First published in 1971, this early second-century papyrus, POxy. 2805, was soon identified as coming from Sophocles’ Niobe.Footnote ¹ It portrays the killing of Niobe's children by Apollo and Artemis during a tragic drama, the staging of which, as we will see, seems to have been among Sophocles’ most striking dramatic creations. This killing, an unvarying element of the myth, was punishment for Niobe's boast that she had borne many children (fourteen, seven sons and seven daughters, is the number specified in Sophocles’ play: fr. 446), Leto only two. A further papyrus, from a mummy case at El-Hibeh (though quite possibly written at Oxyrhynchus) and dated to between 280 and 240 bc, PGrenf. ii.6(a)+PHib. 11, is from the same drama. Aeschylus wrote a Niobe, set after the killings;Footnote ² Euripides is not known to have written such a play. That leaves Sophocles’ Niobe (whose citation fragments indicate that it included the killings) or a play by a minor tragedian; but it is much more likely that two papyri of a Sophoclean drama should survive, from different times and perhaps places too, than that a work by another tragedian should be so unexpectedly favoured. A decade after Barrett argued this, a hypothesis of Sophocles’ play was published which refers to the event described in POxy. 2805: case closed.Footnote ³

The killer urged on in lines four to seven must be Artemis, since she traditionally kills the girls, and is later implored by another girl not to kill her (fr. 442.10–11); the only individual who could urge the goddess on like this is Apollo. The response in lines 8–11 (sung, since the lines are in lyric) belongs to a spectator emotionally affected, but not directly threatened: the chorus.Footnote ⁴ Initially we might think that this scene disproved the convention that death blows were not represented on the tragic stage,Footnote ⁵ since the killing of Niobe's daughter seems to be visibly depicted. The girl, however, is described as hiding ‘inside’ (ἔσω, 4), most naturally interpreted as within the skênê building, the structure at the back of the stage with a single central door. Moreover, under this scenario it is difficult to see how the shootings could have been staged. Shooting arrows, whether real or imagined, inside the skênê door, where their impact would not be seen, is one thing; but shooting an invisible arrow intended to reach its target in full sight of the audience is quite another. Niobe's daughters suddenly collapsing in the face of unseen darts might seem more comical than tragic; the same would hold for deliberately blunt arrows, ricocheting off their targets in a manner unlikely to generate awe at the divine. By contrast, arrows visibly launched but invisibly reaching their targets may be compared with the suicide of Evadne in Euripides’ Suppliant Women, where ‘the audience see her leap, but do not see her land’.Footnote ⁶

If the doomed Niobid is inside, where are Apollo and Artemis? On the skênê building, or at ground level? In his analysis of actors appearing on top of the skênê, Mastronarde includes our fragment among ‘instances from lost plays for which there is insufficient information’, tentatively describing Apollo and Artemis as ‘on stage, shooting into palace door?’Footnote ⁷ In Barrett's view, however, they appeared on the skênê building,Footnote ⁸ which would be visually more impressive. Whether it would be possible from that vantage point to see into a store-room is irrelevant;Footnote ⁹ the dramatic force of the positioning was the primary concern, and Apollo and Artemis shooting from on high would create an extraordinary spectacle, vividly conveying their complete control of the situation, their fatal arrows meting out justice from above as if in lieu of Zeus's thunderbolts.

The intensity of Apollo's words matches that of the staging. His description of the Niobid does so much more than merely point her out. It characterizes her as female, afraid, hiding, alone, cowering: a powerless opponent for the vengeful goddess. The contrast between the gods standing tall on the skênê building and the little girl vainly concealing herself is palpable; the contrast depends on the interplay of words and staging, since only one of the two females can be seen. The specification of the girl's location as inside the πιθών (a store-room for πίθοι, vats used for holding wine) among the κυψέλαι (earthenware containers for storing grain), involving ‘two everyday Attic words (both in old comedy)…previously unattested from tragedy: tragedy has normally no reason to penetrate into the domestic stores’,Footnote ¹⁰ chillingly juxtaposes the supernatural phenomenon of the divinities’ outsized wrath and the ordinary features of a banal everyday home. According to a probably familiar story in Herodotus, the Corinthian tyrant Cypselus was saved as a baby from a group of assassins by being hidden in a chest (κυψέλη), from which he took his name.Footnote ¹¹ But no protection is afforded to Niobe's daughter by the same item; she faces no mere gang of thugs, but the divine children of Zeus and Leto.

The Hibeh papyrus offers a further glimpse of this encounter:

This passage too comes from the killing scene – note the cry in the second line, whether from the chorus or a Niobid – with Apollo stating in lines 3–4 that he can see a pair of girls, denoted first by τάσδ’, then by some phrase such as τὴν μὲν κρέκουσ]αν, τὴν δὲ̣[ κνάπτουσαν ϕάρη (‘this one weaving, the other carding robes’); in the last line ε]ὐ̣αγρ[ο]υ̣ ϕω[ράματος (‘of a creature detected, an easy prey’) is tempting.Footnote ¹² Whatever actions are predicated of the girls, their brief description makes everyday tasks more likely than any reaction to the gods’ presence; hence the killing spree was at an early stage, before Apollo urged on his sister against the girl in the store-room.Footnote ¹³ So if Apollo and Artemis are on the roof for the action recorded in the Oxyrhynchus papyrus, they must already be there during the Hibeh papyrus too, meaning a minimum of four successive killings from that exalted position.

A further Hibeh fragment (fr. 442 TrGF) seems to depict Niobe (inside the skênê) begging Artemis or Apollo not to drive her out of the house, but to slay her; one of her daughters (also presumably inside) begging Artemis not to kill her; and the chorus outside, commenting on the grim exclamations within. By this stage Apollo and Artemis will have descended from the skênê roof and are to be thought of as within the house itself:

Having Niobe's pleas for death punctuated by her daughter's pleas for survival creates new pathos: both pleas, though opposites, are understandable, but neither was granted. Staging again reinforces that pathos: ‘the ultimate victim, Niobe, should be behind the scenes with the daughters whose death is her disaster’, as the audience imagines the daughters being killed in their mother's presence.Footnote ¹⁵ Another Hibeh fragment (fr. 444 TrGF) suggests that one daughter escapes, a version of the story known from the late archaic poet Telesilla of Argos;Footnote ¹⁶ her solitary appearance after so much slaughter further contributed to the shocking scene.

The hypothesis published in 1984 reveals that the slaughter of Niobe's daughters ‘was only the climax to an ascending series of horrors’.Footnote ¹⁷ Early in the drama Niobe sends her sons out hunting, making her boast as she does so, and it is on this hunt that they are killed; a Hibeh fragment, fr. 443, may come from a messenger speech to Niobe or Amphion (Niobe's husband) describing the killings of the sons (who perhaps vainly attempt to battle the gods), delivered as, or immediately before, the corpses are returned to the palace (something mentioned in the hypothesis). The killing scene involved one of the boys invoking his elder male lover (ἐραστής) for help (fr. 448); since the play had an alternative title, Paederastia, ‘the moment cannot have been merely incidental…the theme must somehow have been fully developed’.Footnote ¹⁸ Amphion reproaches Apollo and challenges him to combat, taking up arms against him when he does appear, before being shot dead in his turn. Only then comes the killing of the daughters, to emphasize (the hypothesis states) that pre-eminence belongs to the gods. The pattern whereby some children are killed off-stage, then others are killed closer to home, can be paralleled in Euripides’ Ino, in which Athamas kills his son Learchus off-stage, during a hunt (something almost certainly narrated before the arrival of the body, either by a messenger or by Athamas himself), and his dead body is brought back on stage; Athamas’ wife Ino later leaps with their other son, Melicertes, into the sea.Footnote ¹⁹ It involves a typically tragic intensification; compare how in Euripides’ Electra a messenger narrates Aegisthus’ death ahead of the killing of Clytemnestra, which is heard behind the skênê building.

Sophocles’ Niobe thus combines two tragic scene-types – deus ex machina (whereby gods appear on high to settle matters) and off-stage homicide scene (in which a character is killed behind the skênê)Footnote ²⁰ – while giving each a bitter twist. Here the gods above dispense no wise summing-up of the action, no lofty perspective on mortal affairs, but rather mayhem and violence. True, the daughters’ slaying is a form of justice, in that it punishes Niobe for her boast while simultaneously nullifying its content. But the daughters made no boast; and while the idea that the sins of parents could lead to disaster for their children was certainly familiar from mythology,Footnote ²¹ someone in classical Greece who attacked an offender's innocent child could expect severe punishment. As for the off-stage homicide scene, this is the only one where the death-dealing actions begin their fatal course in the audience's sight. The effect is to emphasize the gods’ dominance, especially if they are standing on the skênê for part of the killing. Through this technique, ‘Sophocles has come as close as he can to presenting violent death on stage, without actually doing so’;Footnote ²² no slave of staging conventions he, but their master.

After the daughters’ deaths, ‘Zethus …[enters] at the end of the play in place of a deus ex machina’ (as per the last line of the hypothesis, π]αραγε̣ν̣[όμ]ε̣νος δὲ Ζῆθ̣ος Νιόβη̣ν μ̣ε̣ν̣).Footnote ²³ He must have been involved in arranging Niobe's transfer to Lydia (known from a testimonium to be part of the play: TrGF iv p. 363), and commented on the action and summed it up, as far as that was possible in the circumstances. Cockle's suggestive comparison of his role to that of a deus ex machina, if the play bore it out, would reflect how the gods have been too directly involved in the main action to perform such a function; instead it falls to a man who can at least assist Niobe on her way, unable though he is to provide comfort beyond this.

Thus in Niobe Sophocles ‘produced one of the most spectacular coups de théâtre ever seen on the Athenian stage’, in terms of the harrowing massacre of the daughters and the events that follow and precede;Footnote ²⁴ the three papyri (more than for any other lost Sophoclean play) reveal Sophocles’ manipulation of stage convention, and his melding of staging with words to create powerful emotional effects. He brings on stage a matter which Aeschylus had only narrated, since Aeschylus’ Niobe is set after the killings: compare the urn supposedly containing the dead Orestes’ ashes in Electra (which in Aeschylus’ Libation Bearers receives only a mention), and Ajax's suicide speech (described by a messenger in Aeschylus’ Thracian Women).Footnote ²⁵

That latter death constitutes a central problem of Sophoclean stagecraft.Footnote ²⁶ An ancient commentator claimed that in Sophocles’ play Ajax kills himself on stage, causing amazement in the spectators.Footnote ²⁷ Such a staging is problematic, since no other tragedy shows a death blow (not itself an insurmountable objection, though, since some conventions can be broken). Moreover, immediately after the suicide the chorus arrives, searching for Ajax, whose body is discovered by Ajax's concubine, Tecmessa, whom the chorus cannot see even after her initial cry, which (they say) comes from a grove; this stage action is inconsistent with Ajax killing himself in view of the spectators. But what actually happened is still extensively debated.Footnote ²⁸ Was the grove some area covered in bushes to the front of the skênê building? Or was it the skênê itself, into which Ajax departs to kill himself? Or was some screen, painted to indicate a grove, set up by scene-shifters in front of the skênê, behind which Ajax's suicide took place? Or was the ekkuklêma, a wheeled platform that depicted on stage scenes imagined as taking place indoors, deployed, decked with bushes, and pulled inside immediately after Ajax fell upon his sword? Was Ajax's sword visible to the audience or not as he delivered his soliloquy?Footnote ²⁹ Discussion of these and other issues remains intense and unresolved.

That sword is one of a series of Sophoclean stage properties – Philoctetes’ bow, Electra's urn, Oedipus’ changing mask, Deianira's robe – whose significance has been such a concern of modern scholarship.Footnote ³⁰ To these we may add the masks of the title characters in Tyro Footnote ³¹ and in Thamyras,Footnote ³² which may have changed during their plays to indicate the violence done to their wearers, as well as Tecmessa's robe or veil in Ajax, which she removes shortly after she discovers Ajax's body:

What is the ϕάρος, the ‘shroud’, which Tecmessa places on Ajax? Traditionally translated ‘cloak’, the rendering ‘veil’ would be just as good;Footnote ³³ and such an action would visually recall how in the Iliad Andromache removes her veil at the news of her husband Hector's death.Footnote ³⁴ Tecmessa's language elsewhere in the drama recalls Andromache's from the Iliad;Footnote ³⁵ perhaps her action had the same effect, with words and staging mutually reinforcing the Iliadic allusion, accentuating the pathos and highlighting Tecmessa's status as in effect Ajax's widow.Footnote ³⁶

Interest in matters of staging is not new: the debate over the staging of Ajax's suicide, for example, has been ongoing for more than two centuries. In 1917 Tycho von Wilamowitz-Moellendorff's Die dramatische Technik des Sophokles (The Dramatic Technique of Sophocles) was published posthumously by his father Ulrich.Footnote ³⁷ Tycho's Sophocles accepted inconsistency of character while he focused on his main goal: creating emotional reactions in the audience through spectacular dramatic effects. Not everyone was immediately convinced of the need to appreciate Sophocles from a theatrical point of view; nearly two decades later, T. B. L. Webster could still write ‘since we are readers rather than spectators, we need not go too deeply into…technicalities of production’, a remark left unchanged in the second edition of the book in question.Footnote ³⁸ But with the publication of Oliver Taplin's The Stagecraft of Aeschylus. The Dramatic Use of Exits and Entrances in Greek Tragedy in 1977, this state of affairs changed for good; although not a book about Sophocles, it galvanized theatrical studies on Greek tragedy as a whole. Stagecraft was succeeded in the following year by Greek Tragedy in Action, a book intended for a broader audience which did deal explicitly with plays by Sophocles, as well as Aeschylus and Euripides. Before the books, Taplin's first article was on Sophoclean stagecraft: ‘Significant actions in Sophocles’ Philoctetes’ (1971) is still an essential guide to the contribution of staging to that play's overall effect.Footnote ³⁹ The gratitude expressed in that article to Eduard Fraenkel, who had died the previous year, is significant: Fraenkel's edition of Aeschylus’ Agamemnon, often considered a monument of purely philological inquiry, refers to the ‘grammar of dramatic technique’, to which any student of tragedy must pay attention.Footnote ⁴⁰

Despite the proliferation of studies of individual points of staging, in articles, monographs, or commentaries, there is still no book-length study of Sophoclean dramatic technique comparable to that of Taplin for Aeschylus. Marshall 2012 provides an incisive overview, however, which examines the plays as works constructed by a theatre director (didaskalos), drawing on not just the seven plays but on the totality of the evidence for Sophocles’ skill as a dramatist. Any future study of the subject will have to follow Marshall's lead. This union of two approaches to the plays, the theatrical and the fragmentary, has the potential to improve our understanding considerably.

So too does a further line of inquiry which has recently begun to receive consideration: the question of the early reperformance of tragedy, which was taking place already in the fifth century, earlier than has often been assumed.Footnote ⁴¹ Paying attention to this issue opens up all kinds of questions. What proportion of Sophocles’ plays was reperformed during his lifetime? To what extent did Sophocles take into account the possibility of reperformance when composing his plays? What did an audience member get out of seeing a play performed for a second time? How did fifth-century tragic reperformance intersect with long-standing traditions of epic or lyric reperformance? What about the reperformance of extracts, at symposia for example? How early were Sophoclean plays reperformed outside Attica? Asking questions is one thing, suggesting answers quite another; but here at least is another approach to Sophoclean tragedy the implications of which have only begun to be teased out.