¹ Pickard-Cambridge, A., The Dramatic Festivals of Athens (2nd edn rev. by Gould, J. and Lewis, D.M., Oxford 1988) 95–9Google Scholar; Pope, M., ‘Athenian festival judges – seven, five, or however many’, CQ 36 (1986) 322–6CrossRefGoogle Scholar; Csapo, E. and Slater, W.J., The Context of Ancient Drama (Ann Arbor 1995) 157–65CrossRefGoogle Scholar; Wilson, P., The Athenian Institution of the Khoregia. The Chorus, the City and the Stage (Cambridge 2000) 98–102 and 346–7 nn.222–37.Google Scholar These works have been used throughout this paper.

² Pope (n.1) 324 and 325.

³ Rightly Csapo and Slater (n.1) 159; Wilson (n.1) 98. For noise in the theatre generally, see Wallace, Robert W., ‘Poet, public, and “theatrocracy”: audience performance in Classical Athens’, in Edmunds, Lowell and Wallace, Robert W. (eds), Poet, Public, and Performance in Ancient Greece (Baltimore and London 1997) 97–111, 157–63.Google Scholar

⁴ Zen. 3.64 cites the proverb, making reference specifically to comic choruses. This restriction likely occurs because he is glossing a line of the comic poet Epicharmus. Hesychius, s.v. πέντε κριταί, evidently has no other source. The scholiast to Ar. Av. 445 draws the same conclusion. P.Oxy. 1611.34–7 indicates Lysippus in Bacchae, and Cratinus in Ploutoi (fr. 177 PCG) also said there were five judges in some context.

⁵ Lucian, Harm. 2 (cited more fully in Ila below)

⁶ See OCD ³s.v. ‘Sortition’.

⁷ ‘The use of the Lot carried with it the implication that all citizens were competent to hold these offices and that no special qualifications or experience were required…’ (Sinclair, R.K., Democracy and Participation in Athens (Cambridge 1988) 195).CrossRefGoogle Scholar This attitude further suggests that what was required of any individual selected for an office by lot in terms of literacy was nothing exceptional. Any citizen should be able to fulfil the requirements of the position.

⁸ Hesychius, s.v. διὰ πάντων κριτής, cited below, suggests that the ability to take notes during a performance was seen as desirable.

⁹ Corruption and intimidation of the judges is alleged at [And.] 4.21; Dem. 21.5, 17, 65; Quint. 10.1.72; Ael. 2.8; Aul. Gell. 17.4. In the context of comedy, bribes are offered at Ar. Eccl. 1140–3, Av. 1102–17, and Nub. 1115–20; threats are offered in comedy at Av. 1102–07, Nub. 1121–30, and Pherecr. Krapataloi fr. 102 PCG.

¹⁰ In the New Testament, the successor to Judas is also selected by lot, after a shortlist has been produced (Acts 1: 21–6). It is in this context that we may best understand Proverbs 18: 18, evoked by Pope (n.1) 323: the Athenian judging process does not ‘arbitrarily disen franchise five tribes’. Rather, sortition places the ultimate decision out of the hands of (corruptible) humans.

¹¹ A complementary discussion of this issue is found in Jedrkiewicz, S., ‘Giudizio “giusto” ed alea nei concorsi drammatici del V secolo ad Atene’, QUCC 54 (1996) 85–101.Google Scholar

¹² Lys. 4.4.

¹³ Isoc. 17.33–4. To tamper with the jars at this point was a capital offence.

¹⁴ Plut. Cim. 8.7–9; Dem. 21.65, 39.10.

¹⁵ Rightly Wilson (n.1) 347 n.231; see also Arrighetti, G., ‘Il papiro di Ossirinco n. 1611 e il numero dei giudici negli agoni’, Dioniso 45 (1971–1974) 302–8.Google Scholar The use of different judges for each contest is also suggested by Plut. Cim. 8.7–9, with the unusual substitution of the ten stratêgoi for the judges of tragedy in 468, a measure which strongly suggests no special preparation or training of the judges occurred.

¹⁶ Ar. Eccl. 1159–62; Pherecr. Krapataloi fr. 102; [And.] 4.21 ; Pl. Leg. 659a; Dem. 21.17,65 and Hypothesis II; Plut. Cim. 8.7–9. Aeschin. Against Ctesiphon 232 indicates that judges making bad decisions could end up on trial.

¹⁷ Pope (n.1) 323.

¹⁸ Pl. Leg. 659a; Plut. Cim. 8.7–9; Ael. VH 2.13 (cited in the next paragraph); Luc. Harm. 2.

¹⁹ Those who believe rankings were used include Haigh, A.E., The Attic Theatre (3rd edn, Oxford 1907) 34Google Scholar; Pickard-Cambridge (n.1) 97, and Jedrkiewicz (n.11).

²⁰ Vitr. 7 pr. 4–7.

²¹ Trans. Csapo and Slater (n.1) 163.

²² Rightly Wilson (n.1) 347 n.220.

²³ Special seats are also attested for the judges in Alexandria, in Vitr. 7 pr. 5. The judges therefore sat with the Priest of Dionysus, who was also in the front row (Ar. Ran. 297): for other indications of seating arrangements in the theatre, see Csapo and Slater (n.1) 298–301.

²⁴ This is not how we would wish to understand Ar. Ran. 1114, as some might, for which see Sommerstein, A.H., Frogs (Warminster 1996) 256.Google Scholar However, a passage in Hsch. s.v. διὰ πάντων κριτής, cited below, does suggest that judges could make notes during the performance of plays. Perhaps then the whole antistrophe, Ran. 1109–18, refers not to the audience generally, but to the judges in particular (as representatives of the audience?): ἐστρατευμένοι γάρ εἰσι (1113 ‘they’re old campaigners’, trans. Sommerstein), familiar with the theatre (so Sommerstein 255–6 and Erbse, H., ‘Dionysos' Schiedsspruch in den Fröschen des Aristophanes’, in Δώρημα Hans Diller zum 70. Geburtstag (Athens 1975) 45–60Google Scholar, at 55).

²⁵ Trans. Waterfield, R., Plato: Republic (Oxford 1993) 325–6.Google Scholar

²⁶ See Adam, J., The Republic of Plato (2nd edn, Cambridge 1963) 373–6Google Scholar, who draws together relevant sources, including inscriptions, and rightly insists that όδιὰ πάντων κριτής must be a technical term. At 340–1 he assumes judges at dramatic festivals ranked competitors.

²⁷ Adam (n.26) 375–6.

²⁸ Carey, C., Trials front Classical Athens (London and New York 1997) 17.CrossRefGoogle Scholar

²⁹ Hansen, M.H., The Athenian Assembly in the Age of Demosthenes (Oxford 1987) 41–2, and 41–4Google Scholar generally.

³⁰ Hansen (n.29) 44–6 applies Plato's procedures to the selection of officials in Athens.

³¹ Wilson (n. 1) 99.

³² Hypothesis II to Nub., and see Nub. 575–6, 610–11, Vesp. 1036–47.

³³ For Aristophanes using a separate didaskalos, see D.M. MacDowell, Aristophanes and Athens. An Introduction to the Plays (Oxford 1995) 34 –6. For the unusual situation of Vesp., see 34 n.12. At Eq. 516, Aristophanes has the chorus say that producing comedies is χαλεπώτατον ἔργον ἁπάντων (‘the hardest task of all’). To this we might compare the prominence given to film directors in title credits and at the Oscars.

³⁴ IG I³ 969 (= SEG 23 (1968) 102), about which see Wilson (n.1) 130–6. We likewise infer that the herald's announcement preceding the play was directed at the didaskalos and not the poet. The form of the announcement is given by Ar. Ach. II: εἴσαγ΄, ὦ Θέογνι, τὸν χορόν (‘Bring on your chorus, Theognis’).

³⁵ Csapo and Slater (n.1) 157.

³⁶ Csapo and Slater (n.1) 158, 163.

³⁷ Pope (n.1). Csapo and Slater (n.1) 158–9 argue persuasively against his interpretation. In a tightly run race – exactly the circumstance when the judges are needed most! – where the ten votes are split between competitors 4–3–3 or when two competitors are favoured against a third such as when the votes are split between competitors 4–4–2, Pope's system would fail to produce a clear victor, or provide any mechanism for determining second place. These cases represent 37.3% of the ways that votes might occur in a random distribution, far too much for a viable system. The discrepancy increases further when there were five competitors, as in the comic competition, about which see section V. In effect, Pope's system reckons all the votes: this removes the democratic, competitive and religious benefits offered by counting only a portion of the ballots.

³⁸ Csapo and Slater (n.1) 159: ‘as many more as necessary to break a tie … with a clear winner emerging by the time the eighth ballot is chosen’.

³⁹ If after five votes had been drawn, the split was 2–2–1 for the three competitors (as it would have to be if more votes needed to be consulted), drawing a single ballot could produce a 2–2–2 result. At least by drawing two following the initial five, an ‘upset’ result (i.e. going from 2–2–1 to 2–2–3) is clear and decisive.

⁴⁰ Pope (n.1) is the only scholar to reckon with this passage seriously. His solution is that the variable number represents the variation of numbers of tribes over time: ‘in Lucian's day there were thirteen tribes, and if there was a judge for each tribe, then seven judges, not five, will have been needed for an unbeatable vote’ (326). Although a clever interpretation of Lucian, Pope's solution requires that one competitor will have always obtained at least five of the ten votes cast, which will not produce a victor when the votes are divided 4–3–3 or 4–4–2.

⁴¹ I.e. if at five votes, they were divided 2–2–1, and with two more votes the result were 3–3–1 (i.e. one vote for each of the frontrunners had been drawn).

⁴² Trans. Csapo and Slater (n.1) 162.

⁴³ Dunbar, N., Aristophanes. Birds (Oxford 1995) 307.Google Scholar

⁴⁴ Trans. Csapo and Slater (n.1) 163.

⁴⁵ With 5–5–0, it is always determined by the initial five ballots; with 5–4–1, it is determined in five ballots 76.2% of the time (16/21); with 4–4–2, it is determined in five ballots 52.4% of the time (11/21).

⁴⁶ Because the votes were removed individually from the urn in the public view, the order of the selection could be automatically preserved.

⁴⁷ Of a random distribution, it represents an outcome that occurs 3/3¹⁰ times, i.e. 0.005% of the time.

⁴⁸ Eur. Med., Hypothesis of Aristophanes the Grammarian.

⁴⁹ Soph. OT, Hypothesis II, citing the authority of Dicaearchus. Strictly speaking, it is not known that Oedipus was presented at the Dionysia rather than the Lenaia, though to our knowledge this has not been doubted.

⁵⁰ Ar. Nub., Hypothesis II. In this case, at least, recent scholarship has suggested that the best play did indeed win the prize. Cratinus' Wine-flask (Pytinê), containing the playwright's self-mockery and wholehearted appropriation of earlier Aristophanic criticism, seems to have been a tour de force for its playwright: see Heath, M., ‘Aristophanes and his rivals’, G&R 37 (1990) 143–58Google Scholar; Sidwell, K., ‘Poetic rivalry and the caricature of comic poets: Cratinus' Pytine and Aristophanes' Wasps’, in Griffiths, A. (ed.), Stage Directions. Essays in Ancient Drama in Honour of E. W. Handley (BICS Suppl. 66, London 1995) 56–80Google Scholar; Luppe, W., ‘The rivalry between Aristophanes and Kratinos’, in Harvey, D. and Wilkins, J. (eds), The Rivals of Aristophanes. Studies in Athenian Old Comedy (London 2000) 15–21Google Scholar; and R. Rosen, ‘Cratinus' Pytine and the construction of the comic self’, in Harvey and Wilkins (this note) 23–39.

⁵¹ Ael. VH 2.8. The Suda s.v. Νικόμαχος, indicates further that at one time Euripides was defeated by some one called Nicomachus, but no more can be said than this.

⁵² Av., Hypothesis I. Since there are no extant fragments of Ameipsias' Revellers, and there are of a Revellers of Phrynichus, it is likely that Phrynichus has entered two plays in the Dionysia of 414, as Aristophanes had done at the Lenaia of 422 (see n.33, above, and Sommerstein, A.H., Birds (Warminster 1987)Google Scholar 1 n.1).

⁵³ Haigh (n.19) 35; Pickard-Cambridge (n.1) 99.

⁵⁴ Eur. Alc., Hypothesis.

⁵⁵ It is often argued that for some of the years during the Peloponnesian War, the number of comedies was reduced to three. For a recent and judicious survey of the evidence and the arguments on both sides, see Storey, I., ‘Cutting comedies’, in Barsby, J. (ed.), Greek and Roman Drama. Translation and Performance (Drama 12, Stuttgart 2002) 146–67.CrossRefGoogle Scholar For the years (if any) where the number of comedies was reduced to three, regardless of whether they were presented in a single day or spread out over three days, the voting procedure would be as described in sections I–IV.

⁵⁶ An overview of relevant dates is found in Pickard-Cambridge (n.1) 124–5. The prize for actors at the City Dionysia was introduced c. 449.

⁵⁷ In the same way that one name is used as shorthand in the adjudication procedure for the play itself (not the name of the chorêgos, but that of the didaskalos, as argued in section 1), so in the acting competition the name of the lead actor is used as shorthand for the team of three performers, which may at times have included the play-wright. However, ‘[o]nly the protagonists could form contracts with the archon, receive payment from the state, or win the actor's prize’ (Csapo and Slater (n.1) 223).

⁵⁸ Ar. Plut. Hypothesis, I.G.Urb.Rom. 216.2–6 (= IG XIV 1097), I.G.Urb.Rom. 218.6–13 (= IG XIV 1098), P.Oxy. 2737, all of which may be found in Storey (n.55) 146–8, 150–1.

⁵⁹ The Hypotheses to Ar. Ach., Eq., Nub., Vesp., Pax, Av., and Ran. list only three competitors (see Storey (n.55) 148–9). Because we cannot say what a likely distribution of votes might be, we cannot determine if this is an unreasonably high proportion of Aristophanes' eleven plays. If one were to pursue this argument, it would suggest that the comic competition even with five competitors tended to present two or three that were significantly stronger than the rest.

⁶⁰ Trans. Todd, S.C., Lysias (The Oratory of Classical Greece 2, Austin 2000) 54–5.Google Scholar

⁶¹ The effect of this would of course be heightened if tribes tended to sit together, either by regulation (a possibility doubted by Pickard-Cambridge (n.1) 270 and Wiles, David, Tragedy in Athens. Performance Space and Theatrical Meaning (Cambridge 1997) 37CrossRefGoogle Scholar) or by custom, which is more likely.

⁶² Wilson (n.1) 34 and 347 n.234.

⁶³ Wilson (n.1) 100–1.

⁶⁴ Plut. Mor. 628A–629B, Quaest. Conv. I. 10. See Clement, Paul A. and Hoffleit, Herbert B., Plutarch's Moralia VIII (Cambridge, MA 1969) 94–103Google Scholar; citation from 95.

⁶⁵ To convert a fraction to a percentage, simply convert it to a decimal and multiply by 100.