Sophocles was an outstandingly successful playwright in his lifetime, with a most impressive number of victories (pp. 5–6 above); he was even Socrates’ favourite tragedian.Footnote ² Shortly after his death he received praise in Aristophanes’ Frogs of 405, and in Phrynichus’ Muses of the same year, in which a character refers to his ‘many beautiful tragedies’.Footnote ³ In particular, the sweetness of his poetry was remarked on while he was still alive, and would become a frequent image in later criticism, paradoxical though that may seem given the dark tones of so many of his plays.Footnote ⁴ For the rest of antiquity he retained his popularity, though from now on he would always be overshadowed by a playwright who during his lifetime had only rarely beaten him. The greater favour enjoyed by Euripides is visible as early as the fourth century, when his plays seem to have been reperformed more often than those of Sophocles, and then in later times is reflected by the considerably larger numbers of papyri, as well as quotations from his plays in other authors. Aeschylus, on the other hand, was far less popular than either Sophocles or Euripides, as measured by the same criteria. The reason for these relative standings, which were to remain consistent, will lie partly in the relative difficulty of the Greek used by each playwright. Aeschylus’ lexicon is the furthest removed from everyday language; Sophocles’ vocabulary may be easier to grapple with, but his syntax is still difficult, certainly more difficult than that of Euripides. Aeschylean tragedy was also less suited to displaying the bravura abilities of individual actors who played such a part in the transmission of tragic texts; in addition, his fondness for connected trilogies may have proved challenging in dramatic contexts unsuited to that elaborate form.

For all that, Sophocles remained firmly in the repertoire during the fourth century. We hear of a performance of Oenomaus in Collytus, one of the Attic demes located within the city of Athens,Footnote ⁵ and Antigone was performed many times in the same city, involving the famous actor Theodorus,Footnote ⁶ who had Electra in his repertoire too.Footnote ⁷ Epigoni was performed by the actor Andronicus at the end of the fourth century.Footnote ⁸ Famously, the actor Polus of Aegina played the part of Electra, whose lament over the urn he once played while holding the ashes of his recently deceased son, creating an extraordinary interplay between reality and fiction;Footnote ⁹ Polus also took the title role in both of Sophocles’ Oedipus plays.Footnote ¹⁰ Timotheus of Zacynthus, an actor of uncertain date, was so effective in his portrayal of Ajax's suicide speech that he was given the epithet ‘The Slayer’, referring to the first words of the soliloquy in which Ajax's sword is so designated.Footnote ¹¹

Actors overall played a vital role in the transmission of Sophocles’ text.Footnote ¹² By the fourth century, they had largely taken over from directors the control of tragic production; actors’ decisions concerning which plays to revive, which to keep in the repertoire, and which to discard, will have both responded to and shaped popular taste. At the same time, actors were no respecters of the texts whose transmission they unintentionally guaranteed. They changed those texts when it seemed to them dramatically desirable, particularly with a view to lengthening parts in order to make them better showcases for their talents. (Ajax's suicide speech, which made Timotheus famous, shows indications of this.) As a result of such changes, the Athenian statesman Lycurgus (active c. 338–325/4) decreed that actors would henceforth have to use the text on an official state copy; such intervention on behalf of Sophocles, Aeschylus, and Euripides indicates how these playwrights were by now canonical.Footnote ¹³

Our evidence for which plays actors were performing during this period is meagre; nevertheless, it may be significant that of the six plays mentioned above, fully five are from the seven plays which reached us complete. If our evidence is representative (and it may not be, since some of it comes from long after the fourth century, when authors may have been more likely to record anecdotes about the plays best known in their day, regardless of their relative popularity during an earlier period), this would indicate that at this time there was already a concentration on a particular set of plays, whose popularity would in time enable them to win out over the others once the pressures of transmission became acute.

The internationalization of tragedy assisted by the travels of star actors will have ensured the spread of Sophocles’ plays beyond Athens. Vase paintings of tragic scenes from south Italy and Sicily indicate that performances of Oedipus the King, Electra, and Oedipus at Colonus, at least, took place in that region.Footnote ¹⁴ The conquests of Alexander the Great opened up unprecedented vistas for the spread of tragedy. The man himself had copies of Sophocles’ plays sent to him while on campaign (along with the works of Aeschylus, Euripides, and the historian Philistus), and is said to have listened to ‘the children of the Persians, of the people of Susa, and of the people of Gedrosia’ singing Sophocles’ tragedies, together with those of Euripides.Footnote ¹⁵

A key staging-post in the transmission lay in another part of Alexander's empire, in Egypt. The ‘Museum’ (i.e. Temple of the Muses, the guardians and inspirers of artistic enterprise) was established at Alexandria by Ptolemy I (ruler 323–283, king from 305), together with a library; his successor, Ptolemy II Philadelphus (sole ruler 283–246), engaged Alexander of Aetolus to edit tragic and satyric texts there.Footnote ¹⁶ Whether or not we believe the late anecdote that Ptolemy III Euergetes (246–221) permanently acquired the Athenians’ copy of the plays (presumably the Lycurgan copy) by forfeiting a valuable bond,Footnote ¹⁷ the Alexandrian edition or editions of Sophocles’ plays decisively shaped all subsequent texts of the poet's works. Work on Sophocles is attested by a range of scholars over several centuries, including Aristophanes of Byantium (257–180), Aristarchus of Samothrace (c. 216–144), Didymus (first century bc), Aristonicus, Theon (both of Augustan date), Nicanor (second century ad), and Horapollon (fifth century ad).Footnote ¹⁸ Unfortunately, the great majority of the ancient scholarship on Sophocles that has come down to us, in the margins of his manuscripts (there called ‘scholia’, abbreviated as Σ), is unattributed to any source, and considerably compressed to fit around the text of the plays, ripped from its original habitat in separate monographs.

A late third-century epigram, roughly contemporary with the work of Aristophanes of Byzantium on Sophocles, highlights the excellence of two Sophoclean plays in particular:

The two dramas praised by the speaker of Dioscorides’ epigram both survive today. The reference to them here suggests that they were particularly admired and well known, less than two centuries after Sophocles’ death; and thus that the process of canon formation, a process whose culmination, or nadir, saw the reduction of the number of Sophoclean plays available to just seven, was already well under way.

If this epigram stood alone, we might have to reckon with the chance of coincidence; but that chance seems to be eliminated by a subsequent epigram from the early first century ad:

Here four plays are mentioned, of which three survive (Oedipus Tyrannus, Oedipus Colonus, Electra): such a pattern across the two epigrams which name specific plays by Sophocles makes it all but certain that at least four of the seven plays that survived into the medieval period had achieved a special popularity in the ancient world long before the other plays were lost. Both epigrams mention Electra, which seems to have been an especially popular play.

Sophocles’ influence on Latin literature was considerable;Footnote ²¹ that influence will have arisen thanks to the Romans’ growing interest in all aspects of Greek culture, and possibly also to the mediation of a continuous performance tradition in southern Italy. It can be identified even among the meagre fragments of early Roman tragedy (for example, those of Accius’ Antigone), and also in classical poetry; Virgil's engagement with Ajax is particularly evident.Footnote ²² Those who Dine Together was performed at Rome in 54 bc,Footnote ²³ and extracts from Electra were sung (in a translation by Atilius) at the funeral of Julius Caesar in 44.Footnote ²⁴ How much Sophocles was translated into Latin, and how easy it was to acquire a copy of his plays in different parts of the Roman Empire, is beyond our knowledge, however.

The Sophoclean text in the Imperial period seem to have been closer to what Sophocles actually wrote compared to the one which made it into the Middle Ages. In the medieval manuscripts of Sophocles, the Messenger's speech in Antigone which announces the death of Antigone and the suicide of Haemon concludes with the following lines:

Were this all the evidence that we had for this passage, the curiously abbreviated οὐ τίθημ᾽ ἐγώ (‘I do not place/consider him’) would no doubt be defended as an effectively oblique statement, perhaps even as a colloquialism. But we do have more evidence. Writing in the late second or early third century, Athenaeus cites this passage with a further line after that phrase (a line today referred to as line 1167), which allows the full construction to be observed: οὐ τίθημ᾽ ἐγὼ | ζῆν τοῦτον, ἀλλ᾽ ἔμψυχον ἡγοῦμαι νεκρόν (‘I do not consider that this man is alive, but regard him as a living corpse’). Somehow this line slipped out of the main manuscript tradition, but was preserved in the secondary tradition via Athenaeus. The chance that this is the only line that has gone missing from the text of the seven plays is infinitesimal. Some omissions will have been noticed; others will probably elude us forever.

In general, the best guides to the texts available during this period are the papyri. Sophoclean papyri are fewer in number than those of Euripides, and as a result the sample that we have is less useful for drawing general conclusions. Nevertheless, the twenty-nine papyri so far discovered do shed light on the survival of different plays.Footnote ²⁵ The data are valid only for Egypt, and only for certain parts of Egypt; nevertheless, if a play was unavailable there by a certain date, we may wonder how easy it was to obtain a copy of it anywhere else. Of these papyri, only one is Ptolemaic: the Niobe papyrus PGrenf. ii 6(a)+PHib. 11² from El-Hibeh (perhaps written at Oxyrhynchus), dated to between 280 and 240 bc. One more, an Antigone manuscript, PMich. inv. 6585a (perhaps from Oxyrhynchus), comes from the first century bc or ad. The great majority come from the second century ad: papyri of Electra (POxy. 693), Antigone (POxy. 875+3686), Niobe (POxy. 2805), and Tereus (POxy. 5292) from the early second century; of Oedipus the King (POxy. 2180 and PSI 1192, both from Oxyrhynchus), Inachus (POxy. 2369 and PTeb. 692), and The Prophets (POxy. 2453), from the second century; of Ajax (PColon. 251, provenance unknown), Eurypylus (POxy. 1175+2081(b)), Ichneutae (The Trackers; POxy. 1174+2081(a)), and an unknown play (POxy. 2452) from the late second century; and of Ajax (POxy. 2093), Ajax the Locrian (POxy. 3151), Trachiniae (POxy. 1805+3687), and Scyrians (POxy. 2077) from the late second or early third century. Then we have third-century papyri of Trachiniae (P Amst. inv. 68, provenance unknown) and Epigoni (POxy. 4807). A further papyrus, of Oedipus at Colonus (PMich. 140) is alternatively dated to the second–third or fifth–sixth centuries.

From this point onwards the only surviving papyri are those from the seven plays. The fourth century offers us papyri of Ajax (POxy. 1615) and Philoctetes (PBerol. inv. 17058, from Hermupolis); the fifth century papyri of Oedipus the King (POxy. 22) and Philoctetes (POxy. 5266); the fifth or sixth century papyri of Oedipus the King (POxy. 1369), Ajax (PBerol. inv. 21208, from Hermupolis), and Trachiniae (POxy. 3688); and the sixth or seventh century a parchment of Electra (PAntin. 72).

The glut of papyri in the second century – fully fourteen in all, almost half the sample, or seventeen if we include papyri assigned to the late second or early third century – should not be taken as evidence that Sophocles was particularly popular during this period, since we have more papyri of just about every classical author at this time. More significant is the division between second-century papyri of the plays that we have today, which number six, versus eight of the papyri that did not make it past antiquity (which represent seven different plays in all). Taking the papyri which could be late second or early third century together with the definitely third-century papyri, there are five, of which two are from the surviving plays, three from the lost ones (with three separate lost plays represented). From one perspective this emphasizes how the ‘lost’ plays were certainly not lost by the second century, nor even by the third. Nevertheless, roughly fifty per cent of the papyri come from the seven plays which did make it down to our own times, even though these represent only about six per cent of Sophocles’ total output. A severe winnowing had evidently already taken place. Some plays may have been lost already; some will have been easier to access than others. But the process which we observed earlier of some plays becoming more popular than others has by this time reached an advanced stage.

A further winnowing, though one less severe, seems to have taken place even among the seven fortunate plays. Of the seven papyri from the fourth century or later, five are from the three plays which in the Byzantine period would be favoured with the greatest number of manuscripts, namely Ajax, Electra, and Oedipus the King (the so-called triad). This is consistent with the hypothesis that there were more copies of those plays available during late antiquity than of the other four. However, we would need a larger sample to be sure; just a couple of papyri dated to this period from the other four plays would make the arithmetic look rather different.

Some of these papyri are quite stylish. The Eurypylus fragments ‘must have come from a luxury papyrus roll, with considerably broad margins, intercolumnium and interlinear space…[the presence of many diacritics and marginal annotations] suggest that it was a scholar's copy, carefully read and studied’.Footnote ²⁶ The Ichneutae manuscript was in the same format and written by the same hand; we seem to have two survivals from what may well have been an extensive library of Sophoclean tragedy, a treasure-house of which the modern scholar can only dream. Similarly, the Scyrians papyrus came from ‘a handsome manuscript, written in a rather large calligraphic script’,Footnote ²⁷ and the Ajax the Locrian papyrus shows plentiful annotation and scholia in the margins. It is tempting to think that, on the whole, the ninety-four per cent of plays were less generally available than the six per cent, and as a consequence those copies which could be found were more likely to be deluxe editions in the hands of wealthy collectors.Footnote ²⁸ Again, though, we lack the data to give such a hypothesis an adequate test.