Kenneth Dover, Aristophanes Frogs. Oxford: Clarendon Press, 1993. Pp. xvi + 398. $69.00. ISBN 0-19-814773-2.
Reviewed by Richard Hamilton, Bryn Mawr College.
Once again D. shows himself master of comedic commentary -- original, incisive, detailed, sensible, witty. Though avoiding the extremes of conjectural emendation and literary criticism, D. discusses almost everything in-between: social values, historical setting, conditions of performance, dramatic conventions, colloquialisms and, perhaps most importantly, gives attention not only to the words of the play but to their physical presentation in the manuscripts, which he finds the best guide to their history. A summary of D.'s substantial introduction will show what I mean.
D. begins with a discussion of Aristophanes' literary career taken from his 1968 edition of Clouds (now with secondary sources) and quickly turns to Composition and Structure of the Play, which is mostly about the absence of Sophocles in the play. Structural anomalies are dismissed ('no worse than Lysistrata') rather than explored: 799-801 are said to give false expectation of technically elaborate measurement as well as weighing [D. does not note that this expectation is held until last so the effect may be one of surprise at fulfilled expectation rather than disappointment because all the instruments listed in 799ff are not employed]; E.'s intention to scrutinize TA\ NEU=RA TH=S TRAGWIDI/AS [862 not 872 as repeatedly cited] is undercut because we don't know what that means [not for lack of theories, which might have been mentioned]. 1251-60 "repeats itself clumsily and is weighted in favour of Aeschylus". [We hear in the commentary that its description of Aeschylus as "the Bacchic king" strikes "a slightly false note" for those who know Euripides' Bacchae, suggesting that the passage is the conflation of two versions -- but this seems unlikely to me.]
III The Contest of Aeschylus and Euripides, has an uncertain start: "We might have expected Dionysos to contemplate a journey to the underworld for a good poet" [but other Aristophanic journeys are for specific individuals like Peace and Hoopoe]. Then a long discussion of sophia (with the startling claim that sophos is never the derogatory "clever" in Old Comedy), whose components are dexiotes ("on its way out from Attic") and nouthesia (which includes making people better, the subject of the first part of the debate 905-1098). A. wins in both categories. Victory in nouthesia we could predict from knowing that the underworld criminals champion Euripides, suggesting that Euripides disturbed a lot of Athenians [acute observations here on the real effect of fictional characters, on quoting out of context, on breaking taboos -- "one careless phrase can wreck the career of a politician"]. Also, A.'s lines are heavier and his criticism of prologues and lyrics is funnier. Still, at 1411 Dionysos can't decide but follows his psyche in an arbitrary intuitive judgment. The chorus explains that E.'s mistake was sitting by Socrates, talking, theorizing (D.'s translation of E)PI\ SEMNOI=SIN LO/GOISIN!) quibbling -- comparison with Clouds shows this is basically new versus old, though the fact of Aeschylean revivals means both styles were "put to the test in the contemporary theatre". D. then with brilliant precision limits the influence of sophists here to orthoepeia: "all else is derived from the comic tradition, anecdotes about poets and ordinary discourse" (D.'s new suggestion for the infamous book of 1114 is "for reading before the plays began or between plays"). "It is unwise to assume when we find words used in the appraisal of poetry both by Aristophanes and by the literary critics of the Hellenistic period, that they already constituted a technical terminology in 405 ... if we claim to see 'sophistic' influence whenever any fifth-century Greek expresses a critical opinion about anything we are rendering the category 'sophistic' useless".
Four sections on characterization follow. Dionysos: it is coincidental that Bacchae may have been produced in same year; Dionysus is actually quite different (bearded, satyrlike), "a collection of functions shaped by comedy itself" (though D. does think Dionysus resilient) therefore we should not speak of development of his character. D.'s lack of interest in themes of disguise and identity is here made explicit: if we "speak of Dionysos as 'searching for himself' or 're-establishing his identity' we must realize that we are translating an ancient comedy into modern categories, perhaps in the faith that a classic author must somehow always be profound even when it looks as if he is frivolous". If one wonders how else to explain the frog chorus, Xanthias gives an answer, Arginusai: "Eupolis' Taxiarchs forbids us to think that Arginusai was the sole inspiration of the rowing-scene, but the implications of a contrast between a foolish master who cannot row and a bold slave who could have won his freedom by rowing cannot have escaped Aristophanes or his audience". The chapter gives us much more, beginning with an exemplary assessment of Xanthias' asides, which mark him as a new sort of comic slave, a type exemplified also by Karion in Wealth and perhaps explained by Arginusai (another careful textual discussion shows Karion is "the only speaker from the household" throughout 823-958, misattributed in most mss). The Doorkeeper of the Underworld should not be called Aiakos, an identification inconsistent in the dramatis personae and manuscripts and already judged implausible in the scholia which is "an unjustified inference of Roman or late Hellenistic date" much like Mnesilochos in Thesmophoriasuzae and Kephisophon in Acharnians [though this latter is not so obvious -- what about Chairephon in Clouds?]. Aiakos in Aristophanes' time was "a hero of exceptional piety ... the decline from tribunal to janitor's lodge is explicable in terms of the overlapping connotations of KLHIDOU=XOS, PULWRO/S, and QURWRO/S". The strongest evidence for Aiakos is the Peirithoos attributed to Euripides or Critias, which resembles Frogs in its Underworld setting, its chorus of Eleusinian initiates and its scene where Aiakos sees Heracles approach. D. suggests that it was never produced or was produced after the Frogs [but this seems to me even more desperate than his argument that "initiates are a fairly obvious category to choose for an underworld chorus"]. The Choruses begins by arguing for the visibility of the frog chorus and -- with a footnote disparaging the search for relevance and unity -- quickly shifts to the principal chorus [though the visibility of the frogs makes the question of their relevance even more pressing]. The discussion attacks the issue of the chorus's role (initiates) versus its function (comic performance): paizein, for instance, fits both role (375, 388, 411, 415, 452) and function (392) and is also blended (333, 407b): "we should not expect our chorus necessarily to enact the procession to Eleusis or any part of the actual Eleusinian ritual, for they are, after all, the souls of people who had been initiated while alive... PAI/ZOUSIN in 319 is an assurance that we are going to witness the festivities of the initiated, not any kind of parody of the very serious business of initiation". Two genuine ritual elements are their invocation of Iakchos and their ragged clothing ("turned to good comic use"). D. then turns to question of a split chorus, which goes back to Aristarchus, with considerable fall-out in the manuscripts and scholia, and which is probably the result of confusion caused by chorus-leader's command and choral response at 370, 383, 394 (416 being an invitation to Dionysos and Xanthias). The apparent departure of the female part of chorus at 445 is probably a sexy joke, which also marks "their last utterance as initiates".
VIII Politics begins with a general assessment of Aristophanic politics, specifically his peace plays, whose seriousness is still (and always) moot, and his criticism of demagogues, whose political style "may have been a novelty after the Periclean years" although "the choice of targets was up to the individual poet". As for the Frogs, not one of the "peace plays", D. argues forcefully for a revival in 404, tied to the honorific olive wreath the Vita says Aristophanes was awarded because of his advice in the parabasis to restore citizen-rights to the 411 revolutionaries, as was done in 404 by the decree of Patrokleides. If so, then the advice about the navy in 1463-65 belong to the first production (since Athens had no navy in 404); hence the considerable problems of the whole scene 1435-66 are solved (1463-5 belong to Aeschylus as well as 1443f, 1446-8, which are placed after 1462).
IX History of the Text continues the discussion begun in D.'s edition of Clouds; here, however, scholia are discussed separately, and colometry becomes the best basis for grouping manuscripts rather than listing of dramatis personae. Having learned a lesson from fully collating 41 mss of Clouds (plus 8 selectively), D. here fully collates only 12 mss (plus 19 selectively). Once again R and V, which cannot by their date have been "improved" by the notorious Triclinius, stand apart: "there are barely a dozen passages in which just one or two pre-Triklinian manuscripts other than R and V give us the right answer, a plausible answer, or a possible pointer to it." The possibility, raised with Clouds, that there was only one ancient ms is strengthened: the fact that two of the three major errors common to all mss "were already in the text in Hellenistic times, as the scholia show" suggests that there was only one Alexandrian copy of Frogs and it was the source of all our mss. Turning to the 84 other manuscripts, which he lists following Eberline (up from Stanford's 74), we must, as with Clouds, accept the impossibility of establishing a stemma based on shared textual errors, given "the universality of correction, the propensity to introduce error, the common use of two exemplars to make a copy and a different two or more as sources of correction, the frequent need to replace lost or damaged sheets from whatever source was available, and the readiness of scholars of all periods to make minor conjectural emendations." D. then looks to scholia, dislocation of verses in stichic passages, colometry of lyric passages and sigla of speakers (but not dramatis personae). Unfortunately "stemmatic investigation of scholia has to be pursued independently of investigation of the text" while study of the sigla shows only that R has blocks of text where change of speaker is omitted or a dicolon is used to mark intralinear change and that half a dozen times in the lesser mss "a missing siglum was imported from a source which designated a character differently". Identical dislocation of lines does help define one group of seven (late) manuscripts but tells us nothing about R and V. D. turns finally to colometry and compares ten manuscripts, which "fall" into two classes, in terms of their differences from R in "overrun" or "split" cola. He finds that one class "has a strong tendency to combine two cola into one". Then D. drops his bomb: "if classification by colometry (and, where, applicable, dislocation) were treated as fundamental evidence for genealogical relationships, and shared textual error relegated to the category of confirmatory evidence, the following steps in the direction of a stemma could be taken." He goes on to describe three sets of affinities: the texts of V and E "though rarely allied in the first two-thirds of the play, show increasing affinity in the last third" while their colometric affinity is consistent throughout; A and Y "so close in colometry" show a "significant degree of affinity through Frogs"; and the same is true of U, Vsi and F. [I have misgivings, particularly given the first affinity, where we have to choose either colometry or text, since I imagine that the probability of independently produced errors in colometry is much greater than in text. (In Clouds D. had concluded colometry "gives meagre results because most copyists had little compunction in 'over-running' colon-end".) Also I am not sure of the origin of the two classes: was this simply the "best fit" of the data for each manuscript separately? And why are pairings of a single ms in one class with a single ms in the other class not counted? The data given in the metrical apparatus criticus marking overruns and splits provide no help since only three mss other than R are recorded (and they show many more cases that mix the two classes than do not). We are left, then, without the basic data to understand the claimed colometric affinities.]
As befits a tradition that has spawned an industry, D. discusses the scholia at length, noting that Frogs scholia "mention Aristarchos more often than the scholia on all other plays put together, and refer ten times to Timachidas". They also lack metrical analyses of lyric passages, and are "particularly fond of the words XARAKTHRI/ZEIN, XARAKTHRISTIKO/S". Charting relationships, D. notes first that textual corrections are rare, observance of word-order is not scrupulous, abbreviation by omission is frequent, and differentiation between accidental and deliberate omission is often difficult. [The table showing how the parts of a scholium can shift should have numbered the parts consecutively.] He then lists places where the R scholia "offer better value", where V differs from R and the others, and finally where "the others" can be divided into E vs M/Vb3/Y. D. ends warning us not to be too deferential to the opinions of Hellenistic scholars gleaned from the scholia. He then discusses the "indirect tradition" noting first the difficulty in determining when an echo counts as a testimonium and then that "out of a thousand testimonia for Frogs only one offers us a significant good reading not already to be found in the extant medieval manuscripts" and so omits many in his text (most single words, vague reminiscences), marking many others as less than a "conscientious quotation".
The final section, Production, argues briefly and cogently for four actors and two doors.
The text generally resembles Stanford's, though we lose Diagoras in 320 and gain Mary at 1345, the most egregious of a surprisingly large number of typos (mostly confined to cross-references). Speaker attributions differ at 78, 181ff, 189, 414, 483, 565-75, 644f, 944, 947, 1018, 1177, 1205, 1209, 1323, 1415, 1415, 1418, 1424. The apparatus criticus reports only R, V, A, K (Milan, Ambrosiana C 222 inf.) fully. As with Clouds, the commentary is perceptive, undogmatic, wide-ranging, amusing, unpredictable and a pleasure to read. Particularly welcome are its exquisitely detailed discussion of meter, frequent reference to secondary literature, and generous translation of colloquialisms, particles, obscure vocabulary, lyric.
This is extraordinary monument to its author and to the craft of positivist scholarship. For once, we should have no complaint about the price!