In order to break out of this circle, we need then, first, to include in the study of religion a broader range of literary sources. Far from there being a distinct and bounded corpus of relevant texts, the entire Greek literary production should be taken into account - as it was indeed by scholars of a previous generation, notably E. R. Dodds (1951), Hugh Lloyd-Jones (1971), or Kenneth Dover (1974; cf. 1972): oratory, historiography, philosophy, tragedy, and comedy, as well as a miscellany of more minor or fragmentary works. Even those rare texts, such as Thucydides’ History, which in certain areas appear to eschew common religious attitudes, are relevant - insofar as they allow us to identify those areas of religious thought that were subject to criticism or skepticism and those that were not (contrast the perspectives of Hornblower 1992 and Marinatos 1981). In the words of John Gould, we should aim to ‘‘take in the whole range of the evidence, liturgical and literary, and to make sense of it as a whole whose parts are meaningfully related to each other’ ’ (1985:32). We also need, secondly, to ask a wider set of questions. The value of such literary texts does not consist solely in isolated mentions of certain ritual practices, or even in their evidence of the consequences of fulfillment or non-fulfillment of ritual in Greek thought (‘‘Every failure of due observance was thought to provoke divine anger and retribution’’: Bruit Zaidman and Schmitt Pantel 1992:28; cf. R. Osborne 1994:144) but also in the attitudes that they express and presuppose in almost any number of other areas.
A crucial point here is not to see these texts as static, presenting us with a mere list of propositions about the gods or their intervention in human life, but instead to recognize the interrelationship between different propositions, and thus to examine how such propositions were sustainable in the light of experience. This is an approach which can be traced back in modern anthropological literature through Godfrey Lienhardt and Edward Evans-Pritchard as far back as E. B. Tylor (Evans-Pritchard 1937, 1956; Lienhardt 1961; cf. Skorupski 1976:4-5), but which can also be found implicit in earlier classical scholarship (Dodds 1951:33; Dover 1974:133-44, 156-7, 241-2, 246-8, 1972:33; Lloyd-Jones 1971:3, 134): that propositions concerning divine intervention require the prior existence of‘‘blocks to falsifiability,’’ or ‘‘let-out clauses,’’ in order to be sustained. It is best explained by way of analogy to a different context. If one feels an unshakable belief in, say, the honesty and publicspiritedness of politicians, evidence to the contrary (the disclosure, for example, that a prime minister or minister gave favors in exchange for gifts or loans) will need to be explained away: for example, by saying that this individual is an unrepresentative ‘‘bad apple,’’ that the problem is only one of unfortunate perception (fueled by the disreputable press), or - especially if one’s concern is for the honesty of politicians of only one party - that it is a fault with the system, one which applies to all parties equally. (One need not suppose that the individual who holds to this belief, regardless of apparently contrary evidence, is disingenuous in their original commitment.)
To take now an extended ancient example, a number of authors reveal the assumption that certain actions will inevitably provoke divine retribution. This is made clear: through repeated proverbial remarks, for example Theognis’ warning to Polypaides on the dangers of deceiving a guest or suppliant (‘‘No mortal... has yet escaped the notice of the immortals,’’ 143-4; cf. Theognis 197-208, Solon fr. 13 West); through extended moral tales (the Herodotean story of Glaucus, 6.86; cf. Herodotus 4.205, Lysias 6.1-2, Andocides 1.29-30, Lycurgus, Leocrates95-7); and through passing assumptions (cf. Herodotus 1.159.4, 2.139.2 with Harrison 2000:103-4). These actions can often be described as acts of sacrilege (so, for example, perjury, the failure to respect suppliants, or the killing of envoys) but by no means always: a number of ‘‘injustices’’ (unprovoked violence, for example) also attract retribution, as well apparently as a man’s intentions or broader disposition (Herodotus 6.86, Solon fr. 13 West, Theognis 897-900). At the same time, reversals in fortune are frequently interpreted as due to an earlier misdemeanor. In some cases, where an individual is held to have committed numerous acts of sacrilege or injustice, an ancient author might concede a lack of certainty in attributing a comeuppance to a particular crime (Lycurgus, Leocrates 91); in other cases, where for example the very same men responsible for sacrilege are singled out for punishment or where there is a degree of appropriateness in the timing or nature of the punishment, there is no doubt (e. g., the Potidaea floodtide, Herodotus 8.129).
Clearly, however, the assumption that all unjust acts will be punished by direct divine intervention would be a difficult, if not actually impossible, one to sustain. But this assumption or belief can be reconciled with experience by means of a number of ‘‘let-out clauses’’:
Retribution is rarely direct. It does not always take the form of direct, divine intervention (for example, through divine epiphanies), but can come through more believable forms of intervention, more believable because they actually happen: roofs falling in, disease, madness, or even through human agency. There is nothing necessarily miraculous (because otherwise impossible) or ostensibly divine about the form of the retribution. The deduction that a misfortune is divine in origin is made on the basis either of timing (i. e., immediate, sudden, or delayed) or appropriateness (i. e., you or your father were known on inspection to have done such and such a terrible thing). So, for example (according to Xenophon in his account of the march of the Ten Thousand), if Clearchus had broken the terms of a truce and so committed perjury, his massacre would be deserved, as it is ‘‘just that perjurers should be destroyed’’ (Anabasis 2.5.38, 41). Prayers to the gods for revenge against an enemy do not envisage any agent of revenge but the author of the prayer (e. g. Theognis 337-50).
The gods do not punish every offence. In many cases they may be happy to leave vengeance to other men, only stepping in when that punishment is inadequate or excessive. ‘‘Leave the undetected sinner to the justice of the gods’’ (Demosthenes 19.70-1). The punishment of the Herodotean Pheretime - whose crime had been to impale all the men of the city of Barca and to cut off the breasts of their wives, in revenge for the death of her son Arcesilaus - revealed that ‘‘the over-harsh vengeances of men are abominated by the gods; she became infested with worms who ate her alive’’ (4.205). (The advantage that the gods have over men in this role of regulating human justice is that they cannot be hoodwinked: Xenophon, Memorabilia 4.4.21; cf. Archilochus fr. 177 West). In order to sustain this rather detached regulatory role on the part of the gods, human vengeance needs to be relied upon as the ongoing backdrop to occasional divine intervention - hence the regular insistence of our sources on the duty of vengeance, a duty incumbent on men from the gods (Lycurgus, Leocrates 146-50; Demosthenes 24.125, [Demosthenes] 59.116; Lysias 13.3; cf. 13.92).
Next, the gods know to look beyond single faults and punish only the pattern of offending behavior; so, though certain single acts may be punished, there is no point in looking for a corresponding punishment for every act. God is not ‘‘angry at mortal men for every fault’’ (Theognis 897). Proper ‘‘respect and fear’’ of the gods, on the other hand, restrain a man ‘‘from impious deed or word’’ (Theognis 1179-80). Xenophon’s Agesilaus, or the Ten Thousand of the Anabasis, focus on acting consistently in such a way that the gods will be their allies - and their enemies’ enemies (they are helped here by their enemies’ repeated perjury: Xenophon, Hellenica 3.4.11, Anabasis 3.121-3). This principle gives the belief in divine retribution a crucial flexibility - and provides a useful response to an obvious criticism, which surfaces for example in the context of the manipulation of oaths, that divine retribution is unduly legalistic in its workings and so allows an unjust outcome in the name of piety (e. g. Lysias 12.98).
Punishment may be delayed. The perpetrator of a crime can never be certain then that retribution does not still await him. ‘‘The minds of men’’ according to Theognis (197-208; cf. Solon fr. 13 West), ‘‘are misled, since the blessed gods do not punish sin at the time of the very act, but one man pays his evil debt himself and does not cause doom to hang over his dear progeny later, while another is not overtaken by justice; before that, ruthless death settles on his eyelids, bringing doom.’’ This possibility of deferred retribution gives some hope to those who are thwarted in their revenges (Theognis 337-50; Lysias 6.19-20). Andocides’ going about his business unpunished is seen indeed as evidence of his very brazenness, his life in advance of punishment full of ‘‘terrors and dangers’’; similarly a fragment of Lysias portrays a man guilty of sacrilege as having been singled out - because of the extraordinary nature of his crimes - for a living death of near-endless illness (fr. 9.4 Albini [Against Cinesias], apud Athenaeus 551a-552b). Even the anticipation of ultimate punishment is considered enough of an interim punishment by Clearchus in Xenophon’s Anabasis, so certain is the fact of retribution (2.5.7-8).
Delay in punishment can, of course, extend beyond the lifetime of the perpetrator of the crime. The punishment of children in place of their parents is something about which a wide range of ancient authors - at one level at least - apparently felt no qualms (contrast Theognis 731-52). As the orator Lycurgus proclaims, in the context of a famous passage on the importance of the oath, ‘‘If the perjured man does not suffer himself, at least his children and all his family are overtaken by dire misfortunes’’ (Leocrates79; cf. Demosthenes 57.27). As this last passage suggests, however, for one’s punishment to fall on one’s children is a kind of longstop rather than the preferred outcome (cf. Isocrates, Peace 120). Punishment may also be inflicted on the perpetrator himself after his death. The idea that unjust or impious actions in life might be, indeed will be, punished after death is common not only in Platonic texts - in the descriptions of the different routes to Hades of pure and impure souls in the Phaedo (107d-108c), for example, or of the judgment of those near death in the Gorgias (523a-524b) - but also in forensic oratory (Demosthenes 25.53, 24.104) or in the words of the semi-philosophical Isocrates (Isocrates, Peace 33-4; cf. Isocrates, Antidosis 282): ‘‘those who live a life of piety and justice pass their days in security for the present and have sweeter hopes for all eternity [tou sumpantos aionos]”; the immediate pleasure of those who take something that belongs to others is like the pleasure ofan animal that has been ‘lured by a bait’: Isocrates, Peace 33-4; cf. Antidosis 282).
The gods or the divine are not always just; people do not always get what they deserve. While misfortunes are frequently seen as evidence of divine justice, they can also be seen as the work of a capricious and essentially amoral divine; human fortune (in general seen as coming from the gods) is by definition changeable and unpredictable. This is an extraordinarily common idea over a wide period, reflected in a set of related aphorisms or gnomic pronouncements repeated throughout Greek literature: that one can never know the outcome of any matter until the end; that human knowledge is never certain; that the gods raise men up and cast them down; that fortune passes from one man to another; that human fortune is always mixed (or more darkly that suffering is inevitable, or that death comes to all men (so leveling their worldly prosperity) (for references see, e. g., Harrison 2000:38-9). Such expressions might seem to modern readers to be mere empty proverbs (indeed they were seen as well-worn lines by contemporaries: Andocides 2.5-6; Herodotus 7.51.3); they remained in currency, however, throughout the classical period.
That the gods are simultaneously represented as characteristically (if not quite exclusively) just and also as characteristically unjust might reasonably be taken to be a contradiction. There are a number of ways of trying (or failing) to reconcile this body of ideas with the assumption of a pattern of just reward and retribution. Reversals of fortune can be supposed to be themselves just, the result of an individual’s overreaching (Xenophon, Cyropaedia 1.6.44-6). Another possibility is to make a distinction between different classes of divinity. (For, as the author of the Aristotelian Magna Moralia puts it, if we assign the dispensation of good and evil to god, ‘‘we shall be making him a bad judge or else unjust,’’ 1207a6-17.) Isocrates, for example, distinguishes the Olympians (exemplarily benign, if not just: ‘‘those who bless us with good things’’) and daimones (‘‘those who are agents of calamities and punishments’’: Isocrates, Philippus 117). For some, the unevenness of justice is simply the way of the world: ‘‘we may both expect blessings and pray for them, but we must reflect that all things are conditioned by mortality’’ (Demosthenes 20.160-1). This can also be expressed in terms of the character of the gods: that the gods have the capacity to do bad things (though it is not their fundamental character, Aristotle, Topics 125a34-b3); that they can take their eye off the ball (Isocrates, Panathe-naicus 186); or, most commonly, that they are resentful of human fortune (Theognis 657-66, Herodotus 1.32.1; contrast Aristotle, Metaphysics 982b29-983a4). For others, the injustice of the gods - the possibility of undeserved misfortune or that ‘‘sinners and the just man are held in the same esteem’’ - is a problem or a cause of complaint (e. g., Theognis 373-400, cf. 585-90, 731-52). In general, however, the availability of different explanations for misfortune (at its simplest, retribution or sheer misfortune) and the absence of any dogmatic certainty as to whether ill fortune comes from the gods, fate, chance or daimones - far from constituting a problem - provides the necessary flexibility whereby the belief in the possibility of divine retribution can be maintained (see further Versnel 1990:1-38). Misfortunes can silently be filed, as it were, depending on the circumstances.
This kind of approach can be replicated in other areas of Greek religious experience. As we will see further below, Greek confidence in the efficacy of oracles and other forms of divination in securing guidance for action (and insight into the future) from the gods depends similarly on a number of let-out clauses (see especially Harrison 2000:122-57; Parker 1985; and Chapter 9 of this volume): the misrepresentation, selection or misinterpretation of a prophecy need not affect the institution fundamentally. Likewise there are a number of potential ‘‘let-out clauses’’ for Greek confidence in the efficacy of prayer and sacrifice: ritual impurity; the nature of the accompanying prayer (i. e., what was asked for); or the proviso, which significantly blurs the distinction between ritual observance and everyday actions, that a man’s life (his consistent propitiation of the god, through good times and bad, his just and sober living) must be taken in the round (e. g., Xenophon, Hipparchicus 9.8-9; Cyropaedia 8.1.23; cf. 1.6.3-4; Isocrates, Areopagiticus 29-30; see further Harrison forthcoming: ch. 2; Pulleyn 1997). The range of other such propositions to which Greek literature attests is such that any attempt at a list would be futile. One idea that requires particular emphasis, however, insofar as it arguably underpins much of Greek religious thought, is the principle of the unknowability of the gods, one which (as has been emphasized by a number of scholars: Gould 1985, 1994:94; Rudhardt 1992:88, 90, 101-6; Sourvinou-Inwood 1990:20, 1997:162; also Harrison 2000:191-2) is common to a range of authors, both those we might term ‘‘religious critics’’ and others usually conceived to be more traditionally pious. Far from being suggestive of a common religious agnosticism, or from qualifying traditional conceptions of the divine (and far from being just a reflection of the lack of clear divine revelation in Greek religion), ‘‘unknowability’’ in fact serves as a necessary complement to traditional conceptions: it was precisely because of the fall-back position that the best way to approach and the best way to envisage the gods were matters inaccessible to men that traditional attributes and forms of worship could continue unchallenged.
World History
