range of sources, but this evidence presents us with more than simply an ordering of
natural phenomena in a static grid of divinities (cf. Bremmer 1994:6). Rather,
the assumption that Poseidon, or the gods in general, are responsible for storms or
earthquakes is marshaled in a whole range of ways in different contexts. So,
for example, the defendant in the trialOn the Murder of Herodes, introduces the fact
that no fellow-traveler on a sea voyage had been involved in disaster, and that
all sacrifices on board ship in his presence had gone smoothly, as evidence of
his innocence (Antiphon 5.81–4). What to do with Andocides, however, accused
of the charge of profaning the Mysteries, but clearly the survivor of numerous
sea voyages? His prosecutors argued (but he denied) that the gods had preserved
him from punishment at sea precisely so that he might undergo trial in Athens
(Andocides 1.137–9; Lysias 6.19–20, 31–2). In part, this variety can be put
down to rhetorical convenience. (Aristotle, in hisRhetoric, and theRhetorica ad
Alexandrumblew the lid off all the seemingly sincere protestations of forensic
oratory by providing a guide to how to play the game of oaths, offering sophistic
arguments to suit every eventuality: AristotleRhetoric1377a12; [Aristotle]Rhetorica
ad Alexandrum1432a34–b4; cf. Demosthenes 54.38–41; Isocrates,Demonicus23.)
It is also, however, testament to the malleability of such religious beliefs, to the
presence of a live religious discourse that can be adapted to different, conflicting
ends. In the subsequent recitation of Andocides’ miseries, we glimpse also the way in
which such beliefs are reinforced through repetitive moralizing: no man should
become ‘‘less considerate of the gods’’ (atheo ̄terous) through seeing Andocides
saved from death, for a long life lived in distress is worse than a short one without
pain (cf. Harrison 2000:247).
The appreciation of the complexity of the Greek religious discourse preserved in
literary sources also leads to a very different understanding of historical change in
Greek religion. Xenophon’sAnabasis, his account of the journey to safety of the ‘‘Ten
Thousand’’ Greek mercenaries stranded in the heart of the Persian empire, has been
exploited largely as evidence of how widespread was thepracticeof seeking divine
guidance, through consultation at Delphi, through dreams, or through the exami-
nation of the entrails of sacrificial victims (Price 1998:1–3; contrast Parker 2004).
Instances in which the outcome of divination is disputed or even questioned as
fraudulent reflect, we are told for example, a ‘‘defensiveness’’ in Xenophon’s attitude
to divination – as if such doubts simplyqualifiedthe widespread practice of divin-
ation; as if, that is, only two discrete stances were possible towards divination:
credulity (Xenophon’s own stance) or skepticism (one which lapped about him). As
Edward Evans-Pritchard famously demonstrated of the Azande, however (1937), and
as others have, in fact, demonstrated of the Greeks (especially Parker 1985), the
possibility of fraudulence may actually serve as one of a number ofsupportsto the
belief in divination: fraudulence provides a way of dismissing inconvenient advice, or
of explaining the non-fulfillment of a prophecy. The failure to appreciate therela-
tionshipof different propositions concerning divination – that is, how together they
operate to reinforce confidence – leads to the underestimation of the resilience of that
confidence, and to a false impression of change: Xenophon’s acknowledgment of
fraudulence in divination (or his ‘‘defensiveness’’) may be interpreted as a reflection
of Greek doubts concerning the validity of divination, rather than, in fact, as a
symptom of its life (see also here Bowden 2005).
Greek Religion and Literature 381