Bryn Mawr Classical Review

Bryn Mawr Classical Review 2005.07.53

Mario Erasmo, Roman Tragedy: Theatre to Theatricality.   Austin:  University of Texas Press, 2004.  Pp. xii, 211.  ISBN 0-292-70242-6.  $45.00.  

Reviewed by Bob Cowan, Brasenose College, Oxford (
Word count: 3661 words

Roman tragedy of the Republican period has been well served by editors but less so by literary critics. Seneca has fared better and the last twenty-five years in particular have seen a massive improvement in his reputation and, accordingly, the quantity and quality of scholarship on his tragedies. The fragments of Livius, Naevius, Ennius, Pacuvius and Accius, however, are still often read as just that, fragments of only linguistic curiosity, which in many cases led to their preservation by grammarians in the first place. Some good recent work has been produced on these texts,1 but the genre still awaits a strong, general discussion of these tragedies as literature and especially as theatre, analogous to the service which Goldberg has performed for the fragments of early Roman epic.2 The ambitious undertaking of Erasmo (E.) is to survey Roman tragedy from its beginnings with Livius Andronicus to Seneca and, beyond him, Curiatius Maternus and the author of the Octavia. His interesting and potentially exciting perspective on the genre is its development from theatre to metatheatre, as the distinction between onstage and offstage reality is blurred so that the 'audience's reality' becomes theatricalised and, in turn, tragedy becomes self-conscious of its own theatricality and, as a result, turns into metatragedy. It is all the more disappointing, therefore, that this fascinating thesis is explored using weak and superficial argumentation based on overstretched and misinterpreted evidence, further marred by an inexcusable number of errors.

The book is structured chronologically and, after a general introduction, there are chapters on Livius, Naevius and Ennius (ch.1 'Creating Tragedy') and on Pacuvius and Ennius (ch.2 'Theatricalizing Tragedy'). Ch.3, 'Dramatizing History' discusses fabulae praetextae, plays on Roman historical themes, from Naevius to the Pseudo-Senecan Octavia, with particular focus on Accius' Aeneadae sive Decius. Ch.4, 'Creating Metatragedy', focuses on the significance of re-stagings of tragedies and praetextae in the late Republic and early Principate, notably that of Accius' Clytemnestra at the opening of Pompey's theatre in 55 BCE, Accius' Tereus (in lieu of his Brutus) at Brutus' ludi Apollinares in 44 BCE, and Varius' Thyestes at Octavian's triple triumph in 29 BCE. The last of these, part of a survey of Roman tragedies based on the story of Thyestes, provides a transition into a discussion of Seneca and the theatrical nature of Nero's principate. The final chapter, 'Metatragedy', is devoted entirely to Seneca and the replacement of theatre with theatricality.

Within this chronological framework, E. develops his argument as follows: Livius Andronicus in 240 BCE introduces theatre to a Rome which had no conception of theatre and thus of theatricality, for which, he claims, the former is a pre-requisite. As time passes, later tragedians become increasingly self-conscious and the barrier between the competing realities of the stage and the audience breaks down, so that 'illusory' gives way to 'nonillusory' drama. (E. uses the term 'nonillusory' rather broadly to mean drama with any connection to offstage reality.) This blurring of boundaries is brought about by two main factors. Firstly, the 'rhetoricization' of drama, E. claims, coupled with the theatricalisation of rhetoric, leads to a confusion of offstage and onstage reality.3 Secondly, boundaries are blurred by allusions to the audience's reality, be it contemporary events, personalities or more general aspects of Roman life. Tragedies proper, i.e. those with mythological subject-matter, can become 'nonillusory' through authorial allusion, audience reception, or through gesture, staging or even interpolation by actors and producers. Examples are discussed further below but include Ennius' reference to plebs in the Iphigenia, Pompey's restaging of Accius' Clytemnestra at the opening of his theatre, and the actor Aesopus' use of lines from and interpolation of lines into Accius' Eurysaces to allude to Cicero's recall from exile. With praetextae, there is the possibility of identification between contemporary figures and their (often homonymous) ancestors represented onstage, as when D. Junius Brutus Callaicus commissioned Accius' Brutus, or even, in cases such as that of Ennius' Ambracia, a situation where someone like M. Fulvius Nobilior might be in the audience and also represented onstage. This process takes on further complexities when tragedies and praetextae are restaged in such a way as to gain further significance through comparison with their original performance context. Just as outside reality enters the world of the stage, so outside reality becomes theatricalised and begins to conceive of itself in terms of drama. Evidence for this is drawn from later oratory and historiography. This process comes full-circle as the theatricalisation of reality reinvades the stage and tragedies, especially Seneca's, represent theatricalised reality onstage, making characters the audience and the audience witnesses, as when Atreus watches Thyestes eating his children.4

Paraphrased thus, E.'s argument is attractive, suggestive and even, at times, persuasive, but it still possesses major flaws. The book does contain many good ideas and discussions. The role of masked actors representing ancestors in funeral processions and the praetextae which perhaps followed (75-80) is well handled. Particularly interesting are E.'s discussions of Suetonius' description of Nero acting tragic roles wearing a mask modelled on his own features (118-21) and of how the audience at the opening of Pompey's theatre might have responded to a spectacle which demanded to be seen at different times as referring to both Pompey and Agamemnon or to Agamemnon alone (89-91).5 In general, however, E. pushes his argument too hard, making claims which are simultaneously too large and too simplistic on the basis of evidence which is too often speculative or irrelevant. The idea that theatricality began with Livius is too simplistic on several grounds. Whatever one's views on the vexed question of Atellane farce and other pre-literary forms of drama (which E. notes at 155 n.3 but does not address or recognise as being of central significance for his argument), the claim that 'we should not confuse the lavish displays found in processions that predate the arrival of theatre in Rome with the terms "theatrical" or "dramatic"' (3) is surely too stark. The sharp division between theatrical and non-theatrical performance, spectacle and representation cannot be maintained, especially in a society where the theatre had no spatial association with a permanent building but rather a temporal one with ludi. Indeed E.'s subtle discussion of the interplay between praetextae and the use of ancestral imagines in funerals underlines the parallelism between the two unless, which he does not argue, the use of imagines postdates Livius' tragedy of 240 BCE. More generally, E. constructs too clean a view of 'reality' so that he can argue that it is blurred by the impact of tragedy. Every society, and perhaps Roman more than most, requires its members to perform roles and act at different times as audience and spectacle. Such performances may well have been conceptualised differently when the referent of the theatre was available, but to argue that Rome was a tabula rasa on which Livius inscribed the beginnings of theatricality is to argue too much. Likewise, E.'s determination to make Livius a πρῶτος εὑρετής leads him to claim too much for his innovations and, in the process, present a simplistic view of dramatic representation. E. argues that the reduced role of the chorus 'resulted in unrealistic expression of emotion by actors who sing in complicated metres rather than speak in a natural way in trimeters or senarius verse throughout the play' (11). Leaving aside the precedent of the kommos in Attic tragedy and the issue of how 'natural' (pace Aristotle) it is to speak in trimeters, to suggest that the heightened emotional level of a (masked) actor singing cantica would be considered 'unrealistic' is to misunderstand the very nature of ancient drama.

The forcing of the argument on the basis of insufficient evidence manifests itself in various ways. This is, of course, a danger inherent in dealing with fragments, but one which needs to be overcome rather than succumbed to. To show tragedians making tragedy relevant to their Roman audience's reality, E. finds a reference to the joining of hands in Roman marriage ceremonies in a fragment of Naevius' Danae where she simply remembers washing her hand (17). To have Ennius heightening 'the pathos of the scene for a Roman audience', he falsely states that the children did not appear in Euripides' Medea (27). He claims that Accius' use of Lemnian shepherds for the chorus of his Philoctetes is an alteration of Sophocles meant to connect with a Roman audience (2 and n.3), a non sequitur in itself and one which ignores the precedents of Aeschylus' and Euripides' versions.6 To emphasise self-consciousness, Livius' Aegisthus is 'a stock tyrant onstage' (12) despite E.'s insistence that this is the very start of Roman tragedy.

Much of E.'s evidence for metatheatre in Republican tragedy seems to come from its use of rhetoric, as he repeats verbatim on both Pacuvius' Antiopa and Accius' Medea, 'rhetoric removes the dramatic reality of the play from the audience's world to a metatheatrical (re)-creation of it.' (42, 50). It seems unconvincing that the use of rhetoric, familiar in political and forensic contexts, of itself renders theatre metatheatrical, especially on the basis of fragments which only sporadically demonstrate overt 'rhetoricization'. Moreover, evidence from the first century BCE of orators such as Cicero alluding to the theatre cannot be used -- on chronological grounds, even if one accepts the dubious logic -- to show that audiences in the second century BCE perceived Pacuvius' and Accius' tragedies as unrealistic because they contain characters who speak like orators. More generally, the paucity of evidence leads E. to make claims based on speculation, and then build (often circular) arguments on that basis. To offer two examples from many: 'for the sake of convenience, it is easier to imagine a linear progression and to assume that some of Livius' plays were produced before Naevius', although for plays written after 235 BCE the converse is equally possible' (14) and 'A correspondence has been detected between Livius' Aegisthus and the Agamemnon of Seneca, but Seneca's play may owe even more to the Clytemnestra of Accius, therefore suggesting that Seneca may have been drawn to metatheatrical elements in Accius' plays and included them in his own' (45, my emphases). Just as common as the use of the word 'may' -- and as frustrating -- is the quantity of unanswered, speculative questions which, though sometimes suggestive, make one long for a few more answers.

The most serious problems with the book (apart from its errors, below) lie with the way in which E. structures and supports his argument. For each tragedian, E. provides all the fragments, in sequence, from one of their best-preserved plays. However, the discussion of these fragments is often non-existent, generally skeletal and, for the most part, extremely superficial. They would at least provide a convenient resource, were not Warmington's Loeb readily available and E.'s translations unreliable. Six and a half pages are taken up by parallel quotations from Aeschylus' Persians, Accius' Brutus and the Octavia, with minimal and jejune comment (59-66), and then the page and a half of the Brutus fragment is quoted in full again in the next chapter (92-4).7 None of these extended quotations has line numbers but, since there is minimal reference to them, this absence is felt less than it might be. The search for evidence even leads E., in the absence of substantial fragments from Ennius' Ambracia, to 'compare similar material in historiographical accounts', by which he means the same author's (epic!) Annales, regardless of the impossibility of recovering precisely those specifically dramatic features which he seeks (including 'how they were staged') from the contrasting genres of historiography and epic (57).

There is also a certain vagueness about terminology. E. repeatedly and rather loosely refers to semiotics without its being entirely clear what he takes it to mean. Even the term metatheatre, central to the entire argument and apparently defined on p.5, suffers a marked degree of slippage. Allusion to events in 'the audience's reality' (a problematic concept in itself, as we have seen) may have something in common with, but are not the same thing as a play's awareness and advertisement of its status as a play. The greatest potential for the discussion of metatheatre comes in the chapters on Seneca and it is here that E. is perhaps most disappointing. He acknowledges the parallel of Boyle's treatment of metatheatricality in Seneca,8 and, apparently to avoid duplication, his discussion largely eschews the roles of Atreus and Medea as dramaturges in favour of rather hackneyed comments on rhetoric in Seneca and, for some not entirely obvious reason, parallels with Ovid. The idea of audiences onstage -- such as Atreus watching Thyestes' feast -- is a good one, and there are nice continuations of the depiction of Nero as imperator scaenicus in the previous chapter, actively blurring the boundaries between onstage and off, even if it does include some startling circularities.9 I wonder whether E. again overstates his case in suggesting that the presence of such audiences onstage marginalises the actual audience and reduces them to the role of witnesses. The distinction between audience and witness is not entirely clear, nor is the reason why onstage audiences cannot be focalisers for the real audience, drawing them in rather than excluding them, as in almost any instance of an Attic chorus viewing the ekkyklema.10 However, this at least is an arguable position, which is more than can be said of many adopted in this book

There are more than a few typographical errors and misprinted references. I have not checked all of the latter, but give here those which I have: 15: a noun has dropped out of ' ... Naevius inserts yet another interpretative [?] for the audience to consider ... '; 19: Enn. fr.99.3 Jocelyn: for institum read institutum; 31: 'whereas ... but not ... '; 35: for 'Telemon' read 'Telamon'; 40: Antiopa fr.8: for minitabilitlerque read minitabiliterque; 60: Aesch. Pers. 193 for εὔαπκτον read εὔαρκτον; there are several medial for final sigmas (lines 180, 198, 211, 212, 213) and a missing breathing in 213; 69-70 and 97: the second line of each of the Latin quotations, from Aeneadae sive Decius and Suetonius DJ 80.3 respectively, are in Roman rather than italics, as are Accius Atreus fr.3 lines 2 and 4 (105), and fr.13.2-3 (107); 76: for θεαο̂ read θεάομαι or possibly the late θεάω; 77: the sentence beginning 'If the stage was in sight of the corpse --' is never resumed after the parenthesis; 81: for insto read isto; 88: for inchohatum read incohatum; 92: for 'Sextus returned to Ardea to rape Lucretia' read 'to Rome'; 103: for euis read eius; 114: [Sen.] Oct. 444: for partriae read patriae; 116: for Cationem read Catonem; 133: Ov. Met. 12.252: for exsilvere read exsiluere; ibid. 12.269: for erviturque read eruiturque; 12.270 for sanquine read sanguine.

Notes: 159 n.55: there is some confusion here as the nonsensical 'Ennius stresses Medea's wealth for a Roman audience even in his Hecuba' provides the transition from discussion of the Medea to the Hecuba; 174 n.34 for [Pliny HN] 36.114 read 36.5; 175 n.46: the second fragment is from Naevius' Equos Troianus, not Livius'; 176 n.60: Cic. Att. 16.1 does not have 11 sections and does not refer to the ludi Apollinares; 184 n.130: for 'Dio Cassius 62.9.4' read '63.9.4' (the page headings in the Loeb are somewhat awry, which may explain this error), likewise 185 n.137, for 62.16.2 read 61.16.2; less accountably, the same note has two references to Dio 16 which should read 61; ibid. n.148 and n.149: for '[Dio] 62.20' read '61.20'; 190 n.30: for 'Grottius' read 'Grotius'. Bibliography: for 'Austin, Alan' read 'Astin'; 'Gorton, Charles' read 'Garton' (also at 29 and 160 n.58); Frank's article in AJP lacks a volume number. Index: for 'Phrynicus' read 'Phrynichus' (also 53); in several, but not all, cases, Gaius and Gnaeus are abbreviated as G. and Gn. rather than C. and Cn. The quality of the proofing is exemplified by the back of the dustjacket, which urges the reader to 'Browse our complete Classics catalog online at', not a URL, of course, but a pro forma in which 'subject' has not been replaced by 'classics'.

More serious and worrying are the number of factual errors and especially mistranslations. The latter are so numerous and, in some cases, so shocking (perhaps most egregiously, translating vinctus on p.44 as 'conquered') that they create an impression of carelessness and even unfamiliarity with the texts, which severely undermines the authority of E.'s discussion. I apologise for the length of the following list, but I would not wish cursory or Latinless readers of the book to be misled: 15: Naevius, not Livius, wrote the Bellum Punicum; 17: Naev. Danae fr.9: quae cannot be feminine accusative singular so read 'Those things which' for 'She whom', hence not a reference to Semele; 20: Enn. Iph. fr.99.8-9: for 'we go here and there, and wherever there is a movement, we are there too' read 'we go here, from here (we go) to there; when we have come there, it is decided that we (must) go away from there'; 26: Enn. Med. fr.9.1: traversa mente is not translated; 27: the children do appear in Euripides' Medea, e.g. 894-975, 1002-80, and even the joining of hands in Ennius fr.14 is closely paralleled at 1069-70; 36: Pacuv. Iliona: read 'and (who) do not pity me' (i.e. indicative not imperative); 38: Pacuv. Antiopa fr.3.6: add 'we do not understand'; 44: Acc. Astyanax, to 'enrich the ears of others' add 'with words', crucial for the antithesis; 60: Aesch. Pers. 213: for 'no city will hold him' read 'he will not be accountable to the city'; ἄγαν in 215 is not translated; 65: [Sen.] Oct. 740: for 'Some force drives the intention of the mind ... sends these things ... ' read 'Whatever (things) the attentive force of the mind is occupied with ... brings those things back ... 95: since Servius Tullius could not be a character in the Brutus, for 'Accius' Tullius refers to himself' read 'is referred to'; 97: E. prints without comment an obscure reading found in certain 16th century manuscripts, rex postremus factus est, as opposed to rex postremo factus est, printed by all modern editors of Suetonius and FPL (E.'s reading makes little sense while Caesar is alive but is essential for his use of the couplet as evidence of an association of Caesar with Tarquinius Superbus); 99: Appian, not Cicero, mentions pro-Republican supporters planted in the audience, as n.70 rightly states; 105: fr.6 should presumably read 'He who did not think it enough to have committed adultery with my wife'; in fr.7, the three infinitives are passive, not deponent, and regias is not a noun but an adjective describing matres, so 'for royal mothers to be contaminated, the stock (to be) fouled and the tribe (to be) corrupted.'; 108 fr.16: though the a of malis could be short at that position in the septenarius, the sense of 'with my jaws' is preferable to 'through these evils', and accepted by Warmington, Dangel and OLD s.v. mala (2). fr.18: the second i of vincite must be long, so for 'Subdue him!' read 'Bind him!'; 111: Odes 1.16.17-21 is not laid out in Alcaics, and it is not the only reference to the Thyestes myth in Horace, who alludes specifically to Varius' tragedy in his mention of saeuam Pelopis domum at Odes 1.6.8; 113: Sen. Thy. 207: tam ... quam ... means 'as much as' not 'rather than'; line 210: omits translation of volet; 114: line 212: for 'false praise comes unless he be powerful' read 'does not come unless'; line 213: for 'for there is no one who does not want the same' read '[then] no one will not ... '; line 214: for 'what is right' read 'only what is right'; line 218: for 'Let whatever qualities are beneficial to kings be prominent' read 'Let kings go where they please'; line 219: for 'It is wrong ... ' read 'Consider it wrong ... '; 115: [Sen.] Oct. line 440: for 'Is it right to treat relatives thoughtlessly?' read 'It is not right that anything be decided thoughtlessly against relatives.'; line 446: for 'ought to be' read 'has even greater (magis) need to be'; line 458: for 'whatever is serious is expressed' read 'whatever is forced is onerous'; 117: E. claims that Cicero uses Accius' oderint dum metuant to criticise Caesar at Off. 2.23, when the (different) quote is explicitly from an unknown play of Ennius (fr.182 Jocelyn) as E. himself notes at 181 n.98; 125: Sen. Thy. 203: for 'Let him destroy ... ' read 'He will destroy ... '; 132: Sen. Phae. 1084: for 'their burdensome yokes lie on the ground' read 'they throw their burden to the ground'; lines 1088-9: for 'the light chariot ... rushed' read 'with the chariot light/since the chariot was now light ... they rushed'; line 1096: for 'destroys' read 'is destroyed'; 133: Ov. Met. 15.525: for 'living wound' read 'living entrails', the point of comparison E. makes with Seneca's description; 12.238: for 'chunks of brain' read 'chunks of blood and brain'; 157 n.25: the quotation from Cic. Rep. 4 does not support the claim in the text at p.15; 184 n.128: Dio 59.13.6 shares Gaius' wish that the Roman people had one neck with Suet. Cal. 30, but not his quotation from Accius' Atreus.

Roman Tragedy contains a number of very good ideas, not least its overarching thesis. Unfortunately, its overall execution is poor. While one might take issue with the various weaknesses of E.'s argument, it is his methodology which most seriously mars this book, presenting scanty evidence on which he either does not comment at all, comments superficially, or builds extremely speculative claims. This, along with the large number of errors, means that, while E. succeeds in his modest aim to provide 'suggestions that [he] hope[s] will form the basis of further discussion' (xi), beyond that, his book cannot be recommended.

[[For a response to this review by , please see BMCR 2005.09.23.]]


1.   Notably the essays in Gesine Manuwald ed., Identität und Alterität in der frührömischen Tragödie (Würzburg: Ergon, 2000). I have, of course, not yet seen A.J. Boyle's forthcoming Roman Tragedy (New York: Routledge, Sept. 2005).
2.   Sander Goldberg, Epic in Republican Rome (New York: OUP, 1995).
3.   'The perpetuation of offstage theatricality affects the audience's perception of dramatic reality (the theatre's illusion of offstage reality), leading to the paradoxical question, "Is this realistic?" when what is being asked is whether the action on the stage accurately reflects the theatricality of reality off the stage (or the audience's perception of it)' (33).
4.   'These illustrative passages show the extent to which stage reaction has replaced stage action. Character reactions appear against a backdrop of spectacle, turning spectators into witnesses' (128).
5.   'The incorporation of offstage reality into the dramatic reality of the plays required the audience to have a selective response to the tragedies, resulting in changing perspectives: Are we watching Agamemnon's or Pompey's triumph? When does Agamemnon cease to allude to Pompey?' (91).
6.   It must be noted that some examples are more convincing, such as the omission of élite rowers and the substitution of contemporary fir-wood for pine in the prologue of Ennius' Medea, though both are already noted in Jocelyn's commentary.
7.   Cf. the substantial quotation from Syme repeated verbatim as evidence for Tacitus' potential to write, respectively, tragedy (164 n.44) and praetextae (170 n.44).
8.   Pp. x and 122 with 187 n.1, citing A.J. Boyle, Tragic Seneca (London: Routledge, 1997).
9.   E.g. 'the inherent theatrical and grotesque aspects of [Nero's] rule found in the art and poetry of the period seem to be reflected in the plays. If such a reflection is even remotely accurate, then offstage reality is just as bizarre and theatrical as anything seen in the theatre' (129).
10.   E.g. Soph. O.T. 1294-6: 'But he will display it to you also; for the bars of the gates are being opened, and you shall soon see such a sight as would drive to pity even one who hates him.'

Read Latest
Index for 2005
Change Greek Display
Books Available for Review

HTML generated at 13:30:14, Friday, 03 April 2009