Socrates may have been swayed by his depiction therein.42 That these Athenians would remember a third-place play that had been staged nearly a quarter of a century earlier, in a city where there had been more than 150 new comedies at the Dionysia and Lenaia in the intervening period should be surprising. Socrates’ followers, Plato and Xenophon, who both were small children in 423, both foreground Clouds in representing the public perception of Socrates by those who did not know him personally. One can envisage a scenario whereby Socrates’ bookish followers, looking for reasons to explain the jury’s disposition against him, found Aristophanes’ Clouds to be the most egregious and perhaps sustained attack on him in Old Comedy. But that does not account for Plato’s suggestion that the jurors had actually seen Aristophanes’ Socrates, and that this depiction would have has such a memorable impact on them that it might still sway their judgment of the man more than two decades after the fact. For Plato to think that this play was of such potential influence to that it would be addressed directly in Socrates’ defence speech, there would have to have been a reperformance of the play at some point in the not-sodistant past. While it would be possible to come up with an Earl of Essex/Richard II scenario for reproduction on the eve of the trial, the internal references within our extant script are absolutely clear that our revised text dates from much earlier.43 It is also possible that Xenophon tells us something about the afterlife of Clouds.44 Xenophon, writing well after the death of Socrates in the fourth century, chose to set his Symposium in the year 421, with himself as a guest, even though the historical reality was that he would have been but a child at the time. We must assume that Xenophon has a reason for the date and occasion at which he has set his work, as well as for the details that he chooses to include. What the dialogue presents cannot represent any historical event, but rather is a deliberately crafted literary fiction. Within this fiction, Aristophanes’ Clouds enters the discussion, despite having been performed decades before Xenophon wrote his work. At Symposium 6.6-10 the Syracusan who has been hired to provide entertainment for the occasion approaches Socrates and asks if he is the one they call the ‘thinker’ (Xen. Symp. 6.7, φθονῶν τῷ Σωκράτει εἶπεν· ἆρα σύ, ὦ Σώκρατες, ὁ φροντιστὴς ἐπικαλούμενος;). He goes on to say that Socrates is supposed to be ‘a thinker on celestial subjects’ and asks him to measure the distance between them in flea feet (Xen. Symp. 6.7-8). There can be no doubt that this is a reference to Clouds 144-54: the question should be, how does the Syracusan know the play? Given that this entire event is a construction of Xenophon’s imagination the question is not looking for a historical certainty. Rather, how does Xenophon think the play was being disseminated among the public. Again, there are a number of possibilities: the reader is meant to assume that the Syracusan saw the play at the Dionysia in 423, and the play is still being discussed more than a year on; or the Syracusan saw the play performed somewhere other than at the Dionysia; or the Syracusan read the play. Whatever Xenophon had in mind, there is no doubt that he viewed the play as damaging to the image of Socrates. When the Syracusan brings up the play with Socrates he does so spitefully (φθονῶν). Xenophon gives us a character who does not know Socrates to see him, and who is not Athenian, who, on account of Aristophanes’ play, thinks of the philosopher as someone who studies natural phenomena and discusses them in the most ridiculous of terms. Somehow, this is also supposed to remain coherent for Xenophon’s audience, for whom the 71 No Laughing Matter layout.indd 71 24/03/2012 16:02 Hallie Rebecca Marshall allusion must also be meaningful. Again, the possibility that Clouds possessed an afterlife in performance following its debut seems the most likely of the three options when we need to account for it apparently being familiar, or least being perceived to be familiar to Syracusan procurers, Athenian jurors and Socrates’ followers. The evidence of both Plato and Xenophon suggests a shared belief that Aristophanes’ Clouds was widely known across a period of more than twenty years and that the play shaped many people’s ideas about Socrates to his detriment. The modern understanding of the history of Clouds, that it was a dismal third-place failure and that a revised or partially revised text circulated as a means of redeeming the play, simply cannot be reconciled with this perspective. On the other hand, a very popular play, that continued to be reperformed following its premiere, to such an extent that it was worth the author’s time to revise it several years later, would account for the hostile view of the play maintained by two of Socrates’ followers in the early years of the fourth century. Even today, in the age of cheap and accessible means of textual production and reproduction, playscripts tend to be very ephemeral objects. Plays that do not enter the literary canon or have regular revivals on the stage disappear with alarming alacrity. Aristophanes’ Clouds survived in two versions, and it continued to be perceived as culturally relevant and influential. The reason for this, I believe, was that it continued to live on the stage in the years following its initial performance.45 Notes 1. See Dover 1968: lxxxi-xcviii, and also Storey 1993: 73-4 and Rosen 1997: 401, all of which rely on the same evidence for non-performance. Revermann 2006: 326-33, with the caveat that the authors of the hypotheses may have been reliant on Athenian records and thus unaware of non-Athenian performance, also accepts that the play was not performed, but argues that the text of the revised Clouds was conceptualized by Aristophanes as a performance text. 2. Tr. Dover 1968: lxxxii. 3. Tr. Csapo and Slater 1995: 6. 4. For a succinct summary of the most serious internal problems, see Storey 1993: 73-4. On the number of actors required by the extant script, and the potential performance problems therein, see Russo 1994: 92-109; MacDowell 1994: 329-30; MacDowell 1995: 144-9; and Revermann 2006: 224-35 and 326-28. 5. Athenaeus 270a says that he knows of two dramatists, Metagenes and Nicophon, who were active in the fifth century, who wrote plays that were not staged. As Rosen (1997: 425 n. 9) cautions, however, Athenaeus gives no indication as to whether these plays had been written with a view to performance and therefore cannot be used as evidence for a strictly text-based tradition of dramatic poetry. And to Rosen’s caution, I would add that Athenaeus also tells us nothing about why he believes that these plays were not performed. It is possible that the issue here, as with the second Clouds, is the lack of an entry for either play in the didascalic record. It is worth noting that Athenaeus does not include Clouds as a play that was believed to have been unperformed. 6. Tr. Csapo and Slater 1995: 5. 7. See Dover 1968: lxxxiv. 8. For a discussion of the phrasing of the Hypothesis on this point and possible interpretations of the phrasing, see Rosen 1997: 402. 9. Tr. Henderson 1998: 83. 72 No Laughing Matter layout.indd 72 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance 10. The date of 423/2 for the revision is not supported by the internal to later events, and is likely a further logical deduction on the part of the scholiast, who has assumed that revision would have occurred immediately following the play’s poor showing in the Dionysia of 424/3. 11. On the date for the revised Clouds see Kopff 1990 and Storey 1993, in response to Kopff. 12. Rosen 1997: 401. 13. On the known ‘double’ plays of Old Comedy and how they are referred to in ancient secondary sources, see Butrica 2001: 55-61, Konstantakos 2004: 9-19, and Sommerstein 2010: 11-29. 14. Konstantakos 2004: 15 has argued that revision ‘was common in the Greek comic theatre’. Revermann 2006: 330-31, on the evidence of an anecdote in Athenaeus (9.374a-b) and KA test. 5.5 regarding Telecleides’ Sterroi, suggests that there may have been a general practice of revision among the poets of Old Comedy. 15. Butrica 2001: 56-8. 16. Though see Konstakos 2004: 14, who argues that sometimes ‘the double title may be the result of διασκευή, a second production of the play in a revised version. ... the poet could well change the title, e.g. in order to give the impression that he was producing a new play and not just rehashing old material’. 17. Storey 2003b: 82-4. 18. Galen, Commentary on Hippocrates’ On Regimen in Acute Diseases 1.4. See Storey 2003b: 83. 19. Storey provides two pieces of evidence for the date of the first Autolycus. A scholiast to Plato’s Apology at 19c tells us that both Eupolis, in Autolycus, and Plato the comic playwright, in his Nikai, made fun of Aristophanes’ statue of Peace. Given that Peace can be securely dates to the Dionysia of 421 BC, Autolycus must date to 420 or later. That it was in 420 precisely is made clear by the second piece of evidence: Athenaeus in the course of discussing the chronological problems of philosophers in respect to Xenophon’s Symposium (216c-d), says: ἔστιν δἐ οὗτος ὁ καιρὸς καθ’ ὃν Ἀριστίων ἄρχων ἦν. ἐπὶ τούτου γὰρ Εὔπολις τὸν Αὐτόλθκον διδάξας διὰ Δημοστράτου χλευάζει τὴν νίκην τοῦ Αὐτολύκου. ‘This is the time when Aristion was archon (421/0); for in that year Eupolis put on through Demostratos his Autolykos making fun of the victory of Autolykos’ (tr. Storey 2003b: 81). 20. Storey 2003b: 82. 21. Butrica 2001: 56. 22. Rosen 1997: 401. 23. We know little about the afterlife of plays following their initial festival performance. Given the number of anecdotes that we have about Athenian tragedians traveling abroad because of non-Athenian demand for their plays and the evidence from vase paintings for an audience in South Italy for Old Comedy, some plays could enjoy new performances on the stage following their performance at the Dionysia or Lenaia. If we think that these plays were being reperformed outside the boundaries of Attica, we must consider the possibility that there were opportunities to restage these plays within Attica as well. We shouldn’t think in terms of going from Broadway to off-Broadway: rather, having made the offering to Dionysus in the appropriate festival context, the spoils of the ritual offerings could be shared more widely among the community. 24. One further piece evidence is potentially relevant to reperformance. The anecdote preserved by Herodotus (6.21) about Phyrnichus’ Fall of Miletus identifies the penalty for reminding the Athenians of their own woes was a substantial fine and ‘they passed a regulation that no one was to make use of this drama in the future’ (ἐπέταξαν μηκέτι μηδένα χρᾶσθαι τούτῳ τῷ δράματι, tr. Csapo and Slater 1995: 11). The phrasing of this leaves open the possibility that reperformance might not just be by the poet in competition, but that others might also seek to stage a play following its performance in competition. Such anecdotes must 73 No Laughing Matter layout.indd 73 24/03/2012 16:02 Hallie Rebecca Marshall always be used with caution, but Herodotus is of particular value for being a contemporary source suggesting that reperformance was a possibility for a play with a disastrous reception. 25. Frogs apparently preserves line variants alongside one another: see, for example, Sommerstein 1996 on lines 1251-60, 1431a-b, 1437-53. 26. See Pickard-Cambridge 1988: 40-2. Aelian, VH 2.13 describes Socrates going to the Piraeus whenever Euripides was competing there. 27. The only exception to this I can find is Russo’s claim that the theatrical crane was available only at the Dionysia (Russo 1994: 3). 28. The extant Clouds is apparently missing a chunk of choral text and has two scenes which require four actors. Again, caution must be urged when reading anything into this when making claims for performance or non-performance. Even the plays of Shakespeare, written with the intent to create a script that would be reperformable and with the wonders of the printing press readily available, frequently circulated in printed editions that are not stage-worthy (the so-called ‘bad quartos’). Even in this relatively recent tradition, for which we have multiple editions that can be compared against each other and good knowledge of the performance context, there is no clear explanation for the origin of these substandard texts. It is only the extant ‘good quartos’ and folio scripts, coupled with external evidence, that assures us that these texts, whatever their origin, have their roots in a performance tradition. A cautionary tale for classical scholars trying to explain the first stages of the textual history of Clouds is the theory put forward by Pollard and Wilson as to why the first quarto of Romeo and Juliet was ‘bad’ and the second quarto ‘good’. Their proposed solution to this textual conundrum was that the first quarto represented an early draft of the play set from Shakespeare’s manuscript, and the second a final stage-ready revision of the play, also set from Shakespeare’s manuscript. On the narratives created for the varying quality of the printed texts of Shakespeare, see Werstine 1990. 29. See Storey 2003b: 83. 30. See Major 2006: 138-43 for the somewhat habitual practice of using the parabasis to chastise the audience and mock other writers of Old Comedy. 31. Clouds 546. Sommerstein 1991: 189, in his commentary on this line, writes that ‘this assertion is audacious even by the standards of the present passage, since it is certain that substantial parts of the revised Clouds are taken over virtually unaltered from the original version’; he then points the duplicated passage in Peace and Wasps. 32. Storey 1990: 22. 33. In addition to complaining about his loss in the revised parabasis of Clouds, at Wasps 1043-5 Aristophanes seems to abuse the audience for their treatment of Clouds in 424/3. 34. See Rosen 1997: 405-11. 35. See Fowler 1989: 257-8. 36. See Marshall and van Willenberg 2004: 95-101. 37. Translated by Csapo and Slater 1995: 163. 38. See Storey 2003a: 281-92. 39. On the date of Eupolis’ Baptai, see Storey 1993 and Kopff 1990. 40. Though perhaps a coincidence, it is worth noting that two of the plays most plausibly identified in vase-painting are plays with known multiple scripts: the Wurzberg Telephus (see Csapo 1986 and Taplin 1987) which almost certainly depicts a scene from Thesmophoriazusae, and the Getty birds, which may depict the agon from the first Clouds (see Taplin 1987, 93-6, contra Green 1985). 41. Tr. Grube 1997: 17-18. 42. Revermann 2006: 329 uses Plato’s emphasis on the jury having seen the play, with no mention of the possibility that they could read the play, to argue that there was a general perception of performance being memorable and reading audiences being small. He 74 No Laughing Matter layout.indd 74 24/03/2012 16:02 5. Clouds, Eupolis and Reperformance acknowledges, however, that there are real questions about ‘how many jurors in 399 could reasonably be assumed to have seen it’, if the first and only performance had taken place twenty-four years earlier. 43. On 7 February 1601 the Lord Chamberlain’s Men, having been commissioned by supporters of the Earl of Essex, put on a performance of Shakespeare’s Richard II, apparently with the intent of fomenting ill-will towards Queen Elizabeth. On this performance, see Dawson and Yachnin 2011: 2-9. 44. If I am correct in suggesting that double texts are indicative of popularity and reperformance, one also wonders about Xenophon’s choice of occasion for the setting of his Symposium: the athletic victory of the young Autolycus in 421. While we know little about the actual content of Eupolis’ Autolycus, it too seems to have been tied to this same victory based on its date, and it too circulated in two texts. This again raises the possibility that Xenophon was using a general familiarity, both for himself and his audience, derived from reperformances of old plays to select the settings of his Socratic dialogues. On the possible content of Autolycus, see Storey 2003b: 84-92. 45. This work has supported by Government of Canada through the Social Sciences and Humanities Research Council. I would like to thank the editors for their helpful suggestions and comments, as well as Tyson Sukava, without whose assistance this paper may well have resembled the incomplete text of Clouds. Thanks are also owed to Ian Storey for his encouragement of my interest in this play over the years, even though my heresies could not have been good for his angina. Bibliography Cooper, J.M. (ed.) (1997), Plato: Complete Works (Indianapolis/Cambridge: Hackett Publishing Company). Csapo, E. (1986), ‘A Note on the Würzburg Bell-Crater H5697 (‘Telephus Travestitus’)’, Phoenix 40.4 , 379-92. Csapo, E. and W.J. Slater (1995), The Context of Ancient Drama (Ann Arbor: University of Michigan Press). Dawson, A.B. and P. Yachnin (eds) (2011), Richard II (Oxford: Oxford University Press). Dover, K.J. ( 1968), Aristophanes: Clouds (Oxford: Clarendon Press). Fowler, D. (1989), ‘Taplin on Cocks’, CQ 39.1, 257-9. Grube, G.M.A. (tr.) (1997), ‘Apology’, in Cooper 1997, 17-36. Henderson, J. (1998), Aristophanes: Clouds, Wasps, Peace. Loeb Classical Library (Cambridge, MA: Harvard University Press). Huss, B. (1999), ‘The Dancing Sokrates and the Laughing Xenophon, or the Other Symposium’, AJPh 120.3, 381-409. Konstantikos, I.M. (2004), ‘Antiphanes’ Agroikos-Plays: An Examination of the Ancient Evidence and Fragments’, Rivista di cultura classica e medioevale 46: 9-40. Kopff, C.E. (1990) ‘The Date of Aristophanes Nubes II’, AJPh 111.3, 318-29. MacDowell, D.M. (1994), ‘The Number of Speaking Actors in Old Comedy’, CQ 44.2 (1994): 325-35. MacDowell, D.M. (1995), Aristophanes and Athens (Oxford: Oxford University Press). Major, W.E. (2006), ‘Alazoneia: Laughing at the Parabasis of the Clouds’, CW 99.2 (2006): 13144. Marshall, C.W. and S. van Willigenburg (2004), ‘Judging Athenian Dramatic Competitions’, JHS 124, 90-107. Phillips, D. and D. Pritchard (eds), (2003), Sport and Festival in the Greek World (London: Duckworth / University Press of Wales). 75 No Laughing Matter layout.indd 75 24/03/2012 16:02 Hallie Rebecca Marshall Revermann, M. (2006), Comic Business: Theatricality, Dramatic Technique, and Performance Contexts of Aristophanic Comedy (Oxford: Oxford University Press). Rosen, R.M. (1997), ‘Performance and Textuality in Aristophanes’ Clouds’, YJC 10.2, 397-421. Russo, C.F. (1994), Aristophanes: An Author for the Stage (London and New York: Routledge) Sommerstein, A.H. (1983), Aristophanes: Clouds (Warminster: Aris and Phillips). Sommerstein, A.H. (2010), The Tangled Ways of Zeus: and other studies in and around Greek tragedy (Oxford: Oxford University Press). Storey, I.C. (1990), ‘Dating and Re-Dating Eupolis’, Phoenix 44.1, 1-30. Storey, I.C. (1993), ‘The Dates of Aristophanes’ Clouds II and Eupolis’ Baptai: A Reply to E. C. Kopff ’, AJPh 114.1, 71-84. Storey, I.C. (2003a), ‘The Curious Matter of the Lenaia Festival of 422’, in Phillips and Pritchard 2003, 281-92. Storey, I.C. (2003b), Eupolis: Poet of Old Comedy (Oxford: Oxford University Press). Taplin, O. (1987), ‘Phallology, Phlyakes, Iconography and Aristophanes’, PCPhS 213, 92-104. Werstine, P. (1990), ‘Narratives about Printed Shakespeare Texts: “Foul Papers” and “Bad Quartos” ’, Shakespeare Quarterly 41.1, 65-86. 76 No Laughing Matter layout.indd 76 24/03/2012 16:02