The Phantom Senex

The Non-Existence of the Old Man in Seneca’s Troades

In: Mnemosyne
Robert Cowan The University of Sydney, Department of Classics and Ancient History Sydney Australia

Search for other papers by Robert Cowan in
Current site
Google Scholar
Open Access


The character designated by the manuscripts as Senex, who accompanies Andromache and Astyanax in act three of Seneca’s Troades, is problematic in many ways. He is not identified or acknowledged by any other character; his entrance and exit are unannounced; his presence onstage in the first half of the act requires that Astyanax’s two words of dialogue be delivered by a fourth actor or through ventriloquism; his very existence conflicts with the obvious interpretation of at least two sections of Andromache’s dialogue. All of these anomalies can be removed if there is in fact no Senex and the dialogue attributed to him by the manuscripts is spoken by the Chorus leader. This level of involvement in the action by the Chorus would itself be unusual in Senecan tragedy, but it does have parallels and would also fit with the exceptional treatment of the Chorus throughout Troades.


The character designated by the manuscripts as Senex, who accompanies Andromache and Astyanax in act three of Seneca’s Troades, is problematic in many ways. He is not identified or acknowledged by any other character; his entrance and exit are unannounced; his presence onstage in the first half of the act requires that Astyanax’s two words of dialogue be delivered by a fourth actor or through ventriloquism; his very existence conflicts with the obvious interpretation of at least two sections of Andromache’s dialogue. All of these anomalies can be removed if there is in fact no Senex and the dialogue attributed to him by the manuscripts is spoken by the Chorus leader. This level of involvement in the action by the Chorus would itself be unusual in Senecan tragedy, but it does have parallels and would also fit with the exceptional treatment of the Chorus throughout Troades.

Troades is Seneca’s most exceptional tragedy and its third act the most exceptional act in the corpus.1 Not least among its anomalies is the character designated by the manuscripts as Senex, who accompanies Andromache and the (almost) mute Astyanax and plays the restraint role in a passion-restraint scene with the former. Many aspects of this character are problematic, anomalous, and in some cases unique. He is not identified or acknowledged by any other character; his entrance and especially his exit are unannounced and problematic; he is (arguably) the only anonymous character in Senecan tragedy who does not have a precedent in an equivalent extant Attic tragedy;2 his presence onstage in the first half of the act requires that Astyanax’s two words of dialogue be delivered either by a fourth actor or through some awkward ventriloquism; his very existence conflicts with the obvious interpretation of at least two sections of Andromache’s dialogue. This article will suggest that all of these anomalies can be removed if there is in fact no Senex and the dialogue attributed to him by the manuscripts (and following them, all editions) is spoken by the Chorus leader. This level of involvement in the action by the Chorus would itself be unusual in Senecan tragedy, but it does have parallels and would also fit with the exceptional treatment of the Chorus throughout Troades.

1 Problems with Stagecraft and a Solution

Senecan stagecraft is problematic, not least because of the continuing lack of consensus about the tragedies’ primary mode of performance.3 Even for those who strongly advocate either recitation or full staging, the precise mechanics of either remain uncertain. Comparisons with the stagecraft of Attic tragedy are at best problematic. It should be emphasized that all references in this article to the dramaturgical practices of Attic tragedy are intended as independent examples for comparison or contrast. They do not imply that Senecan stagecraft was closely influenced by or similar to that of Aeschylus, Sophocles, and Euripides, a notion exploded by Tarrant and by many scholars since.4 Entrances and exits in Seneca are not always as explicitly indicated in the dialogue as in Attic tragedy, though they frequently are. The unannounced nature of the Senex’s entrance and of his exit, taken individually, are thus far from unique in themselves.5 Nevertheless, they have generated particular unease and displeasure in critics, perhaps most pithily expressed by Zwierlein: “Von dem Senex hören wir nichts mehr: so unverhofft wie er dastand, ist er auch wieder fort.”6 To omit indications of either entrance or exit may be regarded as a misfortune; to omit both looks to Zwierlein like carelessness. Instead of attributing this to Seneca’s incompetence in or indifference to stagecraft, I would suggest that the fact that neither entrance nor exit is reflected in the dialogue in any way may be taken as the first (pair) of many clues that this character who leaves no trace may not actually exist. Fantham is not of course herself proposing such a radical scenario, but her comments are extremely suggestive: “Only the old man’s unannounced question tells us that he is present; there is no hint of his origin, or of his departure or destination after his last words at 518.”7 It is but a small step from the question’s being the only indication of his presence to its being the only indication of his existence. Since that question is not only unannounced but expected to come from a different source, the grounds for believing in his existence are tenuous indeed.8

A more overtly problematic aspect of the Senex’s dramaturgy centres on the two words spoken by Astyanax at line 792: miserere, mater. Andromache is onstage throughout act three and is therefore played by a single actor, who is also onstage when Astyanax speaks. The Senex announces the approach of Ulysses at 517-518, at which point he would almost certainly be visible to the audience, so that even a lightning change could not enable them to be played by the same actor, even if he were to exit soon afterwards.9 This accounts for the three actors who are sufficient for all but two scenes in Senecan tragedy.10 However, Astyanax is onstage with the Senex from 409 (or 418-419 at the latest, with its deictic hic) to 512, and with Ulysses from 707 (hic again) till the end of the act, including his single speech. He cannot therefore be played by the same actor as the Senex unless a persona muta were used in the earlier scene and the Senex-actor, having exited sometime after 518, takes over the role and comes out of Hector’s tomb at 707. The latter is not impossible but would go against Senecan practice and, if we envisage a full staging, the relative heights of the actors could be a problem. Sutton asserts that there is no difficulty: “The single utterance of Astyanax in Act III (792) is no more problematic than similar child-utterances in Greek tragedies such as Euripides’ Medea. The Senex is not onstage during the long interview between Andromacha and Ulixes so the tritagonist can speak the boy’s line from offstage.” The comparison with Greek tragedy is less problematic than at other points in Sutton’s reconstruction of Senecan stagecraft, since it only indicates what is possible. However, Sutton is slightly misremembering, since Medea’s two sons cry out offstage, just like adult murder victims such as Sophocles’ Clytemnestra, and their lines can be delivered by one of the three actors without any ventriloquism (i.e. pretending that a different person’s voice is coming out of the silent child actor’s mouth) such as would be needed in Troades.11 There are speaking or rather singing roles for children in Euripides’ Alcestis, Andromache and Suppliant Women, but these seem to have been played by actual child actors and in any case their more extensive involvement does not offer a good parallel for Astyanax’s brief outburst.12 None of these solutions is impossible. Tecmessa in Sophocles’ Ajax is played initially by a speaking actor and then by a persona muta, so the same role-sharing in the opposite order would be possible for Astyanax.13 Euripides’ Hecabe and Electra each feature a mute attendant who re-enters as a speaking (and in the latter case also singing) character.14 The ventriloquism may not be paralleled but neither is it totally unfeasible.15 Alternatively, a persona muta, whether child or adult of short stature, could probably be safely entrusted with a mere two words of dialogue. Nevertheless, it remains an anomaly that requires explanation, another in a series of anomalies whose cumulative evidence tells against the existence of the Senex, and another whose problems can be simply solved if the lines attributed to him are instead delivered by the Chorus.

It is striking that the dramaturgical handling of the Senex is equally problematic in different ways for those who favour recitation (or even reading) as the primary form of Senecan performance and those who argue for full staging. If Troades was a Rezitationsdrama, then the absence of any verbal indication of his identity would be confusing to the audience. Sutton relies on costume and proxemics to clarify the situation: “the individual called by our manuscripts the Senex is Astyanax’ paedagogue and his identity would visually be established both by his age and by the way he hovers protectively about the boy.”16 Yet this assumes staging and underlines the problem that would be faced by a reader or listener with no visual cues. For those who believe the tragedies were staged, the Senex’s identity is unproblematic, but his unannounced exit is. Once more, it is revealing to see how a proponent of the other performance context identifies the problem and uses it as a justification for their preferred solution. Fantham writes, “he is silent thereafter, and no departure is indicated. As a subordinate he naturally keeps silent while his superiors dispute, but in a stage production the dramatist would have found a pretext to take him away once his dramatic usefulness was ended. Writing for recitation, however, Seneca can afford simply to let him be forgotten.”17 As we have seen, hypothetical recitation audiences would have their own problems, but Fantham clearly shows a major issue for the equally hypothetical theatre audiences. The Senex is a problem for every theory about Senecan performance, but that problem can be solved very effectively if the Chorus takes his place.

2 Anonymous Characters in Senecan Tragedy

The elusiveness of the Senex is not limited to his sudden, unexpected, and unannounced entrance and exit. There is no indication in what he or Andromache says that the dialogue is being conducted between two individuals, let alone with an old man or paedagogus specifically. It is true that Seneca’s anonymous characters are rarely as clearly identified as their counterparts in Attic tragedy. There is little resembling the ‘implicit stage directions’ given by the Chorus of Euripides’ Hippolytus, which simultaneously identify the Nurse and tell the audience (as well, perhaps, as reminding the actors) what she is doing: ἀλλʼ ἥδε τροφὸϲ γεραιὰ πρὸ θυρῶν | τήνδε κομίζουϲʼ ἔξω μελάθρων (‘but this is the old nurse before the doors bringing her out of the house’, E. Hipp. 171-172).18 The reasons for this lie partly in the very different dramaturgy of Senecan tragedy and partly in the solipsistic nature of its protagonists. Nevertheless, the identity of all these Senecan characters is made clear at some point or other.19 They are often either addressed by another character or by the Chorus with an identifying vocative, or they themselves address another character in a way that makes their relationship to them clear. The Nurses of the eponymous Medea and Phaedra are called nutrix or altrix by their respective mistresses and, in the latter case, also by the Chorus, while they in turn refer to their former charges as alumna.20 The evidence of the non-Senecan Octavia and Hercules Oetaeus must be treated with caution, especially since they contain some markedly un-Senecan dramaturgical features, such as final choruses. However, it is notable that they replicate—and can be plausibly considered to imitate—this aspect of Senecan practice when the Nurses of both Octavia and Poppaea likewise refer to their alumna and are referred to in turn as nutrix, while Deianira’s altrix calls her alumna.21

Seneca’s Messengers tend to have their identity made clear by their generic function, reporting off-stage action, generally in the fourth act. Their distinction from the Chorus is also underlined by their tendency to enter immediately following a choral ode and provide new information to an audience of which the Chorus forms an important part. Indeed, in Thyestes, the Chorus is the sole recipient of the Messenger’s report of Atreus’ crime and enters into intermittent dialogue with him.22 The severely truncated Messenger scene in Medea also consists of dialogue between Messenger and Chorus.23 Even more clearly, the Choruses of the Phaedra and Oedipus identify the Messenger as he enters, while that of the Agamemnon introduces Eurybates by name.24 The two Messenger scenes in Troades itself may feature the same or different Messengers, and the speaker may be Talthybius in one, both, or neither, an issue to which we shall return. Regardless of his precise identity, each Messenger is clearly distinct from the Chorus, the first being questioned by them, and the second closing his scene (and the whole play) by commanding them to board the ships that will take them as slaves to Greece.25

Two anonymous characters are less clearly identified. Like Phaedra’s, Medea’s, Deianira’s, and Octavia’s Nurses (and to some extent like the phantom Senex),26 they are the champions of restraint in a passion-restraint scene and their lack of definition can largely be accounted for by the solipsism of the character with whom they are futilely arguing. Perhaps the most solipsistic of all Seneca’s monomaniacs is Atreus in Thyestes and it is unsurprising that he has little interest in characterizing the colourless Satelles who vainly tries to urge conventional morality on him before being caught up, like everyone else in the play, by the tyrant’s charismatic villainy.27 Nevertheless, Atreus does address the Satelles directly with surprising frequency and uses second-person singular verbs to refer to him on seven occasions.28 It is true that the play’s Messenger also addresses the Chorus (or Chorus leader) with second-person singular verbs, but these are the more generalizing putas and putes and he does also use the unambiguous second-person plural exhorruistis.29 Although the Senecan chorus was probably much smaller than the Attic, perhaps consisting of as few as three members, it was still a plurality.30 In contrast, the Satelles is repeatedly depicted as singularly speaking, fearing, erring, and believing. Above all, the final injunction to keep Atreus’ plans secret includes the marked singular pronoun tu. Even without these more subtle indications, the differentiation between the Satelles and the Chorus is clearly indicated by the latter’s notoriously and naively optimistic ode about the brothers’ reconciliation. Davis rightly uses the Chorus’ ignorance of Atreus’ schemes as evidence of their absence in the preceding act.31 It is a fortiori evidence that they were not Atreus’ interlocutor as he developed and described those schemes.

An even more elusive figure is the Nurse in Agamemnon. She conforms in many ways to the pattern of the Nurses in Phaedra and Medea, and of Atreus’ Satelles: the protagonist’s soliloquy immediately following the first choral ode is itself followed by the intervention of the anonymous interlocutor, which segues into a duel of sententiae in stichomythia (often further fragmented by antilabe, though not in this scene). The departures from this pattern in Agamemnon—the Nurse’s acknowledgment that Clytemnestra’s speech is notionally inaudible to her, the lack of resolution between the opposing parties at the end of the agon—are interesting in themselves but do not affect the normative quality of the Nurse’s depiction as an individual interlocutor. These formal features are in themselves sufficient to establish her existence and her role. As will be noted in greater detail below, act three of Troades departs from the pattern in ways that are significant for the nature of the interlocutor, above all the fact that Andromache’s opening speech is not a soliloquy but is addressed to the Chorus. It should be observed, however, that, unlike other Nurses, Clytemnestra’s does not address her mistress as alumna and in turn is not addressed as nutrix or altrix. However, this is part of the play’s wider emphasis on Clytemnestra’s regal status rather than an indication that her interlocutor is not in fact a Nurse.32 Even if the formal features were not sufficient, the following stichomythic exchange incidentally makes the identities of the speakers unambiguous:

Nvt. Tuta est latetque culpa, si pateris, tua.
Cl. Perlucet omne regiae uitium domus.33
Nurse: Your guilt is safe and lies hidden, if you will let it.
Clytemnestra: Every vice of a royal house blazes forth.

This insistence that Clytemnestra’s past adultery can be kept hidden, provided that it is not compounded by future regicide, could hardly be made by a Chorus of Argive maidens serving as a synecdoche for the wider population of Argos, but makes perfect sense when delivered by an individual close confidante.

This brief survey of Seneca’s handling of anonymous characters is not intended to show that he carefully marks their identity and the distinction between them and the Chorus. If anything, quite the opposite. Whether these were Rezitationsdramen with placards naming new characters or staged dramas with costumes clearly identifying their status, the audience would know that an individual was speaking and in most cases who they were without hearing them addressed as tu or nutrix. Rather, these pronouns, titles, second-person verbs, formal patterns, and contextual details are precisely the sort of incidental and accidental features that inevitably arise in the creation of dialogue between two individuals, even the peculiar dialogue of Senecan tragedy where “characters are juxtaposed one to the other and … do not give the impression of interacting, but rather of speaking in isolation”.34 Their absence from the scene between Andromache and her interlocutor in the opening sections of Troades act three is not a deliberate signal to the audience. If it were, and if Seneca were in general interested in such dramaturgical markers, the speaker’s identity would surely have been indicated much more clearly and unambiguously. Rather, the aforementioned incidental and accidental features do not arise, as they consistently do elsewhere, because Andromache is not speaking to the phantom Senex or to any other individual.

3 Textual Evidence

It could still be objected that the absence of evidence is not evidence of absence, and that it is statistically possible in a relatively short exchange for none of these incidental features to occur. However, there are also two pieces of positive evidence in the text itself that point strongly, in one case to the absence of another individual onstage, in the other to its being the Chorus with whom Andromache is conversing. When the latter is striving to persuade Astyanax to overcome the pride that she is probably projecting onto him (there are more plausible reasons for the boy’s reluctance to enter his father’s tomb) and accept the inglorious concealment, she conjures for him—and indirectly for the audience—a strong visual image:

succede tumulo, nate—quid retro fugis?
turpesne latebras spernis? agnosco indolem:
pudet timere. spiritus magnos fuga
animosque ueteres, sume quos casus dedit.
en intuere, turba quae simus super:
tumulus, puer, captiua: cedendum est malis.35
Go down into the tomb, my son—why do you flee back?
Is it the disgraceful hiding place you scorn? I recognize your nature:
you are ashamed to be afraid. Drive off lofty spirits
and your old dispositions, take on those which chance has given.
Look! See what a crowd we are remaining:
tomb, boy, captive. It is necessary to yield to evils.

Commentators rightly note the bitterly ironic conceit of the word turba’s being applied to such a small number and its relationship to Ovid’s similar use of the term to describe Deucalion and Pyrrha.36 However, it remains remarkable that Andromache does not include the Senex in her turba of three. The invisibility of slaves—and even the captiua Andromache would presumably consider a paedagogus as such—is a familiar and notorious aspect of Roman culture, so it is possible that Andromache thinks of the Senex as a non-person.37 In addition, Andromache’s pathological obsession with her dead husband, even at the expense of her living son, is in many ways her defining characteristic in the play, so that it is not surprising in itself that her focus should be on Hector’s tomb rather than on an old man standing beside her.38 Nevertheless, even making these allowances, it is surely intolerable for an audience looking at a tableau of three figures onstage, or listeners who have conjured an equivalent stage-picture in their minds, to hear one of them speak of a turba of three that omits one of those who are onstage and adds one (or a monument standing surrogate for him) who is absent. Seneca’s love of paradox and the para prosdokian is famous, but this is jarring without being pointed, more inept than startling. In contrast, if Andromache’s interlocutor is the Chorus, they constitute a cohesive group in themselves who can be ignored, set physically apart, or at the very least easily conceptualized as distinct from the meaningful turba of three, the poignant tableau of tomb, boy, captive.

Andromache’s definition of her turba of three offers a good reason for identifying the Chorus as her interlocutor, but still retains its main force as a reason for that interlocutor not to be the phantom Senex. There is another item of textual evidence that offers more unequivocally positive evidence of the Chorus’ role in this act. We have already noted the ways in which the first half of Troades act three differs from other Senecan passion-restraint scenes in its treatment of the protagonist’s interlocutor. It also differs in its treatment of the protagonist herself. The equivalent scenes in Phaedra, Medea, Agamemnon, and Thyestes all begin with a soliloquy by the protagonist that betrays no awareness of or interest in the interlocutor’s presence. The interlocutor then immediately challenges the course of action proposed or train of thought expressed in the soliloquy.39 Contrast the opening of our scene:

An. Quid, maesta Phrygiae turba, laceratis comas
miserumque tunsae pectus effuso genas
fletu rigatis? leuia perpessae sumus,
si flenda patimur. Ilium uobis modo,
mihi cecidit olim, cum ferus curru incito
mea membra raperet et graui gemeret sono
Peliacus axis pondere Hectoreo tremens.
tunc obruta atque euersa quodcumque accidit
torpens malis rigensque sine sensu fero.
Iam erepta Danais coniugem sequerer meum,
nisi hic teneret: hic meos animos domat
morique prohibet; cogit hic aliquid deos
adhuc rogare, tempus aerumnae addidit.
hic mihi malorum maximum fructum abstulit,
nihil timere: prosperis rebus locus
ereptus omnis, dura qua ueniant habent.
miserrimum est timere, cum speres nihil.
Sen. Quis te repens commouit afflictam metus?40
An. Why, mournful crowd of Phrygia, do you tear your hair,
beat your wretched breast and drench your cheeks
with a flood of tears? Light are the things we have endured,
if we endure things to be wept over. For you Troy fell just now,
for me long since, when, savage in its swift chariot
it snatched my limbs and groaned with deep tones
the Pelian axle, shaking with the weight of Hector.
Because I was then crushed and demolished, whatever happens,
numbed by misfortunes and desensitized, I bear without feeling.
Now, snatched away from the Danaans, I would be following my husband,
if this one did not hold me: this one tames my intentions
and prevents me from dying; this one forces me to ask the gods
for something still, he has added time to my suffering.
This one has taken from me the greatest benefit of misfortunes,
to fear nothing: every opportunity for prosperity
is snatched away, hard times have a path by which they can come.
The most wretched thing is to fear, when you have no hope.
Sen. What sudden fear has shaken you in your misfortune?

All other protagonists talk to themselves and address their own animi.41 Indeed, both Ulysses and Andromache will address their respective animi in asides later in this act.42 However, here at the start of the act, Andromache directly addresses the Chorus. In all the other passion-restraint scenes, the Nurses and the Satelles intervene unprompted to offer unsolicited advice in favour of a change of mind. The interlocutor here responds directly to a speech directed at a specific addressee. Andromache asks the Chorus why they still weep, then proceeds to explain to them (note that uobis maintains the address) that she, in contrast, has been numb since the death of Hector but now she has a new fear prompted by Astyanax. The interlocutor does not, as elsewhere, cut across this speech by telling her not to be afraid.43 Rather they respond to Andromache’s speech directly by asking what the new fear to which she refers is. It makes infinitely more logical and dramatic sense for the response to come from the person (or persons) to whom the speech was addressed, the Chorus, than from an unannounced Senex.44 Zwierlein makes almost precisely this point about the inappropriateness and unexpectedness of the respondent, but uses it as a further example of how the tragedies are more suited to recitation than staged performance.45 Whether Troades was recited or staged, it is surely preferable to assign the response to the appropriate, expected character rather than to scoff at that of the inappropriate, unexpected one.

This initial response by the Chorus naturally leads into the further questions they have for Andromache and, beyond that, advice on the concealment of Astyanax. Indeed, the scene as a whole possesses few features of a passion-restraint scene. In addition to the initial questioning and later advice, the closest the interlocutor comes to ‘restraining’ Andromache’s ‘passion’ are the gnomic reassurances in response to her fears about hiding Astyanax in the tomb and even these are mostly practical advice rather than moralizing maxims.46 Thus even the formal expectation that the restraint-role be taken by a slave or social inferior does not come into play, though it seems likely that the Chorus does consist of lower-status Trojan women rather than unspecified daughters and daughters-in-law of Priam.47 The Chorus’ taking the role of a different sort of interlocutor would feel accordingly less strange to the audience.

4 The Senecan Chorus

This level of involvement by the Chorus would still be unusual in Senecan tragedy. The sort of extended dialogue we find between the Coryphaeus and Aeschylus’ Eteocles, Sophocles’ Tecmessa, or Euripides’ Heracles, to name but a few, is largely absent in Seneca.48 In contrast, the Senecan Chorus’ main function is to deliver choral odes, and its integration into the action or even the world of the play is often doubted.49 However, it does occasionally play a significant role outside the choral odes and these offer good parallels for what we find in Troades act three. The handful of instances where the Chorus speaks dialogue in the middle of an act to provide descriptions of onstage action are not closely comparable, since they constitute a sort of extradramatic function aimed directly at the audience and do not involve interaction with other characters or a contribution to the action.50 Even so, it is notable that this narrative role is more frequently taken by an individual (non-Messenger) character, so that this too is evidence that the Chorus can occasionally fulfil functions more generally associated with individuals.51 On other occasions, the Chorus speaks directly to a character, aiming to change their course of action, as when it advises Amphitryon to avoid being killed by Hercules, Phaedra to stop complaining and honour Diana, and Theseus to give due burial to Hippolytus’ dismembered body.52 As we shall see below, the first of these instances resembles the intervention of an individual character so closely that the manuscripts assign the lines impossibly to Theseus. Also relevant are the scenes where the Chorus acts as interlocutor with the Messenger, as in the brief exchange in Medea and the even briefer one in act two of Troades itself.53 The most extensive of these is that in Thyestes, where they speak a total of fifteen lines (rounding up part-lines), one more than the fourteen lines delivered by Andromache’s interlocutor, while both make nine discrete speeches.54

Critics generally insist that these scenes follow an unwritten rule in Seneca that a character only interacts with the Chorus if there is no one else onstage, though there are clear exceptions, as when the Chorus describes Jocasta’s behaviour to the blind Oedipus, and disputed cases, as when Medea and the Nurse may enter before or after the brief scene between the Messenger and the Chorus.55 If there is no Senex, Andromache would appear to the audience to be alone onstage with the Chorus (until the entrance of Ulysses, after which the interlocutor is silent or perhaps departed) except for the apparent persona muta of Astyanax. With the third actor freed up by not having to play the phantom Senex, he could now play Astyanax, but the audience would have no reason to suspect this. As a result, the audience’s shock at his brief outburst at 792 would be intensified and the effect of an ostensible persona muta’s being revealed to be a speaking actor would be comparable to that of Pylades’ intervention in Aeschylus’ Libation Bearers.56 There are thus excellent parallels for the Chorus’ active (but brief) intervention in the action and for its extended (but passive) involvement in dialogue. The combination of the two in act three of Troades is comprehensible within the limits of Senecan practice, but still exceptional. This is not surprising, since, as Davis puts it, “[t]he handling of the Chorus in Trojan Women is unique in Senecan tragedy”.57

In addition to their brief dialogue with the first Messenger (whom the manuscripts designate Talthybius, a point to which we shall return), which we have already noted, they are addressed directly in act four when Andromache ironically commands them to celebrate with groans and tears the fictional wedding of Pyrrhus and Polyxena that is a cover for the latter’s sacrifice.58 They are even more explicitly addressed (captiuae), perhaps with Hecuba and Andromache subsumed within their numbers, in the play’s final lines, as the second Messenger hurries them to the shore to set sail for their new, Greek homes.59 However, their most startling and radical role is in the kommos with Hecuba that follows the spoken prologue. This responsive chant between the Trojan women and their former queen is in effect a dialogue, almost unique in Senecan tragedy.60 It clearly establishes the Chorus as individuated characters within the world of the play. They are not an extradiegetic, detached equivalent of film musicians, in Hill’s attractive analogy.61 Rather, “the chorus of Trojan women have a character, which is evident visually and in their words”; indeed, they are “the most individualised chorus of any Senecan play”.62 If any Senecan Chorus can be imagined as engaging in a substantial dialogue with one of the protagonists, it is surely that of Troades.

5 Keeping the Chorus Onstage

Two scholars, Davis and Marshall, have made the case for the absence of the Chorus for most of act three and built further interpretations on that absence. Some of these arguments have a degree of circularity (the Chorus must be absent because of the dramatic effect of that absence), but they do offer telling readings of the act and need to be addressed. Davis shows his characteristic sensitivity to Seneca’s theatricality and asserts that “it is dramatically more effective to have them absent during the agon between Andromache and Ulysses. The audience’s attention should focus solely upon the mother’s anguish and her tormentor’s wiles.”63 This is of course a subjective judgment, but no more so than any other assessment of dramatic effectiveness and no reason to avoid that important dimension of a play altogether. However, there are two more objective criteria that make it problematic. The first, with the usual caveats about comparing Attic and Senecan stagecraft, is the fact that many of the most dramatic, focused, and even intimate confrontations between two individuals in fifth-century tragedy take place in the presence of the Chorus. Among many examples, a particularly pertinent parallel can be found in the two scenes in which Euripides’ eponymous Andromache confronts first Hermione and then Menelaus. These vividly allow the audience to focus on a mother’s anguish and successive tormentors’ wiles even in the presence of the Chorus of Phthian women.64 More concretely still, at three different points in the course of act three of Troades, Ulysses gives orders to the Greek soldiers who evidently accompany him, so that, even without the Chorus, he and Andromache are not alone onstage.65 If, despite all this, it is still considered desirable to remove the Chorus before the agon, they can just as easily (and silently) exit at Ulysses’ entrance, when most editors have the Senex leave, as they can at the point when most critics dismiss them, at the phantom Senex’s apparent entrance. Therefore, this is no objection to their being Andromache’s interlocutor in the first part of the act.66

Marshall also requires the Chorus to be absent for act three, after Andromache’s initial address to them. This is part of his sophisticated and important argument that the Chorus’ various presences, absences, entrances, and exits mark the various transitions of location between Hector’s tomb, the Greek camp, and the former battlefield.67 Marshall has the Chorus exit at line 425 stage right, which is his confessedly arbitrary assignment of the eisodos (or Roman equivalent) signifying ‘seawards’ as opposed to ‘Troywards’.68 When they re-enter at the end of the act, after Ulysses has dragged Astyanax away stage left (‘Troywards’) and a brief moment when the stage is empty to ‘re-set’ the location, they do so also from stage left, having performed a ‘backstage run-around’. By doing so, they indicate that they have continued to move in a ‘seawards’ direction and hence that the acting space now represents a location that is no longer in front of Hector’s tomb but closer to the sea, specifically the former battlefield near Sigeum. Marshall’s argument is extremely attractive and makes very important and suggestive points about Senecan stagecraft. However, it does have several points of vulnerability and, even if it were accepted as a whole, it does not in itself preclude the Chorus’ being Andromache’s interlocutor in act three.

The notion that the stage’s momentary emptiness is sufficient to cancel a location is defensible but open to dispute. It could offer particular challenges to the audience to understand that the same eisodos designates a different path (albeit with the same orientation) in the seconds between Ulysses’ exit and the Chorus’ entrance. It is also unclear how the entrance into the skene-equivalent, which has signified Hector’s tomb and as such been the focal point of act three, ceases to do so. Marshall’s further claim that each location has its own “iconic character”, Calchas for the Greek camp, Astyanax for Hector’s tomb, Polyxena for the battlefield, is also attractive but is even more susceptible to objections that he either does not note or answers with special pleading. Although “[w]hen these characters leave, the space which they define collapses”, this is apparently not the case when Astyanax leaves to enter Hector’s tomb, since “[t]hough he is no longer visible to the audience, Astyanax continues to represent the location”, nor when Polyxena leaves (by an eisodos and permanently) at the end of act four: “The removal of Polyxena should close the location, but it does not.”69 Nor is it clear how Astyanax’s ‘iconic’ status for the Hector’s tomb location can be reconciled with his absence in act one and the opening of act two, which Marshall locates there. Marshall also has Andromache exit ‘seawards’ at the end of act three, ready to enter from stage left in act four. The justification for this as a ‘psychologically true response’, for a mother to run away in terror rather than follow her doomed son, is at best too debatable to offer compelling evidence for stage movement.70

If the Chorus is not used in this way to indicate a change of location, then a reason for and consequence of their absence in act three is removed. However, despite these objections, Marshall’s overall argument remains defensible and it is worth pointing out that it does not in fact require the Chorus to exit at 425. He does not actually make an argument for their exit but takes it for granted (note ‘recognized’), declaring that “[m]ost scholars … have recognized that the chorus should not stay on stage throughout act 3 … Their departure is typically marked at line 425, and it is ‘Before 435, surely’.”71 No reason is given. If the Chorus must have exited before the end of the act so as to re-enter from the other side, there is no reason that they could not do so at 518, perhaps frightened by the entry of Ulysses and his men that they have just announced. Davis’ desire that the stage be empty for Andromache and Ulysses’ agon and Marshall’s that the Chorus exit so they can re-enter and change the location can both be satisfied if the Chorus is the interlocutor in lines 426-518 and exits immediately or soon after their final speech at 517-518.

6 Scribal Intervention and the Invention of the Senex

The only evidence for the existence of the Senex is the scene headings and speaker attributions in the manuscripts, replicated in all editions. If he does not in fact exist and the lines attributed to him belong to the Chorus, why do the manuscripts include him? The source and antiquity of these headings and attributions is far from clear and there is no good reason to think that they go back to Seneca.72 Certainly the tragedies show frequent misattributions of dialogue to the wrong character.73 These can sometimes be assigned to purely accidental scribal error but are often evidence of a scribe’s deliberate editorial intervention. This latter phenomenon is particularly manifested in the frequent cases where scribes take characters’ references to themselves by name in the third person as a sign that someone else must be speaking. Thus, the E manuscripts assign Thy. 920-937 to the Chorus because, as Fitch notes, of the word Thyesten at 937.74 Though monodies are not especially rare in Senecan tragedy, the unexpected anapaests may also have influenced the scribe’s decision.75 It should be noted that this is probably not merely a case of making the wrong choice from among the available characters onstage. Since scholars generally believe that the Chorus was absent from the fifth act of Thyestes, the scribe is conjuring an absent speaker because it felt more appropriate to assign lines to them.76

A similar situation arises in act four of Hercules furens, though there, as (I argue) in Troades act three, the manuscripts attribute to an absent character lines that belong to the Chorus. Lines 1032-1034, urging Amphitryon not to commit virtual suicide by exposing himself to the homicidal Hercules, are assigned by the manuscripts to Theseus, the only lines so attributed in act four. Critics have noted the improbability that Theseus should be onstage for the whole of Hercules’ murderous rampage but only intervene at this late point and in such a minor way.77 They also generally agree in assigning the lines to the Chorus. It seems overwhelmingly probable that a scribe has responded either to the absence of any attribution of the lines or to their unexpected assignment to the Chorus by desperately trying to find a character available to deliver them. Lycus and Megara are dead, Juno unthinkable, Hercules insane, inappropriate, and offstage. Amphitryon is just possible, though the volte-face between urging his son to kill him and urging himself to elude being so killed would be extremely jarring, and we have already seen scribes’ discomfort with self-address, so that they would tend to assume (in this case correctly) that someone else is addressing him as senior (1032). For a scribe convinced that the Chorus cannot be involved in dialogue and determined to normalize any evidence that it is, Theseus is the only alternative. A similar discomfort with the Chorus’ unusual involvement in dialogue with Andromache could have led to the attribution of their lines to the phantom Senex.

There is, of course, an important difference between the two cases. Theseus may not be onstage (physically or notionally) in act four of Hercules furens, but he is unquestionably a character in the tragedy and one who, quite apart from other clear contextual indications, is repeatedly addressed or referred to as present by name.78 The Senex in Troades would have to be invented from whole cloth by a scribe and this is undeniably a more radical suggestion. A partial parallel can be found in Troades itself, at the beginning of act two. The Messenger who there enters and, in dialogue with the Chorus (whether of Trojan women or Greek soldiers is another issue on which manuscripts and scholars disagree), describes the appearance of Achilles’ ghost demanding the sacrifice of Polyxena, is designated Talthybius by the manuscripts of both the A and E branches. With one exception, scholars have accepted this designation. In contrast, manuscripts and scholars call the character who narrates the deaths of Polyxena and Astyanax in act five simply Nuntius. Friedrich alone dissents and says that there is as much reason for the second Messenger to be called Talthybius as the first and that, in accordance with standard Senecan practice in identifying named individuals, it is preferable to designate them both as simply Nuntius.79 The name Talthybius is mentioned neither by the character himself nor by the Chorus. There is no individualizing detail that could be used to identify him as Talthybius (or anyone more individual than a generic Messenger) and the only justification for doing so is his general mythological role as Agamemnon’s herald and his specific tragic role in Euripides’ Troades and Hecuba. Yet these are precisely the sort of extratextual factors that could influence an officious scribe to decide that this must be Talthybius and alter the scene-headings and line-attributions accordingly. In contrast, the only securely named conventional Messenger in Senecan tragedy, Eurybates in Agamemnon, is clearly and explicitly introduced by the Chorus, while Creon, a major character fulfilling the role of Messenger, is explicitly dispatched to watch the necromancy (399-400) and in the fractious lead-up to his speech, Oedipus invokes his ‘kindred house’.80 It is true that neither Pyrrhus nor Agamemnon is explicitly introduced in the agon that immediately follows the report of Achilles’ ghost. However, each’s identity is strongly suggested in the dialogue at an early stage (quite apart from possible costumes) and soon made absolutely unambiguous when Pyrrhus refers to Achilles as his father (232, 235) and Agamemnon refers to him by name (252). Agamemnon’s identity is made clear by Pyrrhus’ allusion to the sacrifice of Iphigenia (248-249) and threat to kill him as he had his equal, Priam (307-310). Nothing in the brief scene that opens the act suggests that the speaker is anything other than a generic Messenger. Nevertheless, a scribe has introduced into act two the character of Talthybius, who has no place in it, just as I suggest someone introduced the phantom Senex into act three.

It must be conceded that none of these parallels corresponds in every detail to what I propose must have happened for a scribe to invent the Senex and attribute to him the Chorus’ lines. A scribe invented Talthybius, who is not a character in the play, but there was unquestionably a Messenger who reported the appearance of Achilles’ ghost and it was this existing character that the scribe designated Talthybius. A scribe assigned the Chorus’ lines in act four of Hercules furens to Theseus, who did not speak them and was not even onstage, but who is a character in the play. There is no exact parallel or precedent for the invention of a character from whole cloth and the misattribution to them of the Chorus’ lines. However, the processes by which such a combination of these events that are separately paralleled could have come about are perfectly plausible. We have noted how unusual the Chorus’ role in this act would be and how this could encourage a conservative scribe to normalize its dramaturgical practice by assigning the lines to an individual character. The dramatis personae include no obvious alternative such as Theseus was in Hercules furens. Ulysses is referred to by the speaker, Pyrrhus, Agamemnon, and Helen would be wildly inappropriate as conspirators to save Astyanax, and Hecuba would be too prominent (and emotional) a personality to serve as little more than a passive sounding-board for Andromache. Moreover, the tone of the dialogue and its superficial resemblance to various Nurse- and Satelles-scenes would encourage the scribe to imagine that Andromache is speaking to a low-status (which is the case) individual (which is not). The sort of mind that can infer that a Greek Messenger at Troy must be Talthybius could easily infer that a low-status individual in the presence of a young boy is a paedagogus. All of this is predicated on the more interventionist scenario where a scribe goes out of his way to alter existing line attributions, Messenger to Talthybius and Chorus to Senex. If—as is equally and perhaps more likely—changes of speaker were indicated merely by a paragraphos or similar marginal sign, then the scribe would not even have needed to emend, but merely to gloss what he assumed or inferred to be the most probable speaker for each speech. That a scribe should conjure the phantom Senex in these circumstances is neither improbable nor even surprising.

7 Conclusion

It is undeniably a radical suggestion that the character designated Senex by all the surviving manuscripts of Seneca’s Troades does not exist but was invented by a scribe who either rejected the attribution of so much dialogue to the Chorus or assumed the speaker marked only by a paragraphos must be a slave and hence most probably Astyanax’s paedagogus. It also involves an unusual level of involvement for the Chorus in dialogue. However, all of these apparent anomalies have parallels in the Senecan corpus and its transmission, and it is only their combination that is in any way unusual. On the other side, the existence of the Senex produces multiple, serious problems with what can be empirically reconstructed as the usual practice of Senecan stagecraft, his handling of anonymous characters, and even the most natural interpretation of the text itself. All of these can be solved if the phantom Senex is finally exorcized from outside Hector’s tomb and the Chorus permitted to take its proper place in conversation with Andromache.81


  • Billerbeck, M., ed. (1999). Seneca, Hercules Furens. Leiden.

  • Boyle, A.J., ed. (1994). Seneca’s Troades. Leeds.

  • Boyle, A.J., ed. (2008). Octavia. Attributed to Seneca. Oxford.

  • Boyle, A.J., ed. (2011). Seneca, Oedipus. Oxford.

  • Boyle, A.J., ed. (2014). Seneca, Medea. Oxford.

  • Boyle, A.J., ed. (2017). Seneca, Thyestes. Oxford.

  • Boyle, A.J., ed. (2019). Seneca, Agamemnon. Oxford.

  • Calder, W.M. (1975). The Size of the Chorus in Seneca’s Agamemnon. CP 70, pp. 32-35.

  • Caviglia, F., ed. (1981). Lucio Anneo Seneca, Le Troiane. Roma.

  • Chancellor, G. (1979). Implicit Stage Directions in Ancient Greek Drama. Critical Assumptions and the Reading Public. Arethusa 12, pp. 133-152.

    • Search Google Scholar
    • Export Citation
  • Costa, C.D.N., ed. (1973). Seneca, Medea. Oxford.

  • Davis, P.J. (1993). Shifting Song. The Chorus in Seneca’s Tragedies. Hildesheim.

  • DuBois, P. (2003). Slaves and Other Objects. Chicago.

  • Fantham, E. (1982). Seneca’s Troades. A Literary Introduction with Text, Translation and Commentary. Princeton.

  • Fitch, J.G., ed. (1987). Seneca, Hercules Furens. Ithaca.

  • Fitch, J.G. (2004). Annaeana Tragica. Notes on the Text of Seneca’s Tragedies. Leiden.

  • Friedrich, W.-H. (1933). Untersuchungen zu Senecas dramatischer Technik. Borna/Leipzig.

  • Geue, T. (2018). Soft Hands, Hard Power. Sponging Off the Empire of Leisure (Virgil, Georgics 4). JRS 108, pp. 115-140.

  • Hamilton, R. (1987). Cries Within and the Tragic Skene. AJPh 108, pp. 585-599.

  • Harrison, G.W.M., ed. (2000). Seneca in Performance. Swansea.

  • Haussker, F. (2019). The Child’s Voice in Euripidean Tragedy. Socialization through Challenge. Arethusa 52, pp. 203-229.

  • Hermann, L. (1924). Le théâtre de Sénèque. Paris.

  • Hill, D.E. (2000). Seneca’s Choruses. Mnemosyne 53, pp. 561-587.

  • Keulen, A.J., ed. (2001). L. Annaeus Seneca, Troades. Leiden.

  • Kohn, T.D. (2004-2005). Seneca’s Use of Four Speaking Actors. CJ100, pp. 163-175.

  • Kohn, T.D. (2013). The Dramaturgy of Senecan Tragedy. Ann Arbor.

  • Leo, F. (1878). L. Annaei Senecae Tragoediae, Vol. 1, observationes criticas continens. Berlin.

  • Marshall, C.W. (2000). Location! Location! Location! Choral Absence and Dramatic Space in Seneca’s Troades. In: G.W.M. Harrison, ed., pp. 27-51.

    • Search Google Scholar
    • Export Citation
  • Marshall, C.W. (2014). The Structure and Performance of Euripides’ Helen. Cambridge.

  • Marx, W. (1932). Funktion und Form der Chorlieder in den Seneca-Tragödien. Köln.

  • McAuley, M. (2015). Reproducing Rome. Motherhood in Virgil, Ovid, Seneca, and Statius. Oxford.

  • Ormand, K. (1996). Silent by Convention? Sophocles’ Tekmessa. AJPh 117, pp. 37-64.

  • Sabnis, S. (2012). Invisible Slaves, Visible Lamps . A Metaphor in Apuleius. Arethusa 54, pp. 79-108.

  • Schiesaro, A. (2003). The Passions in Play. Thyestes and the Dynamics of Senecan Drama. Cambridge.

  • Slaney, H. (2013). Seneca’s Chorus of One. In: J. Billings, F. Budelmann, and F. Macintosh, eds., Choruses, Ancient and Modern, Oxford, pp. 99-116.

    • Search Google Scholar
    • Export Citation
  • Sutton, D.F. (1986). Seneca on the Stage. Leiden.

  • Taplin, O. (1977). The Stagecraft of Aeschylus. The Dramatic Use of Exits and Entrances in Greek Tragedy. Oxford.

  • Tarrant, R.J. (1978). Senecan Drama and its Antecedents. HSPh 82, pp. 213-263.

  • Tarrant, R.J., ed. (1985). Seneca, Thyestes. Atlanta.

  • Volk, K. (2000). Putting Andromacha on Stage. A Performer’s Perspective. In: G.W.M. Harrison, ed., pp. 197-208.

  • Wilson, M. (1983). The Tragic Mode of Seneca’s Troades. Ramus 12, pp. 27-60.

  • Yoon, F. (2012). The Use of Anonymous Characters in Greek Tragedy. Leiden/Boston.

  • Zanobi, A. (2014), Seneca’s Tragedies and the Aesthetics of Pantomime. London.

  • Zissos, P.A. (2008). Shades of Virgil. Seneca’s Troades. MD 61, pp. 189-210.

  • Zwierlein, O. (1966). Die Rezitationsdramen Senecas. Meisenheim am Glan.

  • Zwierlein, O. (1983). Prolegomena zu einer kritischen Ausgabe der Tragödien Senecas. Wiesbaden.


Wilson 1983, 56: “the divided dramatic interest and consequent fragmentation of audience sympathy, the illusion of discontinuity, the degree of reliance upon juxtapositions, parallels and contrasts, the length and organization of the extraordinary third act”; Boyle 1994, 176: “the longest act in Senecan tragedy and structurally its most complex”.


Yoon 2012, 153: “anonymous characters in later tragedy are for the most part inherited from particular Greek models rather than invented. … in the seven Senecan tragedies for which there is an extant Greek precedent, he invents only the Old Man in Troades.” Yoon presumably identifies Clytemnestra’s Nurse in Seneca’s Agamemnon with Orestes’ Nurse Cilissa in Aeschylus’ Choephori since none appears in the latter’s Agamemnon. No Attic Thyestes play is extant, so the Satelles’ ancestry is uncertain. Fantham 1982, 74, parenthetically designates the Senex “an innovation, unless he has crept in from another Astyanax drama”.


On Senecan stagecraft, see Sutton 1986; Kohn 2013, though both tend to adopt an extreme position and should be treated with caution. The chapters in Harrison 2000, many focused on Tro., offer a more balanced perspective.


Tarrant 1978. Cf. Boyle 1994, 157; Keulen 2001, 28.


For convenience and brevity, I shall often refer to ‘the Senex’ to designate Andromache’s interlocutor, but this does not of course mean that I believe that such a character exists.


Zwierlein 1966, 53.


Fantham 1982, 278.


The issue of who the audience would expect to respond to Andromache is discussed in section 3 below.


One of Mnemosyne’s anonymous reviewers points out that the E manuscripts head the section of the act beginning at 524 with the names VLIXES ANDROMACHA. This may reflect the Senex’s exit or merely his lack of further speeches. It is also interesting that they do not include Astyanax among the characters, though they did list him alongside Andromache and the Senex ante 409. The A manuscripts list ANDROMACHA, SENEX, VLIXES, ASTIANAX ante 409, but do not mark a new scene with distinct characters ante 524.


Oed. 288-403 and Ag. 867-1012, with Kohn 2004-2005. On Seneca’s general observance of the three-actor rule: Sutton 1986, 28-32; Kohn 2013, 21-22.


E. Med. 1270a, 1271-1272, 1277-1278; S. El. 1404-1405, 1409-1411, 1415-1416. The standard study remains Hamilton 1987.


On these scenes, see Haussker 2019.


On Tecmessa: Ormand 1996. Again, this does not presuppose continuity between Attic and Senecan stagecraft but indicates an ancient parallel.


Entrances at E. Hec. 658 and El. 761. Euripides self-consciously draws attention to the latter switch by having the messenger complain that Electra does not recognise him as her brother’s attendant (766). I am indebted to one of Mnemosyne’s anonymous reviewers for reminding me of the attendant in Hecabe.


One of Mnemosyne’s anonymous reviewers suggests that ‘dubbing’ would be less controversial if Astyanax were an infant or toddler, but this seems inconsistent with his general depiction and especially with his voluntary leap from the tower, as narrated at 1100-1103.


Sutton 1986, 48.


Fantham 1982, 273.


On implicit stage directions in Attic tragedy, see esp. Chancellor 1979; Taplin 1977, 28-39; Marshall 2014, 188-233. This is not to say they are entirely absent from Senecan tragedy, and see esp. Sutton 1986, 43-56; Kohn 2013, 2-3.


The issue does not arise in Her. f., which features no anonymous characters, or Phoen., which does have an anonymous nuntius (320-327, 347-349) and satelles (387-402, 419, 427-442), but no Chorus.


nutrix: Phaed. 178 (Phaedra), Med. 568; altrix: Phaed. 251 (Phaedra), 358 (Chorus); alumna: Phaed. 255, 588, Med. 158, 380.


alumna: Oct. 72, 254 (of Octavia), 691 (of Poppaea), Her. O. 276, 445, 539; nutrix: Oct. 75 (Octavia), 713 (Poppaea); altrix: Her. O. 396, 402, 450, 491.


Thy. 623-640, 691, 716-748.


Med. 879-890. The Nurse addresses Medea immediately after the end of the exchange but it is unclear when the two women enter. Boyle 2014, 346 suggests: “The fact that the Messenger addresses the Chorus and is addressed by them seems to indicate … that there is no one else onstage”.


Phaed. 989-990: sed quid citato nuntius portat gradu | rigatque maestis lugubrem uultum genis?; Oed. 911-914: sed quid hoc? postes sonant. | maestus et famulus manu | regius quassat caput. | ede quid portes noui; Ag. 388-391: sed ecce, uasto concitus miles gradu | manifesta properat signa laetitiae ferens | (namque hasta summo lauream ferro gerit) | fidusque regi semper Eurybates adest (the primary Chorus of Argive women speaking, not the secondary one of Trojan prisoners).


Tro. 166-167: quae causa ratibus faciat et Danais moram, | effare, reduces quis deus claudat uias; 1178-1179: repetite celeri maria, captiuae, gradu: | iam uela puppis laxat et classis mouet.


It should be emphasized that I designate the Senex ‘phantom’ in the sense that he has an illusory existence in the play’s reception (its only adjectival sense recorded in OED). There is no connection with the actual ghosts of Achilles and Hector whose ‘real’ appearances are reported by the first Messenger and Andromache.


Schiesaro 2003, 156: “Atreus’ impassioned speech … has an amazing effect on his advisor, who immediately erases any remaining difference between himself and his master: the trap devised for Thyestes is now ‘ours’”; Boyle 2017, 167: “the Courtier’s movement from moral resistance … to compliance …—and sometimes strong compliance …—seems designed to mirror that of the Ghost of Tantalus in the prologue and to anticipate the analogous movement of Thyestes in Act III”.


loqueris 246, erras 306, times 313, uocas 314, credis 315, nostra tu coepta occule 333.


putes 717, putas 746, exhorruistis 744.


On the size of the Senecan chorus, see esp. Calder 1975. The radical suggestion of Slaney 2013 that the plays were performed by a solo singer (the ‘chorus of one’) and a solo pantomimic dancer is intriguing but would still conjure the image of multiple chorus members in the theatre of the audience’s mind.


Davis 1993, 17-18.


Boyle 2019: “Nowhere in this scene does the Nurse address Clytemnestra in the intimate way of other Senecan nurses, i.e. by addressing her as alumna or era. Clytemnestra’s high birth and regal status are constantly on display in this act as defining constituents of her identity”.


Sen. Ag. 147-148. The text throughout is Zwierlein’s OCT and all translations are my own.


Zanobi 2014, 81.


Sen. Tro. 503-508.


Caviglia 1981, 260: “dolorosa ironia”; Fantham 1982, 289; Boyle 1994, 184; Keulen 2001, 330; Ov. Met. 1.355.


On invisible slaves, see DuBois 2003; Sabnis 2012; Geue 2018.


On Andromache and Hector: Volk 2000, 198-205; Zissos 2008, 199-209; Schiesaro 2003, 194-200; but contrast McAuley 2015, 266-272.


Phaed. 81-177; Med. 116-154; Ag. 108-130; Thy. 176-205 (all references covering only the soliloquy and the interlocutor’s first speech). As noted, the Nurse in Ag. cannot hear the soliloquy and responds to what she rightly assumes Clytemnestra is thinking. The same structural pattern is present with Amphitryon as the interlocutor urging restraint to Hercules at Her. f. 1138-1321. His earlier scene with Megara opens with a long monologue of his own (205-278) but Megara’s speech begins abruptly as if he had not spoken (279-308) and the exchange then proceeds in the usual fashion.


Sen. Tro. 409-426.


Ag. 108 (also later in the scene at 192; Aegisthus addresses his animus in the second half of the scene where his opening soliloquy introduces a passion-passion exchange); Phaed. 112; Thy. 192 (even while in dialogue with the Satelles, Atreus turns to address his animus on three further occasions, at 270, 283, and 324). Medea addresses her animus in act one (41) and repeatedly in act five (805, 937, 976, 988), but not in the act two soliloquy preceding the passion-restraint scene with the Nurse.


Tro. 613 (Ulysses), 662 (Andromache).


Phaed. 130: nefanda casto pectore exturba ocius; Med. 150-151: sile, obsecro, questusque secreto abditos | manda dolori.


I am indebted to Patrick Ryan for pointing out the significance of Andromache’s address to the Chorus.


Zwierlein 1966, 53: “In 409 ff. hatte sich Andromache an die maesta Phrygiae turba, also an den Chor gewandt. Doch in 426 antwortet ihr statt des Chores ein Greis, der plötzlich vor ihr steht.” Cf. Keulen 2001, 299: “Contrary to expectation, not the women, addressed by her (409), but an unspecified senex now speaks to Andromache”.


Tro. 488-495, following Leo’s generally accepted rearrangement of the manuscript’s confused ordering.


I am indebted to one of Mnemosyne’s anonymous reviewers for challenging me on the issue of the interlocutor’s social status. It might also be added that the first passion-restraint scene in Her. f. takes place between the social equals Megara and Amphitryon.


A. Th. 245-285, 652-720; S. Aj. 201-347; E. Alc. 476-508.


On the Senecan chorus, see esp. Calder 1975; Davis 1993; Hill 2000; Slaney 2013.


Ag. 710-719, 775-781; Phaed. 824-834; Oed. 1004-1009, 1040-1041.


Andromache (Tro. 945-950), the Nurses of Phaedra (Phaed. 583-586) and Medea (Med. 382-396), Amphitryon (Her. f. 1042-1048), and Atreus (Thy. 909-919). For the possible influence of pantomime on this narration of mimed action, see Zanobi 2014, 89-127.


Her. f. 1032-1034; Phaed. 404-405; 1244-1246.


Med. 879-890; Tro. 166-167.


Thy. 623-788.


The rule was formulated by Marx 1932, 5, and invoked by Zwierlein 1966, 91, 102-103; Costa 1973; 150 (though he imagines the Nurse and Medea overhearing ‘in the background’); Fantham 1982, 232; Billerbeck 1999, 531; Keulen 2001, 165; Boyle 2008, 255; id. 2011, 317-318; id. 2014, 346; id. 2017, 318; but note the exceptions mentioned by Fitch 1987, 385-386 n. 134.


A. Ch. 900-902. Boyle 1994, 200 also compares Pylades, as a hitherto silent character, but without consideration of which actor speaks Astyanax’s line.


Davis 1993, 220. Cf. Kohn 2013, 112: “The identity of the Chorus is unusually clear, just as remarkable as its level of participation in the play itself”.


Tro. 901-902: celebrate Pyrrhi, Troades, conubia, | celebrate digne: planctus et gemitus sonet. Pace Davis 1993, 20, who claims that “they are presumably responsible for combing Polyxena’s hair”, the combination of Polyxena’s allowing her hair to be handled with her seeking of the beautiful ornamentation of a royal robe (cultus decoros regiae uestis petit | et admoueri crinibus patitur manum, 946-947) suggests that Helen has brought a retinue of handmaids, since the captive Trojans would hardly be able to provide a royal robe.


1178-1179: repetite celeri maria, captiuae, gradu: | iam uela puppis laxat et classis mouet.


Keulen 2001, 126: “Such active participation of the chorus in the dialogue is … rare in these tragedies”. The only other example also features a strongly characterized Chorus of Trojan women, this time in lyric dialogue with Cassandra: Ag. 589-781.


Hill 2000, 587. He does concede that “[t]hey can be involved as characters in the play”, but curiously Troades itself is one of his test-cases for the detachment of the Chorus.


Marshall 2000, 27; Boyle 1994, 33.


Davis 1993, 26-27. Cf., more arbitrarily, Fantham 1982, 39: “Andromache, the old man, and the child need to be alone by Hector’s tomb in act 3”.


E. Andr. 147-272; 309-463.


Tro. 627-629: ite, ite celeres, fraude materna abditum | hostem, Pelasgi nominis pestem ultimam, | ubicumque latitat, erutam in medium date; 678-680: cessatis et uos flebilis clamor mouet | furorque cassus feminae? iussa ocius | peragite (Andromache responds me, me sternite hic ferro prius, 680); 813: abripite propere classis Argolicae moram. These plurals make it hard to agree with Marshall 2000, 33: “even one man can be a full bodyguard for Ulysses in act 3”.


Davis 1993, 26 also cites the probable change of location before act four and the alleged inappropriateness of the third choral ode as a response to the action of the act. The former is more persuasive than the latter, but both can likewise be solved by the Chorus’ exiting at or soon after 518.


Marshall 2000.


It is arbitrary in the sense that stage left could just as easily be ‘seawards’ and stage right ‘Troywards’, not that the main argument is not carefully argued.


Marshall 2000, quoting from 41, 42, and 44.


Marshall 2000, 43.


Marshall 2000, 38-39, citing Hermann 1924, 372, and Fantham 1982, 38.


Leo 1878, 83: “quaerendum autem ante omnia est quatenus personarum notae scaenarumque tituli quales nunc leguntur antiquitus traditi sint”. The subsequent discussion (82-88) remains fundamental. Cf. the interrelated issue of scene titles as discussed by Zwierlein 1983, 249 who asserts that “die Szenentitel … nicht auf Seneca selbst zurückgehen, daß aber ein Grundstock von ihnen bereits im Archetypus wurzelt…. Diese Szenentitel des Archetypus wurden dann … bei der Spaltung der Überlieferung und in den folgenden Traditionsstufen von den verschiedenen Bearbeitern oder Kopisten in unterschiedlicher Weise vervollständigt oder verändert—je nach Maßgabe der oftmals divergierenden Entscheidungen über die Sprecherverteilung im Szeneninnern”.


Fitch 2004, 76: “The MSS are often unreliable in their speaker-attributions”.


Fitch 2004, 130.


Other monodies at Ag. 759-774, Phaed. 1-24, 1201-1212 (trochaic tetrameters, chanted), Med. 740-842, Tro. 705-735, as well as several in Oct. and Her. O.


Chorus’ absence: Davis 1993, 36; Boyle 2017, 389, but note contra Kohn 2013, 129, who asserts without argument that “the Chorus returns to the scaenae frons” at the end of the preceding ode. Tarrant 1985 does not express an opinion.


Billerbeck 1999, 530-531: “Es wäre in der Tat psychologisch befremdlich, wenn Theseus die ganze Mordszene stumm und reglos mitverfolgte und dann, nachdem Hercules die Seinen hingeschlachtet hat, sich endlich zu Wort meldete, um den alten Amphitryon zur Flucht zu bewegen”. Fitch 1987, 385 merely notes his exit 914-917 and the improbability of his return solely to deliver these “colorless lines”.


Her. f. 637, 654, 914, 1173, 1177 bis, 1242, 1318, 1335.


Friedrich 1933, 103 n. 1: “Mit dem gleichen Rechte könnte der Bote des 5. Akte (1056 ff.) Talthybius heißen, und um der Konsequenz willen bleibe er in den modernen Ausgaben auch hier im zweiten Akt besser anonym. Seneca hat den Boten beide Male als namenloses Instrument behandelt wie die Boten der anderen Stücke auch. Zu einer individuellen Persönlichkeit gehört nach Senecas sonstiger Gewohnheit, daß der Name des Neuauftretenden vorher vom Chor oder den sonst Anwesenden genannt werde, oder sich kurz nach dem Auftreten sonstwie herausstelle”.


Eurybates: Ag. 388-391; Creon’s dispatch: Oed. 399-400; cognatae domus: 513; Boyle 2011, 238 ad Oed. 530-658: “the only [messenger speech] to be delivered by a major character”.


The idea for this article grew out of teaching a Roman Drama unit at the University of Sydney (largely on Zoom) in 2020. I am grateful to the whole class for producing such a stimulating environment under difficult circumstances, but especially Patrick Ryan, who made the key observation about Andromache’s address to the Chorus and later made helpful comments on a written draft, as did Paul Roche. The article was researched and written during a period of leave granted by the University of Sydney Faculty of Arts and Social Sciences Special Studies Programme in 2021, for which I would also like to express my gratitude. I am also very grateful to Mnemosyne’s two anonymous reviewers for their insightful and helpful comments and suggestions.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 27 27 0
Full Text Views 239 239 27
PDF Views & Downloads 507 507 54