Oaths, Covenants, and Laws in the Athenian Reconciliation Agreement of 403 BCE

Antichthon http://journals.cambridge.org/ANN Additional services for Antichthon: Email alerts: Click here Subscriptions: Click here Commercial reprints: Click here Terms of use : Click here Oaths (ρκοι), Covenants (συνθκαι) and Laws (νόμοι) in the Athenian Reconciliation Agreement of 403 BC Christopher Joyce Antichthon / Volume 49 / November 2015, pp 24 - 49 DOI: 10.1017/ann.2015.1, Published online: 25 November 2015 Link to this article: http://journals.cambridge.org/abstract_S0066477415000015 How to cite this article: Christopher Joyce (2015). Oaths (ρκοι), Covenants (συνθκαι) and Laws (νόμοι) in the Athenian Reconciliation Agreement of 403 BC. Antichthon, 49, pp 24-49 doi:10.1017/ ann.2015.1 Request Permissions : Click here Downloaded from http://journals.cambridge.org/ANN, IP address: 130.195.253.48 on 23 Mar 2016 Antichthon 49 (2015) 24-49 doi:10.1017/ann.2015.1 Oaths (ὅρκοι), Covenants (συνθῆκαι) and Laws (νόμοι) in the Athenian Reconciliation Agreement of 403 BC* ABSTRACT There has been much discussion in recent years as to whether the Athenian amnesty of 403 BC can be understood in terms of political forgiveness. A number of scholars have denied that it can. Nevertheless, if the oaths, covenants, and laws are properly analysed, it will become clear that, though ancillary to the earliest clauses of the agreement, the promise μὴ μνησικακεῖν was a blanket measure forbidding prosecutions for crimes under the Thirty and before in the courts after 403. The covenants (συνθῆκαι) chiefly laid down conditions for future relations between Athens and Eleusis. The promise not to dredge up the past was just one concern, and should not be confused with the agreement in its entirety. This article re-examines the agreement as a whole in light of recent discussion of the meaning of amnesty in the ancient world, and argues that the oath μὴ μνησικακεῖν, sworn subsequently, affirmed just one of the covenants, not, as is sometimes held, every covenant. Other clauses may have been reaffirmed by separate oaths. The legislation (νόμοι) which followed was designed to give clearer legal definition to μὴ μνησικακεῖν, but was distinct and separate from the covenants of amnesty. There has been in recent years a renewed interest in the concept and practice of amnesty in the ancient world. Most recently Kaja HarterUibopuu and Fritz Mitthof co-edited a series of papers from a wide range of contributors analysing the application and understanding of amnesty from the time of the Pharaohs until the downfall of the Roman Empire in the West.1 Although each examines evidence from a different period, these articles reflect a broad shift in the way in which amnesty in antiquity has been understood. Traditionally, amnesty has been defined in terms of a reconciliation agreement between warring parties, often, but not always, in a civil conflict, which normalised civic relationships and restored the status quo ante.2 In contrast, the thrust of recent scholarship has been to * I would like to thank Edward Harris for his astute comments and suggestions on the final version of this paper. Certain comments which he has made on some of the questions raised by this topic have been included as a separate Appendix at the end of this article (see pp. 46-9). I would like also to acknowledge the useful comments of the anonymous referees for Antichthon. 1 K. Harter-Uibopuu and F. Mitthof (eds), Vergeben und Vergessen? Amnestie in der Antike, Wiener Kolloquien zur Antiken Rechtsgeschichte, 2013. See e.g. the definition by G. Thür in ‘Amnestie’, DNP 1 (1996) 602-3 of amnesty as: ‘Gesetzmässig festgesetzter Verzicht auf Anklage, Wiederaufnahme von Verfahren, Urteils-vollstreckung und Strafvollzug als Mittel, die streitenden Parteien nach internen 2 24 © The Australasian Society for Classical Studies 2015 The Athenian Reconciliation Agreement of 403 BC 25 challenge the idea that the practical application of amnesty in the ancient world, in any real sense, reflected a ‘Staatsnormalisierung’, or a turning back of the clock, which resulted in the reinstatement of affairs prior to a given conflict. Martin Dreher, for example, in a survey of amnesty agreements dating from archaic times to the end of the fourth century BC, has argued that some important attested cases in Greece evince a tendency not to bring about reconciliation and, indeed, that in some amnesty was a politically weighted mechanism which promoted a specific partisan interest.3 Lene Rubinstein has made some similar observations in respect of amnesty measures in Hellenistic times, where often larger power structures interfered in the internal affairs of cities and enforced the claims of one party or grouping for reasons of self-interest.4 From a slightly different angle, but with overlapping implications, Csaba La’da has argued that, while often a means of reconciliation, amnesty in Ptolemaic Egypt was not so much a peace deal between competing factions as an act of grace bestowed by royal edict, and the concentration of amnesties toward the end of the Ptolemaic dynasty might suggest the ineffectiveness of the central power to restore law and order locally.5 If examined on their own terms, amnesty seems in the majority of instances to have been less about the ‘abolition of the past’ than about the creation of new legal and political circumstances for achieving specified political objectives. Within the framework of this theoretical shift, it will be useful to re-examine the evidence for the best attested of all amnesties from the ancient world, the Athenian reconciliation of 403 BC which Cicero famously characterised as the Atheniensium vetus exemplum.6 Very recently, Edwin Carawan dedicated an entire monograph to this episode in 3 oder externen Krisen zu versöhnen.’ Implicit in this definition is the claim that amnesty measures are there to alleviate crisis and to bring about through legislation reconciliation in the wake of a conflict. In the case of Athens, with which this article specifically deals, I will argue that this definition fits the facts which we can glean from the ancient sources, but I leave open the question as to whether other amnesty arrangements in antiquity match the definition. M. Dreher, ‘Die Herausbildung eines politischen Instruments: Die Amnestie bis zum Ende der klassischen Zeit’, in Harter-Uibopuu and Mitthof (n. 1) 71-94. 4 L. Rubinstein, ‘Forgive and Forget? Amnesty in the Hellenistic Period’, in HarterUibopuu and Mitthof (n. 1) 127-61. 5 C.A. La’da, ‘Amnesty in Hellenistic Egypt: A Survey of the Sources’, in Harter-Uibopuu and Mitthof (n. 1) 163-209. 6 Cic. Phil. 1.1.1. For modern discussions as to whether Cicero truly understood the nature of the Athenian amnesty, see Th. Mommsen, Römisches Strafrecht (Leipzig 1899) 458; M. Sordi, ‘La fortuna dell’amnestia del 403/2 a.C.’, in Sordi (ed.), Amnistia, perdono e vendetta nel mondo antico (Milan 1997); E. Carawan, ‘The Athenian Amnesty and Scrutiny of the Laws’, JHS 122 (2002) 1-23, esp. 5-7; P. Scheibelreiter, ‘“Nicht Erinnern” und “Übles vergessen”: Zum Amnestiebegriff im klassischen Griechenland’, in O. Brupbacher (ed.), Erinnern und Vergessen. Tagungsband des europäischen Forums junger Rechtshistorikerinnen und Rechtshistoriker (Münich 2007) 365-84. 26 Christopher Joyce which he proffers some important new suggestions and expands some of the revisionist standpoints of his earlier articles.7 Most significantly, he seeks to disengage discussion of the amnesty of 403 from notions of political forgiveness and argues that, as a legislative measure, its use and scope was limited to a group of written covenants (συνθῆκαι) which the oath of reconciliation simply reinforced. This was by no means a universal amnesty, and to that extent should be understood, not in the Ciceronian sense of oblivio, but as a tightly controlled measure to reconcile the opposing sides under a limited set of legal terms. In its favour, Carawan’s study is in line with the changing Zeitgeist which has taken an ever more sceptical view of amnesty as a measure of forgiveness, but its general argument rests on some deeply problematic interpretations of ancient sources which I intend to analyse carefully here. In particular, his claim that the oath μὴ μνησικακεῖν solidified a series of legislative enactments, or covenants, involves a significant amount of surgery in reconstructing the chronology of events as represented in our most contemporary literary source, the speech by Andocides On the Mysteries, delivered around 400 to defend himself against the charge of impiety. Moreover, the comparanda he offers to support his view of μὴ μνησικακεῖν as a swearing-in of preexisting covenants are at best inconclusive, at worst refuting of the position he seeks to uphold. In two earlier articles, I challenged Carawan over the formula μὴ μνησικακεῖν.8 In contrast, I maintained that μὴ μνησικακεῖν involved far more than the mere solemnisation of a contract exchange and was a genuine attempt by Athenians to ensure, in the interest of restoring harmony, that citizens would not bring charges in court for past wrongs.9 My most recent paper was unable to respond to the expanded argument of Carawan’s monograph, which appeared after its submission, and so one of my present aims is to plug that gap. Specifically, I wish to take issue with his argument that the oath formalised a series of legislative enactments. The evidence indicates, to the contrary, that μὴ μνησικακεῖν was one of a series of provisions in the agreement, most of which had little or nothing to say about annulment of the past. The covenants provided mainly, but not exclusively, for future relations between Athens and the newly created enclave at Eleusis; their nature was different from μὴ μνησικακεῖν. Though the latter featured in 7 E. Carawan, The Athenian Amnesty and Reconstructing the Law (Oxford 2013). For earlier articles expressing comparable views, see n. 6 above and n. 8 below. 8 C.J. Joyce, ‘The Athenian Amnesty and Scrutiny of 403’, CQ 58 (2008) 507-18; ‘Mὴ μνησικακεῖν and “All the Laws” (Andocides On the Mysteries 81-2): A Reply to E. Carawan’, Antichthon 48 (2014) 37-54. The first of these was written as a reply to Carawan’s 2002 article (n. 6), the second a reply to his retort entitled ‘The Meaning of MÊ MNÊSIKAKEIN’, CQ 62 (2012) 567-81. 9 The oaths together formalise the text of a single written treaty, and not two separate ones: thus, A. Heuss, Abschluss und Beurkundung des griechischen und römischen Staatsvertrages (Darmstadt 1967) 7-8; P. Scheibelreiter ‘Atheniensium vetus exemplum: Zum Paradigma einer antiken Amnestie’, in Harter-Uibopuu and Mitthof (n. 1) 95-126 at 101. The Athenian Reconciliation Agreement of 403 BC 27 the covenants also, its effect was to eradicate the scars of the past, and cannot therefore be understood as a sealing clause for the earlier agreements. Orators refer to the amnesty under the umbrella phrase συνθῆκαι καὶ ὅρκοι.10 In all probability these were inscribed and set up in a conspicuous location.11 The main focus was to sort out property rights and other matters to do with the shrine at Eleusis, but they also included a guarantee not to dredge up past wrongs. It was this guarantee, and this alone, which granted that old wrongs would not be mentioned, but, because this needed further definition and clarification, subsequent laws were passed at Athens.12 The twofold nature of the amnesty, both to allay memory of the past and to lay down new conditions regulating the future, is a vital feature of the reconciliation. It might be objected that my view of the Athenian amnesty as a measure against dredging up the past swims against the changing tide of opinion, but I should like to emphasise that, in contrast to more traditional studies, which have understood the reconciliation of 403 mainly in those terms, my approach sees the abrogation of grievances as one aspect of the agreement, but not the only, nor perhaps, its most important aspect. The most important source for the covenants of amnesty shows that the majority of its provisions related to matters of future concern, and that the clause μὴ μνησικακεῖν did not in any way pertain to those; moreover, it indicates that the agreement of 403 aimed less at restoring the status quo ante than at stipulating new terms dealing with fresh and unprecedented circumstances. Once μὴ μνησικακεῖν is disentangled from the other covenants of the agreement, the full impact of this observation will be felt. The earliest negotiations which led to the conclusion of peace in 403 may have said little or nothing about ‘forgetting the past’, as the objective at that time was to provide refuge for those oligarchs who feared reprisals at the hands of returning democrats. The promise μὴ μνησικακεῖν could have been an additional measure once the main provisions for Eleusis had been established, and once the democrats decided that it was more in their interest to have an expanded community embracing erstwhile oligarchs than two small, weak, divided communities. 10 Lys. 13.88, 25.23, 28.34; [Lys] 6.39; Isocr. 18.29. 11 Thus J.L. Shear, Polis and Revolution: Responding to Oligarchy in Classical Athens (Cambridge 2011) 198-9. 12 I do not wish to revive here the now discredited view that a separate agreement was concluded after the re-integration of Eleusis in 401, to which Xen. Hell. 2.4.43 might possibly allude. What I do wish to suggest, however, is that there was more to the amnesty than just the covenants, or συνθῆκαι. The various laws which ensued made provisions not incorporated in the original terms. Against the view of two separate amnesties, see J.M. Stahl, ‘Über athenische Amnestiebeschlüsse’, RM 46 (1891) 251-86; T.C. Loening, The Reconciliation Agreement of 403/2 BC in Athens, Hermes Einzelschr. 53 (Stuttgart 1987) 26-7. 28 Christopher Joyce 1. DEFINITIONS AND CHRONOLOGY The ancient sources which detail the provisions of the amnesty agreement of 403 refer to three self-contained phases of reconciliation. The first, in chronological order, consisted of contracts drawn up between Athens and Sparta following the victory of the democrats over the oligarchs in the summer of that year. Ancient authors refer to these negotiations variously as συνθῆκαι, διαλλαγαί, or διαλύσεις, but they concur that they formed the earliest basis for the reconciliation following the civil conflict. These agreements were succeeded by a series of oaths, one of which was μὴ μνησικακεῖν, often, but not always, translated as ‘to forget the past’, and which may or may not have been backed by a decree of the people. Finally, we hear of supplementary laws (νόμοι) which came into effect once difficulties brought about by the amnesty came to light. Our most chronologically precise account, Andocides’ self-defence, details the main events in the sequence in which they occurred (On the Mysteries 81-90): when the party of Piraeus (the democrats) prevailed over the men of the City (the oligarchs), the people swore μὴ μνησικακεῖν, but it soon transpired that many were liable for offences under existing laws, and so a legal scrutiny followed, and some new laws were ratified which the orator summarises. In order to understand better the circumstances which led to the peace it would be profitable to review briefly the circumstances of the Thirty and their overthrow. When Athens lost to Sparta in 404 and a peace treaty was drawn up, a commission of thirty tyrants was set up with the official remit to research the ancestral laws and establish a new constitution for Athens.13 Very soon, however, they ignored their constitutional mandate and inaugurated a reign of terror in which around five percent of the resident population are estimated to have perished.14 Isocrates and Lysias mention a blacklist of targeted opponents who in some cases were murdered, in other cases had their property confiscated and were driven into exile.15 To ward off external threats, three thousand armed cavalrymen, or Hippeis, were sent to the Piraeus where they held down the population by force.16 Little by little, resistance took shape in northern Attica and, under the leadership of Thrasybulus, the democrats seized Piraeus. The Thirty withdrew to Eleusis, and the men who had stayed behind in the city chose ten 13 Xen. Hell. 2.3.11-12. 14 Thus, G.A. Lehmann, ‘Überlegungen zu den oligarchischen Machtergreifungen im Athen des 4. Jahrhunderts v. Chr.’, in W. Eder (ed.), Die athenische Demokratie im vierten Jahrhundert vor Christus. Vollendung oder Verfall einer Verfassungsform? (Stuttgart 1995) 139-50, esp. 145; W. Nippel, ‘Bürgerkrieg und Amnestie: Athen 411-403’, in G. Smith and A. Margalit (eds), Amnestie und Politik der Erinnerung in der Demokratie (Frankfurt am Main 1997) 103-19, esp. 107. 15 Isocr. 18.16, 21.2; Lys. 25.16. 16 Xen. Hell. 2.3.20. The Athenian Reconciliation Agreement of 403 BC 29 men – henceforth known just as the Ten – to negotiate with Thrasybulus and entreat the help of Sparta.17 King Pausanias and his general Lysander gathered an army to lay siege to Piraeus, but the democratic resistance remained firm, and so the Spartans appointed a commission of fifteen deputies to oversee negotiations between the two sides at Athens.18 The aim of the negotiations which ensued was to provide for the return of the exiles under the Thirty, restoration of property, and the possibility that anyone from the city who feared the return of the democrats could join the Thirty at Eleusis.19 In his account of the negotiations of early summer 403, Xenophon omits to mention an oath. The first time we hear from Xenophon of a clause μὴ μνησικακεῖν, it refers not to the events of 403 but to the later circumstances of 401, when Eleusis was reintegrated into the Athenian polis and new oaths were sworn.20 Nevertheless, it seems most likely that this was a re-affirmation of the oath sworn two years earlier, which Andocides cites verbatim, and which resembles the wording of the terms spelled out in Xenophon’s narrative for 403 barring the Thirty, the Ten and the Eleven from the city.21 Andocides attests that μὴ μνησικακεῖν was sworn after the compacts (μετὰ τὰς διαλλαγάς). He does not say that the clause μὴ μνησικακεῖν reinforced every covenant, but he seems to think that the pledge not to nurse grudges was taken after the initial bargaining. The chief aim of the διαλλαγαί was to reverse decrees of the Thirty which had driven opponents into exile and to provide circumstances whereby partisans of the Thirty need not fear future reprisals. The evidence indicates that the initial purpose of the διαλλαγαί was to provide for the return of the democrats and recovery of some, but not all, of their sequestered goods, but this developed into a further discussion about Eleusis and provisions for the settlement there of the Thirty and any other collaborator who wished to join them. The promise not to harbour grudges for past wrongs was included among the covenants of reconciliation, and Andocides clarifies that this was re-iterated in a separate oath (Myst. 90). When the democrats returned to Athens, their principal concern was to ensure that the decrees of the Thirty under which they had been exiled were annulled and that measures should be taken to specify the relationship between the two political communities. [Arist.] Ath. Pol. 39 shows that the pacts of reconciliation mostly pointed to future arrangements, whereas only one of 17 18 [Arist.] Ath. Pol. 41.2; Diod. 14.33.5-6; Lys. 12.58-60. Xen. Hell. 2.4.29-30. 19 Xen. Hell. 2.4.38 20 Xen. Hell. 2.4.43 21 Compare Xenophon’s wording at Hell. 2.4.38 with that of the oath cited by Andocides at Myst. 90. 30 Christopher Joyce its provisions, the clause μὴ μνησικακεῖν, had anything to say about the past.22 Carawan’s view of μὴ μνησικακεῖν as a formality which bound the συνθῆκαι in their entirety simply ignores the fact that this was one of the provisions stipulated in the agreement, and cannot therefore be read as a byword for the agreement as a whole.23 As we shall see, the oath μὴ μνησικακεῖν re-affirmed, though perhaps not with as much verbal detail, the covenant spelled out at §6, and was concerned to re-affirm that covenant alone. The fact that it begins with καὶ, as reported by Andocides, almost certainly indicates that it was preceded by at least one other oath, if not more than one, which affirmed other covenants of the agreement. In a strict legal sense, μὴ μνησικακεῖν was one clause of the agreement. When the orators refer to the amnesty in a blanket way they use the expression συνθῆκαι καὶ ὅρκοι, but do not refer to one clause only. It is vital to observe a proper distinction between μὴ μνησικακεῖν and the amnesty as a whole. The relationship between the oaths and covenants has been debated for over a century. P. Cloché supposed that the decree of general amnesty did not take place until 401, when the oligarchic enclave was dissolved.24 Cloché’s view was questioned by A. Dorjahn, who discounted the notion that the oath came about by decree, and preferred instead to speak of μὴ μνησικακεῖν as a morally binding commitment which never found official legal expression, but which carried the same force as law.25 Thomas Loening, in contrast, understood μὴ μνησικακεῖν as a legal term which prohibited action in a court of law for earlier offences and, on the grounds that the oath expressly provided against such litigation, argued against the existence of any formal covenant which spelled out the rule.26 More recently, Stephen Todd has argued that the oath formally forbade offences committed in the time of civil conflict but was taken only half-heartedly, as political show-trials rapidly ensued which violated the spirit of the oath, and which were allowed to take place while Sparta occupied its 22 It always used to be thought that the homicide rule spelled out in §5 was an exemption to μὴ μνησικακεῖν, but a careful reading of the text of Ath. Pol. 39 shows that it cannot point to the past; for a better reading, see Harris’s appendix to this article and, in a similar vein but with rather different implications, B. Gray, ‘Justice or Harmony? Reconciliation after Stasis in Dikaia and the Fourth-Century BC Polis’, REA 115 (2013) 369-401, esp. 385-7, 398-401. 23 In my two earlier articles (n. 8) I examined three inscriptions which show decisively that μὴ μνησικακεῖν did not in fact seal earlier agreements in any of the parallel examples which Carawan adduced in his 2002 paper (n. 6). Comparative studies of inscribed texts of other amnesty agreements confirm that μὴ μνησικακεῖν was, in the best documented cases, separate and distinct from the other covenants of reconciliation which brought together internal or external parties. 24 P. Cloché, La restauration démocratique à Athènes en 403 avant J.-C. (Paris 1915). 25 A. Dorjahn, Political Forgiveness in Old Athens: The Amnesty of 403 BC (Evanston IL 1946). 26 Loening (n. 12). The Athenian Reconciliation Agreement of 403 BC 31 attention elsewhere.27 From a broader historical standpoint, Nicole Loraux understands μὴ μνησικακεῖν to imply an attitude of oblivion to the past which, since Homeric times, was essential to the cohesion of society, and intended forgiveness for offences committed before the amnesty.28 Andrew Wolpert and Julia Shear have individually argued that, within the prevailing ideology of the restored democracy, blame was attributed to a very small group of outstanding political criminals who did not come to terms, and that a blame culture was, perhaps paradoxically, avoided by heaping blame on a limited group of miscreants; consequently, the only people who committed ‘wrong’ in the eyes of the restored democracy were the Thirty and their immediate cohorts who were not covered.29 Throughout this long tradition of scholarship there exists substantial disagreement as to the function of the oath and its relationship to the covenants. Now, Carawan proffers a daring solution.30 He holds that the amnesty came about in a series of legal covenants which μὴ μνησικακεῖν formalised. The purpose of the oath was to do no more than ratify pre-existing covenants; thus, a separate decree of amnesty is redundant. But in order for Carawan’s reconstruction to work, some tampering with the chronology is required. In those sections from his self-defence (81-90), Andocides differentiates between the stages of the reconciliation. First came the negotiations, or διαλλαγαί, which preceded the conclusion of peace (90); then followed the oath μὴ μνησικακεῖν (81); next arose objections by citizens liable under old laws (82), so a scrutiny and publication of the laws followed (82-85); once revised laws were written up, new laws were passed (85-89). The order is clear: discussions, oaths, re-publication of old laws, passing of new laws. Carawan’s treatment of the evolving agreement, in contrast, draws no clear distinction between the covenants of amnesty (συνθῆκαι) and the additional laws passed after the oath was sworn.31 His approach ignores the etymology of μὴ μνησικακεῖν, which 27 S.C. Todd, ‘Lysias against Nikomachos: The Fate of an Expert in Athenian Law’, in L. Foxhall and A.D.E. Lewis (eds), Greek Law in its Political Setting: Justification not Justice (New York and Oxford 1995) 101-32, esp. 120. 28 N. Loraux, La cité divisée: L’oubli dans la memoire d’Athènes (Paris 1997). Loraux’s monograph has been translated into English by C. Pache and J. Fort, The Divided City: On Memory and Forgetting in Ancient Athens (Cambridge MA 2002). 29 A. Wolpert, Remembering Defeat: Civil War and Civic Memory in Ancient Athens, Baltimore 2002; Shear (n. 11) esp. 295-301. In a similar vein, see also R. Thomas, Oral Tradition and Written Record in Classical Athens (Cambridge 1989) 132-8; A. Dössel, Die Beilegung innerstaatlicher Konflikte in den griechischen Poleis vom 5.-3. Jahrhundert v. Chr. (Frankfurt 2003) 110-12, 141-2; S. Forsdyke, Exile, Ostracism and Democracy: The Politics of Expulsion in Ancient Greece (Princeton NJ 2005) 262-3. 30 Carawan (n. 7). For his understanding of the relationship between the oaths and covenants, see esp. 69 with n. 4 and 184-9. 31 Carawan (n. 7) 68-90, esp. 86-9, where provisions (4) and (5) of the second part of the agreement according to his schema refer to laws cited by Andocides at Myst. 85-8. On p. 86 he does appear to recognise that further nuancing took place after the covenants were concluded and even claims that the rule ‘to apply the laws from the time of 32 Christopher Joyce refers to crimes of the past, not to agreements affecting the future.32 Moreover, he maintains, without good supporting evidence, that the covenants (συνθῆκαι) entailed, or implied, legal principles which the νόμοι listed at Myst. 85-8 re-stated.33 Unless Andocides has muddled his chronology, the oath could not have entailed any of laws which Andocides spells out at Myst. 85-8. Carawan’s effort to read the νόμοι back into the original covenants is problematic. The orator differentiates between at least two separate and self-contained stages of legal enactment, the earliest being the διαλλαγαί, or ‘exchanges’, that preceded the oath-taking, the second being the νόμοι enacted after the oath was sworn. In §90 he states that the oaths were sworn μετὰ τὰς διαλλαγάς, which cannot be identical with the laws to which he alludes in the previous sections. Andocides’ chronology shows that those νόμοι were additional measures taken after the ratification of amnesty, and not before. The laws cited cannot have been part of the agreement concluded after the democrats returned. In some sense, they added to, refined, or nuanced the agreement, but they cannot be regarded as identical or synonymous with it. Xenophon (Hell. 2.4.38) refers to fifteen men sent from Sparta to Athens to negotiate (διαλλάξαι). The verb implies the same διαλλαγαί to which Andocides and Lysias allude,34 arbitrated by the ten (or fifteen) arbiters, or διαλλακταί.35 He tells us that the arbitrations brought about peace between the warring factions, but Ath. Pol.’s account is more substantial: the covenants listed at 39.1-5 pertain to oligarchic émigrés to Eleusis and focus on property rights. The term used is not διαλλαγαί but διαλύσεις. We know from other evidence 32 Eucleides (403/2)’ did not feature in the original terms. But there is a strong implication that the essential stipulations of the new laws were stated, or at least implied, in the earlier covenants. For example, at p. 82 he writes: ‘. . . the testimony suggests that the new rules were based on the covenants or inspired by them [my italics]. Often we can only conjecture about the connection, whether the legislation adopted wording from particular clauses or followed the wider implications.’ This quotation suggests that the laws which came afterwards re-stated principles spelled out in the terms of the peace. But it is unclear why any of these later laws, the chief purpose of which was to clarify μὴ μνησικακεῖν in legal terms, should have had precedents in the covenants. The comparative case studies, which form the backbone of Carawan’s argument in all three publications listed above, that μὴ μνησικακεῖν in Greek legal terminology closed off contracts of reconciliation between the two sides, do not support any of Carawan’s contentions as regards the amnesty agreement of 403: see especially my 2014 article cited at n. 8. 33 Andocides’ chronology that the oath was sworn after the covenants were drawn up is followed by Ath. Pol. 39.6 and Xen. Hell. 2.4.37-8, both of which attest that μὴ μνησικακεῖν was sworn once the exchanges had taken place. It is noteworthy, however, that neither mention any of the legal principles which Carawan reads back into the covenants. 34 Andoc. Myst. 90; Lys. 12.53; see also IG ii 2 10 (= SEG 30.54) line 8. 35 Cloché (n. 24) 239-44 believed that two separate agreements were reached in 403, but his view has since been discounted by P.J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia, rev. edn (Oxford 1993) 463, and P. Krentz, The Thirty at Athens (Ithaca NY 1982) 102-30. The Athenian Reconciliation Agreement of 403 BC 33 that some of those covenants addressed property. The pledge μὴ μνησικακεῖν is separate. Andocides (Myst. 81) specifies that the oath was sworn once the negotiations had taken place. Its terms are not to be confused with the earlier covenants between Athens and Eleusis, as its provisions were relevant only to Athens.36 The διαλλαγαί, συνθῆκαι, or διαλύσεις were not the same legislative enactments as those Andocides calls νόμοι at Myst. 85-88. One of the aims of those νόμοι was to iron out imperfections in the oaths and covenants which had resulted in the ensuing legal quandary (Myst. 82), though this would not have been the sole aim, as Andocides seeks to relate the legislation specifically to his own circumstances. The etymology of the relevant terms militates against such identification. With διαλλαγαί, some exchange or bargain is implied; συνθῆκαι implies ratification of a treaty between two sides of a negotiating table.37 The implication is that the covenants were brokered with Sparta and the remnants of the oligarchy over property and citizenship. In contrast, the subsequent νόμοι were carried unilaterally by newly created lawgivers, and their concerns were mainly separate from those of the covenants (συνθῆκαι).38 The issue of criminality in the time of Thirty, and how to quash both civil and criminal action against oligarchs, were handled after the oaths and covenants were sworn. 2. THE COVENANTS (συνθῆκαι) AND THE OATH μὴ μνησικακεῖν When referring to the amnesty package, the orators used the blanket expression συνθῆκαι καὶ ὅρκοι. The main purpose of the συνθῆκαι and the oaths which ratified them was to lay down rules by which the restored democracy at Athens and the newly created oligarchic enclave at Eleusis could co-exist; their focus was directed towards the future. In contrast, the provision μὴ μνησικακεῖν concerned past actions solely. This is clear not only from the wording of §6 (τῶν δὲ παρεληλυθότων μηδενὶ πρὸς μηδένα μνησικακεῖν ἐξεῖναι, πλὴν πρὸς τοὺς τριάκοντα καὶ τοὺς δέκα καὶ τοὺς ἕνδεκα καὶ τοὺς τοῦ Πειραιέως ἄρξαντας) but from an inscription governing affairs at Iulis on Ceos immediately following the suppression of an anti-Athenian rebellion, in which the same formulation (τῶν δὲ παρεληλυθότων) is used in reference to past misdeeds of the losing side, which were to be wiped from the record under the terms of 36 Thus Joyce (n. 8, 2014) 42. 37 For the etymology of διαλλαγαί, see Y. Garlan, ‘Études d’histoire militaire et diplomatique I’, BCH 89 (1965) 332-48, who connected the term historically to the exchange of hostages. For attestations of συνθῆκαι, see LSJ s.v. This is clearly implied by Andocides (see below). The speech of Isocrates Against Callimachus is aware of a conceptual distinction, as it refers to the law of Archinus on παραγραφή in a separate sense from the συνθῆκαι (Isoc. 18.2). 38 34 Christopher Joyce the amnesty of 363/2.39 Though, as we would expect, a number of oaths would have been used to seal the various covenants of the amnesty, the one and only aim of the oath μὴ μνησικακεῖν was to provide immunity from prosecution for earlier crimes against the demos. It cannot have sealed every covenant of the peace, as the majority of the terms of the agreement had nothing to say about crimes of the past. In approaching the covenants of reconciliation we run up against an important methodological problem. Though Ath. Pol. 39 is almost certainly based on a reliable documentary tradition, it is clear from two extraneous authors, Lysias and Isocrates, that the account was abridged. Carawan holds that the narrative draws on two amalgamated documents, one dating from 403 and the other from 401, but the evidence for such an amalgamation is tenuous.40 It seems probable that Ath. Pol. referred either to an inscription or to an archival record, but still does not recount the covenants completely. Wherever Xenophon says anything of material value, his evidence confirms what Ath. Pol. says.41 We therefore have good reason to trust the account of Ath. Pol. at this point, despite its omissions. Importantly, we must not insert covenants into the schema where the term συνθῆκαι is not mentioned. As a starting-point for discussion I therefore offer in place of Carawan’s schema this twenty-point reconstruction of the covenants (the source references are always to Ath. Pol. unless specified otherwise): 1. 2. 3. 4. 5. 6. 7. 8. The reconciliation agreement took place in the archonship of Eucleides (403/2) (39.1). Those of the city party who wish to emigrate may have Eleusis (39.1; Xen. Hell. 2.4.38). They retain full citizenship rights with full power and authority over themselves (39.1). They are entitled to draw revenues of property elsewhere (39.1). The sanctuary of Demeter at Eleusis is to be common to both parties (39.2). It is to be controlled by the Ceryces and Eumolpidae according to custom (39.2). Those at Eleusis may not go to the city (39.2). Those in the city may not go to Eleusis except for the Mysteries (39.2). 39 IG ii 2 111 = Rhodes and Osborne 39, esp. lines 57-61. For my most recent discussion of this document, see Joyce (n. 8, 2014) 47-8, where I argue that, as in Ath. Pol. 39.6, the phrase μὴ μνησικακεῖν cannot refer back to earlier covenants but must pertain to misdeeds of the Cean rebels, who under the terms are re-integrated into the political community. 40 Carawan, ‘Amnesty and Accountings for the Thirty’, CQ 56 (2006) 57-76; contra Shear (n. 29) 191 n. 9. Compare Xen. Hell. 3.4.38 with Ath. Pol. 39.1 and 6. The only additional point Xenophon reports is that the Thirty, the Ten, and the Piraeus Ten are not to return to their homes, but this looks like an interpretation of the covenants mentioned in §§1 and 6. 41 The Athenian Reconciliation Agreement of 403 BC 35 9. The Athenians at Eleusis must contribute funds to the alliance (39.2). 10. If anyone departing take a house at Eleusis he must persuade the owner, but if they cannot come to terms each party is to choose three assessors (39.3). 11. Those natives of Eleusis whom the new settlers accept may stay (39.3). 12. Rules about registration for residents in Attica and abroad (39.4). 13. No one at Eleusis may hold office in the city before being re-registered (39.5). 14. Homicide trials are to take place at Athens according to ancestral custom (39.5). 15. No one is to remember past wrongs against any except the Thirty, the Ten, the Eleven [and the Ten in Piraeus] (39.6; Xen. Hell. 2.4.38; Andoc. Myst. 90). 16. If any listed above submit to accountings, they are to be protected (39.6). 17. Those ruling in Piraeus are to appear before courts in Piraeus; those in the city before courts of men with timemata. If they wish it will be possible for these men to emigrate (39.6). 18. Each side in the former conflict shall pay off its former war loans (39.6). 19. Items confiscated under the Thirty and not sold off may be recovered, but items sold off to a third party may not be recovered (Lys. Against Hippotherses lines 34-48). 20. Anyone who has informed against or denounced someone in the time of the Thirty is to be immune from criminal prosecution (Isocr. Call. 20). The first priority was to restore those who had lost their citizen rights under the Thirty. This was inevitably linked to matters of property, as citizens alone were entitled to own land. Athenians had to be able to show parentage on both sides. At least two separate decrees were passed in or shortly after 403 reaffirming the principles of Pericles’ citizenship enactment of 451/0, but there is no evidence that the covenants included rules about parentage.42 Concerning Eleusis, any member of the defeated could move to the enclave with full citizen rights there, which entailed autonomy and enjoyment of proceeds from property.43 Significantly, the specifications did not limit land ownership within the geopolitical boundaries of the newly created entities.44 Those who migrated to Eleusis appear to have held on to property elsewhere in Attica, as was the case half a century later when Athens put down a revolt at Iulis and permitted émigrés enjoyment of the proceeds of existing property.45 Thus, the covenants endowed the oligarchs, who in many cases owned estates, with extensive concessions, 42 Schol. Aeschin. 1.39 = FGrHist 77 F 2; Athen. 577b-c. See A. Dössel, (n. 29) 135-7. 43 Ath. Pol. 39.1. 44 Pace Cloché (n. 24) 251-3, who held that this applied only to Athens. 45 IG ii 2 111 = Rhodes and Osborne 39, lines 65-6. See also Xen. Hell. 4.1.35; Dem. 7.41; [Dem.] 59.102. 36 Christopher Joyce and, despite the emerging political division between them, a considerable overlap remained between Athens and Eleusis as far as ownership of land and estates was involved. Another matter to be clarified was access to the sacred precinct. This was to be used by inhabitants of both communities, and was to remain under the supervision of the Eumolpidae and the Ceryces. Apart from the sharing of sacred rites, traffic between the two communities was forbidden. Within this provision, those who emigrated were initially obliged to contribute funds to the Spartan alliance, though this was later abrogated.46 Carawan claims on the basis of an allusion in Hypereides that the rule against trespass was enforced on both sides by a covenant on arrest (ἀπαγωγή) and denunciation (ἔνδειξισ), but it is difficult to see how any of this is relevant to the situation in 403; Hypereides was speaking of a much later law which applied to his own time and which relates to the circumstances of Chaeronea in 338.47 The reference to contributions to Sparta can be read to imply either that the émigrés retained Athenian citizenship or that they did not, but I take the words καθάπερ τοὺς ἄλλους Ἀθηναίους to mean that oligarchs were obliged just like other Athenians to contribute. The implication here and elsewhere seems to be that citizenship was officially shared, but in practice a limitation on the ability of oligarchs to participate in the governance of Athens and vice versa was imposed by the exigencies of having to live apart.48 In the eyes of the restored democracy, the oligarchs at Eleusis were still Athenian citizens, but in their case a special provision was taken to practise a form of government which was different and separate from the democracy of Athens. Those who took possession of a house at Eleusis were to reach an agreement with the existing owner over price, but if an agreement could not be settled, the price would be decided by assessors.49 The political advantage of the émigrés over the original inhabitants entailed that those with whom the oligarchs did not wish to share polity would be required to leave. A narrow time-limit of twenty days after the swearing of the oaths was prescribed for an exchange of populations. Within the first ten days the decision to migrate would have to be taken.50 If émigrés had a change 46 Ath. Pol. 39.2. 47 Hypereides, Athen. col. 14, §29 refers to a law prescribing ἀπαγωγή and ἔνδειξιϛ against any who leave the community during war and later return. We have no reason to think that this law dates from the time of the amnesty; the orator seems to refer to a much later statute passed in the time of Philip of Macedon. 48 Thus Loening (n. 4) 35; contra Carawan (n. 7, 2013) 72. 49 Ath. Pol. 39.3. The implication is refusal to sell was prohibited: thus G.A. Lehmann, ‘Die revolutionäre Machtergreifung der “Dreissig” und die Staatliche Teilung Attikas (404-401/0 v. Chr.)’, in R. Stiehl and G.A. Lehmann (eds), Antike und Universalgeschichte, Festschrift Hans Erich Stier (Münster 1972) 222. 50 Ath. Pol. 39.4. The Athenian Reconciliation Agreement of 403 BC 37 of heart and decided at a later date to reside in Athens, citizenship at Eleusis could be renounced and re-enrolment at Athens was permitted.51 Carawan understands this permission to extend exclusively to natives of Eleusis, on the grounds that rules against trespass had already been laid down, but his view ignores the force of παλίν in reference to those who re-register.52 The prohibitions on trespass remained in effect against those who were registered in one community or the other, but if an oligarch decided to renounce enrolment at Eleusis and re-enrol at Athens, it was permitted for him to do so. Presumably, this would have meant that property bought up at Eleusis would under those circumstances have to be relinquished. Both sides, in addition, were expected to contribute to the Spartan-led alliance.53 A fragment of Lysias’ Against Hippotherses attests that among the συνθῆκαι was a rule that goods bought in the time of the Thirty would remain in possession of the new owner, but anything not sold would return to the possession of the original owner (38-43). As the speech of Isocrates Against Callimachus confirms, the Thirty had confiscated money and possessions of those who had gone into exile. These fragments of Lysias suggest that some of that property was put up for auction and sold off but the rest remained in the possession of the state. The question was how to compensate the returnees. Evidently, a complete restoration of property wrongfully extorted was impossible, as it may have changed hands more than once, and tracing lengthy transactions was impractical. Thus, it was decided that the only category of property which former owners could legally reclaim was that which had not been bought from the state by a new owner. The law of Archinus on παραγραφή later clarified that civil suits (δίκαι) brought contrary to the covenants of amnesty were illegal from 403 and were subject to a penalty of a sixth of the value of the claim (Isocr. Call. 1-4). The covenant to which the law referred was the same to which Lysias here alludes in this fragmentary speech. The aim was to ensure that simple matters of reclaim could be handled quickly and satisfactorily, but more complex cases which might have resulted in the outbreak of fresh hostilities were quashed. The majority of scholars have thought that the rule ‘no reclaim’ applied to land as well as to movable goods.54 This had been questioned more recently on that grounds that it contravenes the law invalidating 51 Ath. Pol. 39.5. For the interpretation that registration at Eleusis had to be cancelled, see P.J. Rhodes (trans.), Aristotle, The Athenian Constitution (London 1984) 83. 52 Carawan (n. 7, 2013) 73. 53 Ath. Pol. 39.2. 54 J.-H. Kühn, ‘Die Amnestie von 403 v. Chr. im Reflex der 18. Isokrates-Rede’, WS NF 1 (1967) 31-72, at 35; Loening (n. 4) 52-3; S.C. Todd, The Shape of Athenian Law (Oxford 1993) 234 with n. 4. Contra: M. Sakurai, ‘A New Reading in POxy 13.1606’, ZPE 109 (1995) 177-80; Carawan (n. 7, 2013) 84-5. 38 Christopher Joyce arbitrations under the Thirty.55 But there is no evidence that the law to which Andocides and Demosthenes later refer re-iterated a clause stated in the covenants; Carawan’s insistence that it did stems from a conceptual confusion between the laws and the covenants themselves. There is little evidence that a separate rule for immovable goods was included. Lysias was a metic and could not by law own land. The issue at stake in the speech Against Hippotherses is the recovery of slaves which had been taken from the plaintiff, and for some of whose value Hippotherses had sued. The issue of land is irrelevant to this speech; the contested matter was whether confiscated goods had been bought. Among the covenants was a rule barring lawsuits against informers and denouncers. In the speech Against Callimachus Isocrates refers to a covenant barring proceedings against informers and denouncers (20) and refers to the rule which upheld civil suits in the time of democracy (21-24). This was later reinforced by the law of Archinus which imposed financial penalties against any who litigated contrary to the oaths. This would have included murder cases brought by δίκη, and is further evidence that no special exemption was made for homicide in the terms of the covenants (see below). We learn that in the wake of the amnesty agreement many had started to ignore its terms. In this particular case, it seems that money was not the only issue; in addition was the claim of Callimachus that the defendant (here, the plaintiff in the παραγραφή) had denounced. What is clear from this and most of the other covenants is that the treaty of amnesty was weighted in favour of the defeated. The principal beneficiaries were those who otherwise would stand most to lose from the peace.56 The most often misunderstood of the covenants stated that homicides who killed or wounded by their own hand (αὐτοχειρίᾳ) would be liable to stand trial at Athens in the traditional way. This has conventionally been understood as an exemption to the rule μὴ μνησικακεῖν.57 But if the passage is read in the correct way, there is no need to read this as a special exempting clause. Its focus is not on deeds committed in the past but ones committed in future (see the comments of Edward Harris in the Appendix).58 Indeed, there is no sign in Lysias speech Against Agoratus that any special exception for specific 55 Andoc. Myst. 1.88; Dem. 24.26. Carawan (n. 7, 2013) 86 takes it for granted that this rule was specified in the earlier covenants, but Andocides ascribes it to a law passed after the covenants and oaths were ratified. 56 Carawan (n. 7, 2013) 88. 57 Among those who have taken the view that this covenant is retrospective, see e.g. R.J. Bonner, ‘Note on Aristotle Constitution of Athens XXXIX.5’, CP 19 (1924) 175-6; P.J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia (Oxford 1981) 69-70; S.J. Todd, A Commentary on Lysias. Speeches 1-11 (Oxford 2007) 639 with n. 55; E. Carawan, (n. 40), esp. 271-2. For Carawan, this has been crucial in developing the case that the amnesty had legal gaps and cannot be understood as a universal measure. For a slightly different formulation, which Harris in the Appendix criticises, see Gray (n. 22) 399. Like Harris, however, Gray understands Ath. Pol. 39.5 to refer to a future, not a past, arrangement. 58 The Athenian Reconciliation Agreement of 403 BC 39 types of homicide was made. If such a distinction had been made, we would expect the prosecution to have harped on a distinction between αὐτοχειρία and accessory, but it is noteworthy that this line of attack is not canvassed. Ath. Pol. 39.5 does not refer to crimes of homicide under the Thirty. The clause at 39.5 is about homicide trials in the future; it does not create an exception for those who committed the crime of murder with one’s own hand in the past. As the trial of Eratosthenes shows (Lys. 12 with Harris’s remarks in the Appendix), it was still possible to bring a case for homicide under the terms of exemption laid out at 39.6. Emphatically, this is not a qualification of the rule μὴ μνησικακεῖν. Only 39.6 relates to crimes committed in the past. Ath. Pol.’s account attests that the covenants contained a clause enjoining μὴ μνησικακεῖν. The orators, in contrast, refer to an oath. It appears that the pledge was included among the συνθῆκαι but re-affirmed by a separate oath. The fact that the covenants contained a clause μὴ μνησικακεῖν is sufficient proof that the phrase cannot have meant what Carawan reads into it. As with the amnesty at Iulis, μὴ μνησικακεῖν was one of many promises made by the terms of agreement; it cannot have ratified the agreement in its entirety.59 Its basic provision was that the Thirty, the Ten, the Eleven (and those officers in Piraeus?) under the oligarchy should not be forgiven. The exemption list in Ath. Pol.39.6 is not fully compatible with the other sources, but the evidence together shows that the Thirty and their cohorts were not protected under the covenant.60 The exempting clause entailed a further exemption, which stated that among the aforesaid those who were willing to submit to accounting were protected. The case of Rhinon, one of the Ten, shows that those who submitted to and passed scrutiny could not be prosecuted thereafter for their part in the oligarchy.61 This was the ‘exception to the exception’. While the scope of μὴ μνησικακεῖν was stipulated in the covenants, how the principle of forgiveness was to work in legal terms was specified in subsequent legislation. Still, it is likely that part of this definition had already been supplied by a term which Ath. Pol. does not cite but which Isocrates reports in the speech Against Callimachus (20): anyone who had denounced in the time of the Thirty was immune from prosecution. This was not as Carawan holds a guarantee against lawsuits involving property, but a promise to rule out criminal proceedings against collaborators.62 59 See the arguments of my 2014 article (n. 8). 60 For a full discussion of the problems of the sources and how to decide between them, see Rhodes (n. 59) ad loc. 61 We know that Rhinon was commended for his goodwill and benefactions towards the people and was later elected general under the restored democracy; thus, Ath. Pol. 38.3-4; IG ii 2 1371 line 10. 62 Carawan (n. 7) 88 misunderstands this guarantee; it was a shield against criminal, not civil, proceedings. 40 Christopher Joyce When the covenants were ratified, oaths were sworn probably on 12 Boedromion.63 The aim of μὴ μνησικακεῖν was to bar prosecution against all except the Thirty, the Ten and the Eleven. Ath. Pol.’s account refers not to the oath itself but to the covenant it reinforced; our most reliable witness for the oath is Andocides (Myst. 90): καὶ οὐ μνησικακήσω τῶν πολιτῶν οὐδενὶ πλὴν τῶν τριάκοντα <καὶ τῶν δέκα> καὶ τῶν ἕνδεκα: οὐδὲ τούτων ὃς ἂν ἐθέλῃ εὐθύνας διδόναι τῆς ἀρχῆς ἧς ἦρξεν. Notably, it re-applies the covenant (Ath. Pol. 39.6) about amnesty, its exemptions, and their exemptions, but does not re-affirm all of the covenants mentioned in the agreement.64 Carawan’s failure to note this is the fatal weakness of his entire case. The aim of the oath μὴ μνησικακεῖν was to reinforce one provision, and one only, which dealt with the past. The plural expression συνθῆκαι καὶ ὅρκοι implies there was more than one oath. It would be misleading to infer that μὴ μνησικακεῖν was the single principle to which the Athenians swore, since we know they also swore a bouleutic oath, as Andocides subsequently reports. If the oaths solidified covenants, it is important to recognise which of the covenants they reinforced. If Ath. Pol. 39 is read correctly, it ought to be clear from context that μὴ μνησικακεῖν covered none of the provisions spelled out at §§1-5, but only those of §6.65 In my 2008 article I claimed that the oath of amnesty came about by a separate decree. This claim has been criticised since on the strength of an old argument of Douglas MacDowell that ἔδοξε at Andoc. Myst. 81 was not meant in the technical sense.66 But, as Julia Shear recognises, it is equally possible that the covenants and oaths of amnesty, when published, were prefaced by a decree. The stele containing Draco’s homicide law (IG i 3 104 = ML 86) also contains a decree for re-publication. The decree of 63 Thus, Shear (n. 29) 209. 64 It is not completely compatible; the list of exemptions in Andocides does not mention the Ten or the Piraeus Ten. I am not persuaded by Shear’s claim (n. 29) 197 that the wording of the oath gleaned from Andocides cannot be taken as read, or that it did not detail those who were exempt and under what circumstances. The words given form a bleeding chunk of a larger oath, as D.M. MacDowell recognised (Andokides: On the Mysteries, with Introduction, Commentary and Apppendices [Oxford 1962] 130), but this would confirm my argument that μὴ μνησικακεῖν was not the only oath sworn. 65 66 Joyce (n. 8, 2008) 508; contra Shear (n. 29) 199 n. 34); Carawan (n. 7) 185 n. 25. The argument that there was no separate decree of amnesty was originally made by MacDowell (n. 65) 120 and 128, who argued that the decree to which Myst. 81 refers was the enabling decree for the government of the Twenty which was installed straight after the return from Piraeus, and that Andocides here refers to a resumptive clause, not a separate measure. But MacDowell’s argument is predicated on the claim that the oath of amnesty was brought into effect by the law cited later, barring litigation for crimes predating 403/2. As argued here, that statute was part of a later legislative package which gave further definition to the amnesty, but did not bring it into existence. For Carawan, MacDowell’s interpretation is crucial, as he claims that μὴ μνησικακεῖν cemented the covenants of reconciliation and did not have any separate decree authorising it. The Athenian Reconciliation Agreement of 403 BC 41 408/7 to which it refers sanctions publication of an older text. In the case of 403, the decision of the people to authorise the amnesty should not rule out a separate decree, just because the amnesty was defined by earlier enactments. Later sources refer to a decree of amnesty, and it seems clear that the Athenians decreed to swear μὴ μνησικακεῖν and to publish the oaths and covenants.67 If the assembly voted to have everyone swear an oath, then they would have to have passed a decree.68 Old coals were raked up after the conclusion of the oaths, as both Andocides and Isocrates confirm (Andoc. Myst. 82; Isocr. Call. 2). This has led some to infer that it was not taken seriously in the climate of vendetta that followed, or, as Carawan claims, that the scope of the oaths was limited and therefore permitted certain types of trial. Nevertheless, the trial of Socrates in 399 concerned the present crime of impiety and corruption; the fact that Socrates was condemned four years after the amnesty has little bearing on the terms of the amnesty itself. Andocides engages in a lengthy legal argument to show that, though his offence predated the fall of democracy and might not, according to the prosecution, have been protected under the terms of 403, Athenians took every measure after the restoration to ensure that those like him were sheltered.69 The παραγραφή against Callimachus (Isocr. 18) shows that, though many grey areas remained over property, criminal proceedings against former denouncers were barred. The trial of Eratosthenes (Lys. 12) involved a member of the Thirty and by the terms of exemption (Ath. Pol. 39.6) did not therefore violate the rule μὴ μνησικακεῖν. The trial of Agoratus (Lys. 13) is more difficult to assess as it inculpates a man who was not himself one of the Thirty. It did not, as Carawan maintains, entail a debate on whether a murder was committed αὐτοχειρίᾳ or otherwise: its aim was to suggest that Agoratus had committed homicide through his act of denunciation.70 As a murder case it seems that this one violated terms of amnesty, as it relates to a period of political instability, and it was decided in 403 that crimes of this sort would not be 67 Plut. Mor. 814B; Dion. Hal. Lys. 32; schol. Aeschin. 1.39; schol. vet. Ar. Wealth 1146a; schol. Tzet. Ar. Wealth 1146; schol. rec. Ar. Wealth 1146b; Val. Max. 4.1 ext. 4; Vell. Pat. 2.58.4; Nep. Thras. 3.2. 68 See the oaths in decrees of the assembly listed by Canevaro and Harris (n. 75). 69 Thus, Joyce (n. 8, 2014) 44-5. 70 Carawan (n. 7) chap. 6, esp. 125-35. In none of the passages he cites can he show that the phrase αὐτοχειρίᾳ was used against Agoratus, or that this case had any relevance at all to the specifications of Ath. Pol. 39.5. For the meaning of the controversial phrase ἔπ’ αὐτοφώρῳ, see M.H. Hansen, Apagoge, Endeixis and Ephegesis against Kakourgoi, Atimioi and Pheugontes (Odense 1976) 48-53; E.M. Harris, ‘“In the act” or “red-handed”? Apagoge to the Eleven and Furtum Manifestum’, Symposion (1994) 169-84. It is worth noting that Carawan cites Harris’s paper (n. 25) 126, only to misrepresent it. Hansen claimed that the phrase changed its meaning from the fifth to the fourth century. By contrast, Harris analysed all the passages in which the term is found and showed that the phrase did not change its meaning from one period to the next. 42 Christopher Joyce prosecuted.71 There is no reference in the speech to a special ‘rule’ in the covenants which exempted certain categories of killer from protection under the amnesty. Carawan’s tortuous reading of the case to imply that the amnesty did not cover every kind of crime ignores the fact that, as with the trial of Andocides, litigation could take place in spite of the amnesty which forbade it, and it remained for the defendant to show by the covenants of amnesty that the trial was illegal. 3. THE LAWS (νόμοι) (ANDOC. MYST. 85-88) When the democrats returned, the legally more shrewd among the oligarchs may well have foreseen complications, and this might explain, in some cases, why the decision was taken to migrate to Eleusis, despite the promise of shelter. It soon became obvious that a general proclamation of amnesty was not as simple or straightforward as expected. What was to be done about those who had committed murder and who, with blood on their hands, were barred from sacred precincts, or others who might have been exiled previously for crimes against the gods, like Andocides? What about slaves or metics who claimed citizenship? What measures should be taken to ensure that arguments over property would not escalate and threaten the peace? How were those who had lost sizable estates to be compensated? What about repossession of movable goods? What about shady cases where direct acts of violence had not been participated in but where individuals had been accessories to particularly heinous crimes? And what about old laws under which citizens were still liable? Were they to be abrogated? Or should limiting clauses be inserted to prevent application to crimes predating 403? These, and other, similar questions must have weighed upon the minds of legislators in the immediate aftermath of the amnesty decree. Letting go of the past had to be qualified in some way. Vitally, these later enactments were modifying measures; they make sense only against a backdrop of universal amnesty, since they resolved problematic cases where total oblivion was not practical, or where letting go could only be accomplished in a practical way if older laws did not conflict with the decree which guaranteed against recrimination for past grievances. These qualifying statutes (νόμοι) did not supply the main legal framework for the amnesty, which was guaranteed by the earlier oaths and συνθῆκαι. In an important sense, they were further calibrations to the rule μὴ μνησικακεῖν. Their most valuable effect was to define the hierarchy of laws against decrees, mentioned by Andocides who needed to show that the decree of 415 which had declared him ἄσεβης was inapplicable. But the legal issue, from the point of 71 On memory as a civic duty, see now A. Chaniotis, ‘‘Normen stärker als Emotionen? Der kulturhistorische Kontext der griechischen Amnestie’, in Harter-Uibopuu and Mitthof (n. 1) 47-70, esp. 50-5, on the trials of Eratosthenes and Agoratus. The Athenian Reconciliation Agreement of 403 BC 43 view of the legislators of 403, was to ensure that the decree of amnesty did not override, or cancel out, the enactments (νόμοι) which followed. Andocides (Myst. 85-8) does not recount the legislation (νόμοι) exhaustively. The laws he mentions are only those which had a direct bearing on his case. But the picture that emerges is that the enactments gave a more precise legal definition to the problems which the amnesty had unleashed. It has been suggested that the reference to ‘all the laws’ at Myst. 82 means only the laws of amnesty, those which Andocides partially recounts at 85-88 and that there was no comprehensive redaction of the laws of Athens, as once widely believed.72 That recent suggestion, however, ignores the context in which the reference appears. Andocides states (81) that the original contract of the lawgivers was to provide Athens with a constitutional blueprint, but that the remit was then expanded once the problems of amnesty came to light. There is no good reason to doubt that a complete compilation of the laws of Athens was authorised.73 The laws to which Andocides refers in the passage that follows are not related to the scrutiny; these were additional laws necessary to define the amnesty legally. The purpose of these νόμοι was to define more clearly what μὴ μνησικακεῖν amounted to in terms of the law. Evidently, a complete eradication of the past was not possible. As we have seen, democrats exiled in the time of the Thirty whose land had been confiscated demanded compensation. Without some legal modification, μὴ μνησικακεῖν would have failed to guarantee this, and, indeed, might have guaranteed against it. Furthermore, without the necessary adjustments, the oath might inadvertently have cancelled out legal transactions made in the time of democracy, before the period of the Thirty. This, no doubt, was one of the legal entanglements with which the legislators of 403 needed to deal urgently. And so it was decided that civil suits (δίκαι) and arbitrations (διαίται) taken in the time of democracy should remain valid, but as part of the same principle, such decisions taken under the Thirty were now invalid.74 As far as matters of private law were involved, there was a legal continuum from the old democracy to the new, interrupted only by the transactions of the Thirty now stricken from the record. But criminal procedures (γραφαί, φάσεις, ἐνδείξεις, or ἀπαγωγαί) for public liabilities were forbidden. Andocides (Myst. 88) draws on the distinction between the private domain (τῶν ἰδίων συμβολαίων) and the public (τῶν δημοσίων) to show that the case against him in 415, which had been a public liability, could not now be revived. Though the chief aim of the law negating past public liabilities was to reverse any action of this sort committed under the Thirty, from Andocides’ point of 72 Carawan (n. 7) makes this claim in all three works cited. For the most recent formulation of his ‘minimalist position’, see 2013, 7-19. For a totally different view of the legal redaction, see now Shear (n. 29) 286-312. 73 Joyce (n. 8, 2014) 48. 74 For slightly different wordings of this rule, compare Andoc. 1. 88 with Dem. 24.56. 44 Christopher Joyce view the crucial by-product was that any public liability from before the time of oligarchy was eradicated in addition. Apart from the matter of property, the legislation that came after the oath did not undermine the guarantees of amnesty. Its aim, however, was to ensure that Athenians understood in precise legal and procedural terms what μὴ μνησικακεῖν meant. It has been argued here that the legislation which Andocides cursorily summarises provided a clearer legal framework in which the pre-existing pledge μὴ μνησικακεῖν should be understood and practically applied. The important specification of the laws was that old liabilities to the polis were now swept aside. This did not contradict the oath, but clarified that its provisions did not pertain to matters of private law. The rule ‘to apply the laws from the archonship of Eucleides’ (Myst. 86-87) was to ensure that old laws under which citizens were liable for offences prior to 403/2 could not apply: a notable example is the case of Epichares, who had served on the Council under the Thirty and was liable under an ancient statute which proclaimed all conspirators against democracy ἄτιμοι (Myst. 95-102).75 These and other similar laws came into the limelight in the wake of the amnesty, and so it was decided that a limiting clause should be applied.76 To prevent other forms of vexatious litigation, the law of Archinus provided a special procedure of παραγραφή, so that, in the event crimes of the past were revisited, special legal procedures were put in place to block action.77 These laws were passed in the wake of the amnesty, so that the oath μὴ μνησικακεῖν could properly be understood in its legal sense. As Andocides attests (Myst. 82), this understanding was not firmly in place when the oath of amnesty was issued. The purpose of the legislation which followed on from the oath was to define in procedural terms exactly what kinds of legal process were forbidden for crimes predating 403/2. In addition to new laws, older statutes were also published. I have already addressed the aim of this process of redaction extensively in my previous two articles, and there is not space to repeat all of those arguments in detail. In brief, the purpose of the legal redaction was to clarify which of the older statutes passed before 403/2 were still valid and which not. The aim was to invalidate statutes like the decree of Isotimides of 415 which exiled Andocides, and under which the orator might still have been liable if a clear distinction between decrees and laws had not been made. This was part of a broader process of re-publication which had been going on since 410, and which perhaps finished at some point around 399 when Nicomachus, who had headed the commission, was tried. 75 For the problematic citation of the decree of Demophantus and other quoted documents in the speech On the Mysteries, see M. Canevaro and E.M. Harris, ‘The Documents in Andocides’ On the Mysteries’, CQ 62 (2012) 98-129. 76 Thus Joyce (n. 8, 2014) 53. 77 Isocr. 18 (Against Callimachus) esp. 1-4. The Athenian Reconciliation Agreement of 403 BC 45 4. CONCLUSIONS This article has suggested that the Athenian amnesty of 403 came into being in two main stages. The first was a set of legal covenants providing for Athens and Eleusis and guaranteeing protection for all except the most notorious oligarchs, with the oaths of amnesty solemnised. The second was a set of additional laws which limited the possibility of vexatious litigation against those who had fallen foul of the law before the archonship of Eucleides. The covenants were mainly, but not exclusively, directed to the future and affected both communities. The second stage focussed on the past and pertained exclusively to Athens. The most important and vital point to recognise is that the oath μὴ μνησικακεῖν solidified only one covenant of the amnesty agreement. The covenants of amnesty did not deal exclusively with the past, and it might be said that there was a greater degree of innovation in than ‘normalisation’. But letting go of past evils was a crucial component, and it was this, above all, which for later antiquity rendered the Athenian amnesty of 403 BC a lasting moral exemplum. The Haberdashers’ Aske’s Boys’ School Elstree, Hertfordshire APPENDIX BY CHRISTOPHER JOYCE [email protected] EDWARD HARRIS Suits for Homicide at Ath. Pol. 39.5 The Constitution of the Athenians attributed to Aristotle states that there was an agreement during the archonship of Eucleides to end the conflict between the men of the city and the men of the Piraeus (Ath. Pol. 39.1). The terms of this agreement dealt mainly with the relationship between the two communities created as a result of the agreement, one in the city of Athens, one in Eleusis. Among the terms is a clause about private suits for homicide (39.5). The first part of the clause presents no textual problems: τὰς δὲ δίκας τοῦ φόνου εἶναι κατὰ τὰ πάτρια (‘there are to be private suits for homicide according to traditional rules’). What follows is very corrupt: αυτοχιραεκτισιοτρωσασ, which has been emended in various ways.78 Thalheim proposed: εἴ τίς τινα αὐτοχειρίᾳ ἐκτείσαιτο τρώσας (“if anyone should kill anyone with his own hand after wounding”).79 Several scholars have thought that this clause makes an exception to the general amnesty for actions in the past, which follows in the next sentence (39.6: τῶν δὲ παρεληλυθότων μηδενὶ πρὸς μηδένα μνησικακεῖν ἐξεῖναι). It provides that the ‘amnesty is not to apply to cases of homicide or of wounding not covered by the homicide law ( . . . ) in which a man is accused not merely 78 See Rhodes (n. 35) 468 for various suggestions. 79 Thalheim’s emendation has recently been endorsed by Gray (n. 22) 399-400. 46 Christopher Joyce of arresting or of procuring a condemnation (that is of helping to implement the policies of the Thirty) but of doing the deed in person.’80 This Appendix does not propose a new emendation of the corrupt passage, but shows that this clause does not provide an exception to the Amnesty. It is important to study the structure of the entire passage so as to place the problematic clause in context. The passage gives the regulations for the two communities.81 These are to apply in the future. One of the issues the drafters of the agreement had to face was what had to change and what would remain the same. At Ath. Pol. 39.2 the temple at Eleusis is under the jurisdiction of both communities (innovation), but the Kerykes and Eumolpidae are to retain the priesthoods of the temple (traditional – kata ta patria). Those residing at Eleusis cannot go to the city and those residing in the city cannot go to Eleusis (innovation) except to attend the Eleusinian Mysteries (traditional). Ath Pol. 39.3 deals with the acquisition of houses – if the buyer/oligarch cannot persuade the resident to sell (traditional), assessors will be appointed to fix a price and arrange a sale (innovation). The last phrase has to do with registration of those leaving Athens and taking up residence in Eleusis (innovation). Now we come to 39.5. Those who live in the city cannot hold office at Eleusis, and those who live at Eleusis cannot hold office at Athens. All these provisions are prospective – they look to the future. Next we come to the phrase about homicide trials. The question arises, what happens if someone from the city kills someone from Eleusis or vice versa – this should be the issue addressed here, because the rest of the document regulates relations between the two groups. In other words, this clause should apply to any homicides that occur in the future. Here the agreement follows traditional procedure – kata ta patria – which should mean that the standard procedures for homicide should apply. This is the way the phrase is used in 39.2 about the Kerykes and Eumolpidae: they are to retain their traditional jurisdiction over the Mysteries.82 The clause about homicide trials should therefore mean that all the procedures followed in the past are to be followed in the future. As far as we can tell, this is in fact what happened. There is not much evidence for homicide procedure before 403, but we know from Antiphon’s speech On the Chorister (6.35-36; cf. Soph. OT 236-242) 80 Rhodes (n. 35) 468, following Cloché (n. 24) 259-61 and Bonner (n. 57) 175-6. For more recent bibliography see Gray (n. 22) 385 n. 49. Carawan (n. 7) 139-70 accepts this interpretation and claims that because of the rule contained in this clause, Lysias could not have delivered his speech Against Eratosthenes. As this Appendix demonstrates, this view is untenable. 81 Cf. Gray (n. 22) 386. 82 Rhodes (n. 35) 468 believes ‘There would be no special point in stipulating that for homicide trials traditional procedure was to be followed.’ But then why in a previous clause did the agreement use the same phrase (39.2: κατὰ τὰ πάτρια) and state that for the administration of the shrine at Eleusis the Kerykes and the Eumolpidai were to hold their traditional privileges? If the phrase was not pointless in that clause, it should not have been pointless in this clause. The Athenian Reconciliation Agreement of 403 BC 47 that after the Basileus received a charge of homicide, he made a proclamation that the defendant had to keep away from ‘from lustral water, libations, bowls of wine, holy places, and the marketplace.’ This practice continued after 403 (Dem. 20.158; Ath. Pol. 57.4). The speeches of Antiphon Against the Step Mother and On the Chorister also indicate that there were separate charges of deliberate homicide and involuntary homicide.83 This distinction was continued after 403 (Ath. Pol. 57.3-4; Dem. 23.22-61). What is important for Ath. Pol. 39.5 is that before 404 one could bring a charge of homicide not just against the person who caused death by direct violence or giving poison, but also against someone who gave an order to kill (Antiphon 1) or gave an order to someone to administer a potion that resulted in the victim’s death (Antiphon 6).84 This principle may have gone back to Draco’s law on homicide, if a common restoration of the first clause is accepted (IG i3 104, lines 11-13: δ]ι|κάζεν δὲ τὸς βασιλέας αἴτι̣ο[ν] φόν̣[ο] [τὸν ἐργασάμενον] ἒ [β]ολ|εύσαντα).85 One can see this principle at work in Attic tragedy. In Euripides’ Hippolytus, Theseus causes the death of his son by uttering a curse, which causes a monster to rise from the sea and frighten the horses of his son’s chariot, who drag the young man to his death. Theseus never touches his son, yet considers himself responsible for killing him (Eur. Hipp. 1448-50). In Aeschylus’ Agamemnon, Aegisthus plots with Clytemnestra to kill her husband. Even though she carries out the murder, he is still responsible for killing him (Aesch. Ag. 1613-14). As Andocides (1.94) states, the same principle applied after 403 (τὸν βουλεύσαντα ἐν τῷ αὐτῷ ἐνέχεσθαι καὶ τὸν χειρὶ ἐργασάμενον). The next clause starts with the words ‘but about past events it is permitted for no one to recall past wrongs against anyone’ (Ath. Pol. 39.6: τῶν δὲ παρεληλυθότων μηδενὶ πρὸς μηδένα μνησικακεῖν ἐξεῖναι). There are good reasons not to see the clause about homicide trials as an exception to the promise not to recall past wrongs in this clause. First, there is a strong contrast between the previous clauses, which concern the present and future, and this clause about the Amnesty, which looks to the past, creating a clear break between the clause about homicide trials and the clause about the Amnesty.86 Second, when one makes a general rule and provides an exception to it, one normally states the general rule first, 83 On the charges in these two speeches, see E.M. Harris, Democracy and the Rule of Law in Classical Athens: Essays on Law, Society and Politics (Cambridge and New York 2006) 398-400. 84 For this and what follows about liability in homicide law, see Harris (n. 83) 391-404, apparently unknown to Gray and Carawan. 85 For discussion, see M. Gagarin, Drakon and Early Athenian Homicide Law (New Haven and London 1981) 37-41. 86 Cf. Gray (n. 22) 387: ‘The sentence about the amnesty ( . . . ) seems quite clearly to mark a shift . . .’ 48 Christopher Joyce then the exception. There is a good reason for this: an exception only becomes meaningful in the context of the general rule. In fact, the following phrase gives a very good example of this principle: there was a general ban on recalling past wrongs, except for the Thirty, the Ten, the Eleven and those who ruled in the Peiraieus, unless they submit to their euthynai (Ath. Pol. 39.6: πλὴν πρὸς τοὺς τριάκοντα καὶ τοὺς δέκα καὶ τοὺς ἕνδεκα καὶ τοὺς τοῦ Πειραιέως ἄρξαντας, μηδὲ πρὸς τούτους, ἐὰν διδῶσιν εὐθύνας. Cf. Andoc. 1.00). Third, the way to indicate in Greek that one is making an exception is to use the word πλήν (‘except’). This word does not occur in the previous clause about homicide procedure.87 But the strongest evidence against the view that there was an exception to the Amnesty limiting cases of homicide to those in which the defendant was accused of doing the deed in person is the evidence of Lysias’ speeches Against Eratosthenes and Against Agoratus.88 In Against Eratosthenes the defendant is accused of arresting Polemarchus, the brother of Lysias, who was then condemned tο death by the Thirty and executed (Lys. 12.12-17). Lysias admits that Eratosthenes did not kill Polemarchus with his own hand, but performed actions that caused his death. The same is true for the case against Agoratus: the accuser says that Agoratus denounced his brother-in-law Dionysodorus to the Council (Lys. 13.29–33). As a result, he was tried by the Thirty and condemned to death (Lys. 13.39–42). I have no new proposal to emend the corrupt portion of the text of Ath. Pol. 39.5.89 For the moment all that can be done is to place brackets around the phrase starting with εἴ τίς τινα. At this point, there is no satisfactory solution to the problems posed by this passage, which may have been a marginal comment that crept into the text. What is relatively certain is that this clause did not provide an exception to the terms of the Amnesty or limit accusations of homicide to those who killed with their own hand. Both the democrats and the oligarchs appear to have agreed to have cases of homicide committed anywhere in Attica tried by the traditional courts according to the traditional procedures. The reason for this provision is possibly because the Areopagus, which was responsible for trying cases of intentional homicide, was viewed by both sides as an impartial body, which had stood above the conflict and not participated 87 Compare IG i3 40, lines 71-74; IG ii2 244, lines 90-95; SEG 26:72, lines 4-8. 88 Gray (n. 22) 399-400 realises that this speech creates a serious obstacle to his view. 89 The proposal of Gray (n. 22) 399-400 to defend Thalheim’s emendation (εἴ τίς τινα αὐτοχειρίᾳ ἐκτείσαιτο τρώσας) is vulnerable to several objections, as he himself recognises. One is that Athenian laws sometimes contain a conditional clause followed by a clause indicating a procedure to be followed (see e.g. Dem. 23.26, and E.M. Harris, The Rule of Law in Action in Democratic Athens (New York and Oxford) 139-40, but the conditional clause always contains ἐάν followed by the subjunctive, not the optative. Gray wants to restore εἴ followed by the optative ἐκτείσαιτο, which is unparalleled in Athenian laws. Another is that cases of wounding were tried under a separate procedure, the graphê traumatos, not by the dikê phonou. The Athenian Reconciliation Agreement of 403 BC 49 on either side.90 Both sides respected the traditional role of the Kerykes and the Eumolpidae; in a similar way they also respected the traditional role of the Areopagus. An additional reason for the provision was the fear of pollution, which would have affected both communities. But pollution was not just incurred by murder committed with one’s own hand, but also by causing death indirectly (Lys. 13.82; Eur. Hipp. 1448-50).91 Durham University EDWARD HARRIS [email protected] 90 For the role of the Areopagus as a guardian of the land that stood above political factions, see Dem. 23.66: ‘More recently no tyrant, no oligarchy, no democracy has dared to take trials for homicide away from this court alone. All men think that in this tribunal alone no defendant who has been convicted or accuser who has lost has ever proved that his case was wrongly decided.’ Cf. Zelnick-Abramowitz, ‘Guardian of the Land’, in Gabriel Herman (ed.), Stability and Crisis in the Athenian Democracy (= Historia Einzelschriften 220) 103-26. 91 This point is missed by Bonner (n. 57).