Bryn Mawr Classical Review

Bryn Mawr Classical Review 2002.12.05

James Barrett, Staged Narrative: Poetics and the Messenger in Greek Tragedy.   Berkeley:  University of California Press, 2002.  Pp. 250 + xxiv.  ISBN 0-520-23180-5.  $49.95.  

Reviewed by Ruth Scodel, Classical Studies, University of Michigan (
Word count: 1497 words

Barrett addresses a real and fascinating tension in Greek tragedy. A messenger, as I. F. de Jong has shown in her study of the Euripidean messenger, is a focalizer, with his own inevitably restricted and partial point of view.1 Yet spectators and critics alike intuitively trust messengers as they do not trust what other dramatis personae tell them. Barrett explores the problem sensitively. He convincingly argues that the messenger does not necessarily have extraordinary authority but nonetheless has a conventional claim to it, a claim particular messenger speeches rely on or subvert. This claim is based both on the eyewitness status of the messenger and on characteristics of his reports that assimilate him to the poet.

Barrett does not rely on older studies of the "epic" characteristics of messenger-speeches (it is well into the book before he cites them). Instead, the first chapter examines the messenger in Persians, the oldest extant, whose narrative is in many ways impersonal and who goes distinctly beyond what he could realistically have seen: he is omnipresent (Barrett extends this omnipresence too far, by taking the Greeks as the subjects of 374; there is no reason to do this except some critics' desire to make Aeschylus as propagandistic as possible. This does not affect the overall point, however). At the same time, Barrett stresses that the messenger's authority rests on his presence at the events he describes, and his omnipresence is a conventional license. The second chapter discusses Homeric messengers and the image of messenger or herald in archaic poetry before looking at the complexity of the tragic messenger's movements. In Trachiniae, the Nurse separates herself from the other servants, giving herself an apparently perfect invisibility; in Alcestis, in contrast, the slave who speaks is merged with the others, as the report of the combat between the brothers in Phoenissae identifies the response of the speaker with that of the crowd. The messenger, with his (usual) invisibility and disembodiment, thus implicitly claims extradiegetic status. Barrett suggests that the messenger's typically low status is valuable in depersonalizing him but also that the tragic messenger is feminized through the loss of bodily power.

A chapter on Bacchae as metatheater argues that the messengers succeed where Pentheus fails, having the same ability as the spectator to see hidden things and survive. The lying messenger in Sophocles' Electra depends heavily on the narrative of the chariot-race in the Iliad, and especially on Nestor's advice to Antilochus. Orestes relies on metis in the race, just he adapts the Odyssean mode in his revenge. Barrett sees a close relationship between the repeated question of whether to act "openly or by cunning" in the Odyssey and Orestes' question to the Delphic oracle. The speech thus has powerful metatheatrical implications; it points to how its own ability to convince is based on its being a messenger-speech and a Homeric imitation. In Rhesus, in contrast, the charioteer who reports the death of the hero is so personally involved that his report is inadequate. His ignorance and confusion are thematically typical of the play.

The best chapters are very good. The sections on Bacchae and on the messenger's movements and visibility are especially valuable. The discussion of the limits of the messenger in OT is very interesting. Throughout, the writing is clear, and even elegant, making the book far more enjoyable to read than most theoretically sophisticated studies. Undergraduates will be able to handle it.

Not everything is equally helpful or convincing. Barrett may have focussed too narrowly on messengers in Homer and tragedy, instead of more broadly on the authority of narrators. For example, I do not find his connection of tragic messengers with Homeric heralds and messengers very useful. Homeric heralds and poets both speak with an authority that comes from the source of what they say or sing and their fidelity to it. Tragic messengers, however, are not transmitters of someone else's speech, and their authority is not based on being a faithful speaker for another. Indeed, many have not been sent by anyone, but report for their own reasons. They may therefore be more like the speakers of character-narratives more generally than they are like Homeric messengers. Some Homeric first-person narratives show a "drift towards omniscience" that provides a useful model for the authority of messenger-speeches.2 Eumaeus at Od. 15.403-484 goes far beyond what he could have understood as a child, and he quotes speeches he surely did not hear. Achilles at Il. 1.380-82 seems to know what took place between Chryses and Apollo at 33-43. If the external narrator has given no reason to distrust a speaker, and what the speaker says corresponds to the audience's expectations, other knowledge, or general plausibility, the audience is unlikely to be careful in distinguishing the messenger's facts from his inferences.

The reader should know in advance that this book does not address the issues in practical criticism raised by the messenger's combination of subjectivity and depersonalization. It is helpful in offering a richer basis on which to think about such problems, however. For example, de Jong has demonstrated that Euripidean messenger-speeches regularly include the speaker's inferences and judgments. At Heraclidae 813-17, for example, the messenger criticizes Eurystheus' cowardice. De Jong argues (77) that this is only the messenger's judgment, so that interpreters are wrong to be surprised that Eurystheus, when he finally appears, is dignified and resigned. The basic argument of the book would confirm the intuition of many interpreters that messengers' evaluations are indeed subjective but that they invite the spectator to accept them anyway. As de Jong points out, only at Medea 1171-72 does the text mark inference as inference by using που. Here, though, the messenger is reporting a response that is itself an uncertain inference: the old woman guesses that the princess' illness is caused by Pan or some other god, but she does not know which one. The exceptional particle is probably generated by the exceptional level of uncertainty -- he admits that he is guessing that she guessed. In Heraclidae, however, since we have heard nothing good about Eurystheus so far, we are likely to believe the messenger's explanation of his motives, especially since they fit the actions described by the messenger. So Eurystheus' behavior when he appears really is surprising.

Because Barrett does not survey all the speeches, the idea of what is "conventional" is hard to pin down. It may be possible to decide more precisely what the conventions are and where they are overturned. For example, messengers' understanding frequently fails in divine interventions. The herald in Agamemnon does not know what happened to Menelaos or how his own ship was saved (636-80). In Antigone, the guard's report at 249-67 defines precisely the distinction between what he has seen (the traces of the burial) and what he did not see (the action); the messenger in OC does not see what finally happened to Oedipus, though he enumerates what did not (1656-62). These examples suggest that the divine is a constant limit on the knowledge of messengers and that the tragedians are very aware of the distinction between the omniscient Homer and his characters in this respect ("Jorgensen's rule"). Rhesus would then be an extreme case.

I am not convinced by the treatment of Electra, for a number of reasons. First, Barrett overemphasizes the weight of the Homeric background in the chariot race. Although Orestes drives in accordance with Nestor's advice in the Iliad, there is no special emphasis on cunning as the essence of good charioteering in Sophocles, even though the point where special skill is demanded, the turn, is the same. The Sophoclean narrative stresses instead Orestes' care in holding back until only one other competitor is left. Much of the narrative is distinctly non-Homeric. I think he misreads the dispute between Menelaus and Antilochus in Iliad 23, where the issue seems to be less what happened than how it should be evaluated: seeing is not always enough. In any case, it is surely not the Homeric borrowings that make the speech persuasive. One factor is surely that the chorus has seen Myrtilus' death as the origin of the family's suffering (503-15), which makes Orestes' death in a chariot-race "poetically" right. It might have been useful to compare Neoptolemus' lie in Philoctetes (343-90), which is not a messenger-speech but relies on Homer more profoundly. Philoctetes finds the speech persuasive because it corresponds to his knowledge of the personalities involved, and because he does not recognize that it is based on epic stories with which he is unfamiliar. Within the fictional world, Neoptolemus uses his father's experience and Ajax's to create a lie that fits how things happen in the world as Philoctetes knows it. That is more interesting metatheatrically than the lie in Electra.

These are the quibbles and disagreements one has with a lively and thoughtful book. I hope that not only the specialists in tragedy will read it but our many colleagues in related fields who teach these texts.


1.   I.F. de Jong. Narrative in Drama: the Art of the Euripidean Messenger-Speech (Leiden: Brill, 1991).
2.   I discuss this in Credible Impossibilities (Stuttgart and Leipzig: Teubner, 1999) 65-66.

Read Latest
Index for 2002
Change Greek Display
Books Available for Review

HTML generated at 13:28:18, Friday, 03 April 2009