Scholars have repeatedly lamented the more limited attention that Knights have attracted compared to Aristophanes’ other plays[1]. This new edition by Zachary Biles and Douglas Olson, two eminent specialists in Aristophanes and Greek comedy, contributes greatly to overcoming this neglect. This volume will remain a reference work on Knights for many years. Scholars and students of all levels of expertise will now be ideally served by this commentary, together with the useful 2020 edition by Anderson and Dix.
Biles’ and Olson’s edition of Knights adopts a standard structure, consisting of an introduction, the critical text of the hypotheseis and of the play, an extensive commentary, and the indices.
The introduction is divided into six sections: (1) “Aristophanes and Knights”; (2) “The poetic form and dramatic character of the play”; (3) “Historical background”, with sub-sections (a) “Pylos” and (b) “The evolution of Athenian political leadership”; (4) “The politics of Knights”; (5) “Production practicalities and their implications”, with sub-sections (a) “Division of parts”, (b) “Masks, costuming and props”, (c) “Staging”; (6) “The text”, with subsections (a) “Papyri”, (b) “Manuscripts”, (c) “The Triclinian and Post-Triclinian text”, (d) “Other witnesses”. This introduction is effective in providing a clear picture of the main interpretative problems in the study of Knights. The first four sections cover a range of topics, with useful discussion of some central issues: the formal and thematic consequences of Aristophanes’ design of Knights as a frontal attack on Cleon; how Aristophanes has dramatised the interaction between the Athenian democracy and its leaders; the elusive and self-contradictory presentation of the Sausage-seller and Demos throughout the play. The analysis is lucid, enabling the reader to form their own opinion. The section on the text is likely to have the more far-reaching consequences for the study of Aristophanes. Thanks to Biles’ and Olson’s extensive investigation of the manuscripts, we can form a clear picture of the relationship between the sources. One point of detail which I find especially interesting is Biles’ and Olson’s conclusion that, contrary to the view of previous scholars, the Suda is not part of the same sub-branch as R (Ravennas 429), since it agrees in error both with R against the sub-branch y and with the sub-branch y against R.
Coming to the critical text, it is only natural to compare this edition with Nigel Wilson’s 2007 OCT. As explained in the preface (pp. vii-viii), Biles and Olson are less inclined than Wilson to adopt emendations and supplements (vv. 62, 295, 358, 414, 540, 580, 589, 599, 628, 711, 1035, 1294, 1328, 1399) and defend the received text in places obelised by Wilson (vv. 193, 260). Biles and Olson are also ready to print and obelise a corrupt text when no convincing solution is available (vv. 304, 1230). Nevertheless, in several cases Biles and Olson accept conjectures from earlier scholars (or variant readings that are probably Byzantine emendations) in places where Wilson retains the received text (vv. 396, 536, 546, 819, 969, 1108, 1242, 1311, 1334), delete or obelise the received text against Wilson (vv. 114, 742, 1062), or change the transmitted order of lines (vv. 649-650). Therefore, Biles’ and Olson’s text is not simply a more ‘conservative’ text than Wilson’s. Biles and Olson also challenge the manuscripts’ assignment of speakers’ parts (vv. 54-57, 145-146) and reject the traditional identification of Slave I and Slave II as Demosthenes and Nicias.
Besides accepting other scholars’ conjectures, Biles and Olson adopt four new emendations of their own: (1) v. 79, ὁ νοῦς against ὁ νοῦς δ’ in Rj and ὁ δὲ νοῦς in v a and the Suda; (2) v. 819, Ἀχιλλείαις against Ἀχιλλείων in x and the Suda; (3) v. 1091, ἀρυταίνᾳ against ἀρυταίνῃ in x; (4) v. 1339, τί δ ̓ ἔδρων, κάτειπέ μοι, πρὸ τοῦ; ποῖός τις ἦ; against κάτειπέ μοι, πρὸ τοῦ; καὶ in y and πρὸ τοῦ, κάτειπέ, καὶ in R. These are mostly very light modifications, which normalise the syntax (2) or aim to impose order on divergent manuscript readings (1, 4). But (3) is baseless, and the justification is perplexing: “The MSS offer ἀρυταίνῃ, but the form in eta is attested elsewhere only in Athenaeus (2nd century CE), and we have accordingly regularised the spelling” (p. 564). This seems to imply that ἀρυταίνᾳ is widely documented, but this is not the case: ἀρυταίνᾳ is never attested. On the contrary, the transmitted dative ἀρυταίνῃ is precisely what one expects in Attic for an ă-stem with alpha impurum (e.g. τράπεζα, τραπέζης, etc., where -ης, -ῃ in the genitive and dative singular is the outcome of Ionic-Attic ᾱ > η). Therefore, ἀρυταίνᾳ is simply to be rejected. Biles and Olson suggest four more emendations in the apparatus: (1) v. 216, ἀπογλυκαίνων for ὑπογλυκαίνων; (2) vv. 264-5, transposing ἀμνοκῶν | πλούσιος into πλούσιος | ἀμνοκῶν; (3) v. 407, παιδοπίπην in place of πυροπίπην based on the scholia; (4) v. 523, βατράχειος or βατραχειοῦς as a proleptic predicative for transmitted βατραχείοις (Blass had previously suggested βατραχειοῖς, which is linguistically superior to the manuscript reading). (3) and (4) are tempting, (1) and (2) are in my view unnecessary.
In their critical text, Biles and Olson have sought to adopt the spellings more likely to have been employed in Aristophanes’ time. Commendably, Biles and Olson do not approach this matter with a heavy hand. For instance, they rightly resist (unlike Sommerstein and Wilson) Brunck’s ὑπολίσφοις at v. 1368; considering that the evidence from Atticist lexicography is split, Biles and Olson (like Coulon) are likely to have chosen wisely[2]. Given this attention for orthography and spelling, it is puzzling to read ΑΡΙΣΤΟΦΑΝΟΥΣ ΙΠΠΕΙΣ on the title page (p. 55) and then top of every page of the critical text. In fact, ἱππῆς is the spelling rightly adopted by Biles and Olson at vv. 225 and 242, and so ΙΠΠΕΙΣ is probably just a repeated typo. Also, at v. 822 Biles and Olson print Brunck’s ἐλελήθης in place of transmitted ἐλελήθεις; that -ης (< -εας) is the expected ending of the pluperfect in Aristophanes’ times is certainly right, but the same emendation should have been adopted at v. 1044, where Biles and Olson print ἐλελήθεις instead[3]. Also with regard to spelling, Biles and Olson make a different choice from Wilson with regard to ποιέω/ποέω: while Wilson prints ποι- at all places (except πεπόηται at v. 63, which is perhaps a slip), Biles and Olson print πο- where the first syllable must be short (cf. p. 230, where one misses a reference to Willi 2003, 236-237). In such cases, the spelling adopted in modern editions is largely a matter of convention. Still, I find it inconsistent when editors alternate -ο-/-οι- in ποιέω/ποέω but retain -οι- in other words even where the syllable is short (cf. ἑκεινοιί at v. 1196). Finally, it is surprising that Biles and Olson adopt j’s συνιστάμενον at v. 863 against ξυνιστάμενον in Rva. (They also adopt συν- over ξυν- at v. 427, but here συν- is better supported in the manuscripts.)
One of the greatest merits of this edition is the thoroughness of the critical apparatus. Compared to Wilson’s, Biles’ and Olson’s apparatus is far richer and provides a clearer picture of the manuscript evidence, including the Triclinian manuscripts. An example is at v. 379, where Biles and Olson indicate that the reading σκεψόμεσθ’ is a (typical) Triclinian emendation, whereas Coulon attributed it to the Aldine edition, and Wilson does not note anything in the apparatus. Also, thanks to their thorough investigation of the manuscripts, Biles and Olson show that the reading αὐτή at v. 1172 is already in Γ2 (Laurentianus Pluteus 31.15), while Wilson indicated it as a conjecture by Bothe. This eloquently illustrates the importance of Γ2 as a source of extra-stemmatic readings. One shortcoming of the apparatus lies in the difficulty of tracing the bibliographic references for the conjectures.
The commentary stands out for clarity, conciseness, and sharpness. The general approach of Biles and Olson is pragmatic. In the preface, Biles and Olson present the aim of this commentary as “to open up the text of Aristophanes’ play on a more fundamental level, allowing readers of all sorts to then do with it what they will” (p. viii). This goal has undoubtedly been achieved[4]. Biles and Olson provide abundant information on matters of language (especially, but not only, vocabulary, with a keen eye for the shortcomings of modern lexica) as well as on textual and staging issues, constantly guide the reader in the minute exegesis, and collect a wealth of parallel passages.
In many cases, Biles and Olson illuminate the meaning of difficult passages and improve on interpretations by previous scholars[5]. In other cases, they show the limits of our knowledge. Inevitably, one cannot possibly agree on every point of interpretation. For instance, I remain sceptical about Biles’ and Olson’s interpretation of vv. 411-412[6]. The evidence from the lexicographers collected by Biles and Olson cannot be considered compelling, because it may very well derive, as is often the case with the lexicographical entries which distinguish between the meaning of very similar and nearly identical words, from a rough generalisation based on a limited number of instances. (Indeed, Eup. fr. 300 PCG is likely the locus classicus on which the doctrine about μαχαιρίδες meaning only ‘barber’s shears’ depends.) Further, Pollux might well be wrong when he comments on the passage of Knights, but we have no way of ascertaining it. Besides, the comparison with v. 1236 strongly suggests that already at vv. 411-412 the Sausage-seller refers to his professional apprenticeship. An actual mistake seemingly occurs in the commentary on ἧσθ’ at v. 1405. Against previous scholars, Biles and Olson convincingly show that at vv. 1402-1405 Demos makes a double invitation. Commenting on ἧσθ’ at v. 1405, however, Biles and Olson write (p. 667): “ἧσθ(α): thus Meineke for the paradosis ἦν (probably a deliberate correction when ΗΣΘ was misunderstood as standing for elided ἧστο)”. But ἧσθ(α) with a rough breathing does not exist, and ἧσθ’ undoubtedly stands for ἧστο rather than ἦσθα, which would make no sense in the passage.
Here and there, some characteristically blunt remarks may strike the reader as somewhat unfair. For instance, in their note on v. 754 (p. 454) Biles and Olson describe Sommerstein’s suggestion that Demos takes a seat on a stone in the orchestra as “too extraordinary a proposal to be accepted without positive support from the text itself”[7], and argue that the most economical hypothesis is that Demos sits on the couch on which, according to Biles’ and Olson’s reconstruction, Slave I took a seat at v. 98. However, if one is looking for “positive support from the text itself”, we must admit that there is no evidence in the text to prove that Slave I reclined on a couch at v. 98 (κατακλινήσομαι would also be compatible with lying down on the ground). By contrast, the Sausage-seller’s ἐπὶ ταυτησὶ […] τῆς πέτρας at v. 754 (cf. also vv. 783-787) clearly indicates that Demos sits on a stone (or on a stage prop resembling one). To be sure, one cannot prove conclusively that Demos was sitting in the orchestra rather than on stage, but such a hypothesis would prove intriguing and is worthy of consideration[8]. To mention another case, the comment at p. 451 against Kock’s conjecture ὑποτρεμόντων that “every step further away from the paradosis makes a conjecture less likely to be right” is an oversimplification, at least if so formulated (cf. pp. 548-549). One is reminded of the even blunter warnings by A. E. Housman against such an approach to textual criticism in “The Application of Thought to Textual Criticism”[9].
The volume is well-produced. However, the font size of the critical text and apparatuses is extremely small, which makes the reading unnecessarily tiring. Typos are rare and mostly without consequence. A few are worth flagging. At p. x, the edition of the Synagoge is misattributed to Hansen, which could confuse readers who are less familiar with lexicography. At section VI.B of the introduction, I cannot understand the oscillation between x and χ. In the critical text, I have noticed παλαί (4x) at vv. 1155-1156, while the nota personae Ἀλ. καὶ Πα. at v. 1161 has been split, and καὶ Πα. has become part of the verse. In the commentary, “Slave II” at pp. 608-609 is likely a mistake for “Slave I” (i.e. “Demosthenes”). Indeed, there seems to be a recurring confusion in the identification of Slave I and Slave II. See also p. 134, where as part of a discussion on vv. 54-57, assigned to Slave II (i.e. “Nicias”) in the text and later in the commentary (cf. p. 163), one reads that “Slave I is certainly identified briefly with Demosthenes at 54–7”. Whatever the reason behind these inconsistencies, perhaps reflecting different stages of Biles’ and Olson’s work, they are likely to confuse the reader.
To conclude, Biles’ and Olson’s edition is a lasting achievement that will remain a reference work for the years to come.
Bibliography
Anderson, C. A.; Dix, T. K. (2020). A Commentary on Aristophanes’ Knights. Ann Arbor.
Belardinelli, A. M. (1981). ‘Nota a Men. Sicyon. 421’. Eikasmos 2, 163-166.
Bravi, L. (2020). Cavalieri. I canti. Pisa, Roma.
Bravi, L.; et al. (2023). Il teatro greco. Strutture, società, eredità. Atti del convegno, Chieti, 15-18 marzo 2022. Pisa, Roma.
Diggle, J.; Goodyear, F. R. D. (1972). The Classical Papers of A. E. Housman. Volume III: 1915– 1936. Cambridge.
Lautensach, O. (1896). ‘Grammatische Studien zu den griechischen Tragikern und Komikern I. Personalendungen’. Programm des herzoglichen Gymnasium Ernestinum zu Gotha.
Olson, S. D. (2016). ‘Scenes from an Ill-Spent Youth (Aristophanes, Knights 411-412)’. CQ n.s. 66, 774-775.
Slater, N. W. (2002). Spectator Politics. Metatheatre and Performance in Aristophanes. Philadelphia.
Storey, I. C. (2023). Review of Mike Lippman, Wilfred E. Major, Aristophanes: Men and their horses. Fargo: Theran, 2022. Pp. 97. ISBN 9781944296223. BMCR 2023.08.40.
Willi, A. (2003). The Languages of Aristophanes. Aspects of Linguistic Variation in Classical Attic Greek. Oxford.
Wilson, N. G. (2007). Aristophanis fabulae. Tomus I. Oxford.
Notes
[1] Cf. Slater 2002, 68: “The Knights is today one of the less studied and least produced of Aristophanes’ plays”; Storey 2023: “Knights (424 BCE) is one of Aristophanes’ unloved comedies”.
[2] It may well be that Moeris simply applies to λίσφος vs. λίσπος an Atticist tenet that the forms with φ are “more Attic” than those with π. (By contrast, in the case of φιδάκναισι at v. 724, the Atticist sources are unanimous in their identification of the spelling with φ- as the proper Attic one).
[3] Cf. also Lautensach 1896, 11.
[4] Biles and Olson engage with a wide range of scholarship. There are some recent items that they do not cite, such as Bravi 2020 and the articles on Knights in Bravi et al. 2023.
[5] With regard to vv. 546-547, Naber’s exegesis (followed by Biles and Olson) is strengthened further comparing Men. Sic. 421 (cf. Belardinelli 1981).
[6] Cf. already Olson 2016.
[7] Incidentally, it is not entirely correct to write that, according to Sommerstein 1981, after v. 754 “the action that follows takes place not on stage but in the orchestra” (p. 454). The stage-direction given by Sommerstein 1981, 81 is: “All move away from the house; Demos seats himself on a rock at one side of the orchestra”. Sommerstein’s hypothesis also presupposes that there existed “a natural rock-seat near the edge of the orchestra” (Sommerstein 1981, 182).
[8] Cf. Slater 2002, 77-78.
[9] Cf. Diggle–Goodyear 1972, 1065.