Tragedy and Athenian Religion (review)

This book, an impressive addition to Christiane Sourvinou-Inwood's wide-ranging body of work on Greek culture, argues that tragedy was "a discourse of religious exploration." Sourvinou-Inwood is always original and often, as here, combative; her great love of the Greeks and passionate commitment to understanding them communicate powerfully in everything she writes. The book's three chapters cover the fifth-century response to religion in tragedy, the ritual background, and religious aspects of all of Aeschylus' and thirteen of Euripides' plays. It is required reading for anyone seriously interested in the subject.

In the case of such an important and predictably influential work it may be most useful to express some doubts. Sourvinou-Inwood sets out to pursue independent lines of investigation whose conclusions will be validated if they converge, but the various subsets of evidence she deals with, almost all of which could be read in a variety of ways, are consistently read in a way that suits her general thesis, which is clearly (if unconsciously) playing a determinative role. She is, in other words, as guilty as the rest of us of constructing what Christopher Pelling (in Literary Texts and the Greek Historian) has disarmingly called a "Poirot," an explanation that accounts for the evidence, but need not be the only or the best explanation of it.

The discussion of some centrally important evidence in the opening pages of the book (6–8) is a good illustration. Sourvinou-Inwood's view is that gods ex machina were perceived by the audience as representations of the gods they worshipped, that they evoked epiphanies in the audience's lived religion (e.g., 514), that they were anything but "simple" (7, loaded word!) theatrical devices. Plato (Cratylus 425b) and the comic poet Antiphanes (fr. 189.13–17 K–A) both speak of tragic poets employing the deus ex machina as a shift when they are at a loss. Sourvinou-Inwood claims that since both are talking about the poets' motivations, neither has anything to tell us about popular perceptions of tragedy. The claim would hold only if we could conclude that the Greeks' extratragic experience never affected their response to tragedy, but someone capable of responding to Plato's or Antiphanes' jests is capable of responding directly to a tragic deus ex machina in the same way. It is therefore illegitimate to disallow such a response when constructing the "perceptual filters of the Athenian audience." (The unconvincing fallback position that "perceptions may have changed" (6) between Euripides and Plato/Antiphanes is another good indication that an a priori conception is at work.) Then there is Aristotle, who says that the mechane should be used not to resolve the plot but to introduce matters external to the drama, such as "subsequent events, which need to be reported or announced. For we ascribe (¢pod…domen) to the gods the power to see everything. But there should be nothing irrational within the action . . ." (Poetics 1454b2–7). Sourvinou-Inwood glosses thus: "the gods are appropriate for this, since in real life 'we' ascribe to the gods the power to see everything—in other words, this use of the gods is, in a limited way, mimetic of the audience's realities" (8, my emphasis). Here again the general thesis is guiding the interpretation of the evidence. The language of "attribution" is not the language of unquestioned certitude. Aristotle is justifying the use of the deus ex machina as a device for introducing things that human agents cannot know; if the capacity of gods to know and reveal everything were not a convention but a universally accepted proposition he would not be banishing it as "irrational" from the action. It is true that Plato and Aristotle are at an extreme on this issue, but the attitude they held was available to others, and it is therefore unjustified to assume "ancient perceptual filters" that filter out that attitude. This rare explicit evidence for ancient...