The manuscript of Andocides' speech On the Mysteries contains a series of documentary inserts culminating in the decrees of Patroclides (§§77–9), Tisamenus (§§83–4) and Demophantus (§§96–8). These decrees seem to fit their historical context and they are presented at length, with at least a few of the formalities that we would expect to find in the official record. Modern commentators have relied upon them as substantially genuine, allowing for the usual errors in transmission, but now their authenticity is contested. A close reading by Mirko Canevaro and Edward Harris rejects all three of these documents as products of feckless ‘forgery’. Alan Sommerstein responded with a strong defence of Demophantus' decree, and Mogens Hansen has recently defended Patroclides' decree and Tisamenus' as well.Footnote 1 On many points the defenders are persuasive, but the sceptics make a good case, and so it seems reasonable to reconsider how these documents took shape.
Both sides seem to proceed from a presumptive model of what the inserted documents represent: accurate or not, these decrees are presented as copies from official text, in original wording and substantially complete.Footnote 2 Such close copies seem plausible because transcripts of that sort are indicated elsewhere in the corpora of orators and historians, and (it is often assumed) such copies were readily accessible in a collection of decrees such as Craterus compiled (FGrHist/BNJ 342); for the latter was a compendium of transcripts arranged chronologically, where an enterprising editor could easily find what Andocides seems to describe. But that approach to the decrees in this case seems mistaken on both counts: in regard to the editor's methods as well as his resources. For gaps in the documents themselves and discrepancies between them and the orator's comments suggest that these decrees were reconstructed from abridged versions or fragments, not from close copy of official records. Such abridgement may be the work of an intermediary, but it is likely to begin with the decree collector himself. After all, Craterus often seems to have focussed on distinctive features and in many instances may have quoted only the relevant excerpts.
The manuscript documents that go back to such sources may be factual in content but ‘forgeries’ of a sort: not copied from the original decrees but patched together from excerpts and context that the source historian provided. Indeed, to call them ‘forgeries’ is misleading, for they seem largely free of deceptive imitation. They are, for the most part, ‘reconstructed’: the main content is reliable, but the formalities (such as the prescripts) are often haphazard, not crafted to beguile the ancient readers. Of course, there are bound to be mistakes in transmission, but the major defects may derive from gaps in the documentary source: where the historian provided context but quoted only select material, the editor felt free to expand the quote from the contextual clues. Where the document is most at odds with the orator's perspective, it bears witness to this editorial patchwork.
With this ‘reconstructive’ model as a working hypothesis, the present study proceeds as follows. First, it will be helpful to review the minor documents that illustrate Andocides' narrative; for these, too, seem substantially reliable, but the defects suggest a pattern for the major documents, the three decrees. From that perspective we should reconsider those three decrees case by case, with some regard for the legal argument that frames them. Then, in conclusion, we compare the documentary practices of Craterus and rival source collectors, whose material was likely to be available to the ancient editor. But before delving into the details, it may be helpful here to outline the supporting material that was brought into the speech and what was omitted altogether. Documents indicated in Andocides' speech On the Mysteries (hereafter simply Mysteries) include:
-
(1) Exhibits in the narrative section (§§11–70), especially lists of those implicated in 415 b.c., that have little bearing on the issue except to situate Andocides' case among others. All of the lists are represented in the text, in some form, but supporting testimony is sometimes omitted.
-
(2) Statutes in evidence on the legal issue: Andocides argues that the decree of Isotimides, which rendered him ἄτιμος, is cancelled and no longer enforceable because of a sequence of legislation that overruled or superseded it. These documents fall into two sets:
-
(a) The major decrees at our focus: Patroclides' decree cancelling ἀτιμία (§§77–9) is discussed at some length (§§73–6, 80) in order to introduce important distinctions affecting Andocides' case (though Patroclides does not shield him directly). Tisamenus' decree (§§83–4) stands where Andocides called for a measure of the early restoration (403/402) for δοκιμασία of old laws including those that authorize the ἔνδειξις against him; those standing laws were subject to limitation ‘from Euclides' archonship’ (not to apply to events prior to that year). Then, where we find Demophantus' decree (§§96–8), Andocides illustrates that rule: the Solonian law against councilmen or others who serve under a subversive regime cannot be invoked against Epichares (one of the accusers), only because of the limitation ‘from Euclides'. All three of the decrees are represented in some detail, but each document is doubtful to some degree.
-
(b) Following Tisamenus' decree, we find shorter texts of early restoration laws (§§85–8); Andocides confirms the gist of these, but some seem patently concocted from his comments.
-
-
(3) The rambling conclusion (§§101–50) consists largely of amplification on the legal issue and ranting against the adversaries. Witnesses are called for points of emphasis, but the speaker provides enough content for the reader to follow the argument; no documents are inserted.
In the first and the last sections, documents are omitted where Andocides gives context or explanation that seems sufficient. The brief documents that are included in the first section (the narrative) sometimes fill the gaps where the orator says little or nothing; these would have helped the original readers to see what the evidence was supposed to show. But then in section 2, at the heart of the legal argument, there are the three decrees, Patroclides', Tisamenus' and Demophantus', which serve a different purpose, as each is framed by substantial explanation to which it does not entirely conform.
I. THE EVIDENCE IN THE NARRATIVE
The argument in Andocides' Mysteries, as we have it, probably followed closely the prepared text, with perhaps minor adjustments after the trial.Footnote 3 In the draft that the orator prepared for himself before trial the documents would be omitted; these were for the clerk to read from the dossier at court, not for the orator to recite in his own voice. Documents were then brought into the narrative at two stages in the transmission: (1) In the version for early circulation, soon after the trial, there was little incentive to include documentation for its own sake, but there are two inserts that were probably included in that first edition precisely because the material was not summarized in the main text and the turn of thought would be unintelligible without it. (2) There is also a sequence of short documents that slightly expand the narrative details and seem largely reliable: the denunciations (μηνύσεις) at §§13, 15 and 35. These are often treated as authentic documents though, most probably, copied into the manuscript by a later hand (not in the author's version for circulation); for Andocides himself would have no reason to include them merely for documentation (he names notables in the main text); and if he had, he would surely have mentioned more than first names.
Likely inserts in the version for circulation. For early readers the interrogation of Diognetus at §14 seems practically indispensable, as the narrative reveals nothing of what Diognetus would say.Footnote 4 The inserted text reveals that Diognetus was one of the investigators in 415 (ζητητής), and he confirms the list of names represented in §13 as men denounced by Andromachus. Without the brief interrogatory, we have merely the prompt, κάλει Διόγνητον, before Andocides proceeds directly to the next information.
By contrast, there is no insert at §17, where Andocides calls for confirmation from kinsmen of the accused and explains their relationship; no document is needed. Likewise, in §§27–8 Andocides describes the content of the testimony (payment for the informants Andromachus and Teucrus), and at §46 Andocides calls for confirmation from those who were πρυτάνεις, in charge of the agenda when Dioclides informed against so many. In each instance we need no illustration and have only the lemma, μάρτυρες.
But then, at §47, we find a list of men who were denounced by Dioclides (later discredited), with a line-by-line comment emphasizing Andocides' kinship to the men he claims to have saved by his ‘confession’. Translating this passage, MacDowell italicized the names to indicate what the clerk actually read:Footnote 5
[Y]ou'll hear the names of the rest. Read out the list for them.
Charmides son of Aristoteles. That's my cousin. His mother and my father were sister and brother.
Taureas. That's my father's cousin.
Nisaeus. Taureas' son.
Callias son of Alcmeon. My father's cousin.
Euphemus. The brother of Callias son of Telocles.
Phrynichus the former dancer. A cousin.
Eucrates. Nicias' brother. He was Callias' brother-in-law.
Critias. He also was a cousin of my father's; their mothers were sisters.
All those were among the forty men he listed.
The list includes some patronymics such as we would expect in the text of record, but personal relationships are also noted, beyond what the official text would say. Again, the speech that Andocides prepared before trial probably included no such list, with or without annotation: the copy of Dioclides' information would have been included in the dossier for the clerk to read (not for the orator to recite). But it would certainly be effective, in the version for circulation, for Andocides to add these notes on his family connections, whether or not he interrupted the reading to add those details at trial. So I conclude that Andocides himself probably inserted this text and added these personal notes to the circulation copy for the benefit of his readers.
The reconstructed lists. Equal authenticity is often assigned to three other lists of names presented as denunciations (μηνύσεις) against those implicated in the affairs of the Herms and the Mysteries. Among those listed, some of the names are introduced in Andocides' narrative, so it may have seemed superfluous to copy the documents into the speech for circulation.Footnote 6 And the ‘documents' we have are introduced in a perfunctory way which the editor evidently supplied from the narrative: (§13) ὀνόματα. τούσδε Ἀνδρόμαχος ἐμήνυσεν· (§15) ὀνόματα. τούσδε Τεῦκρος ἐμήνυσε· (§35) ὀνόματα. Τεῦκρος ἐπὶ τοῖς Ἑρμαῖς ἐμήνυσεν. That format does not match the epigraphic evidence,Footnote 7 but we have good reason to suppose that the inserted documents are accurate, as far as they go, because the Attic Stelai (listing property confiscated in 415) verify some of those that Andocides does not name in his narrative.Footnote 8
This material, however, though confirmed by the Stelai, differs significantly from the annotated list in §47; for in these denunciations (§§13, 15, 35) we are given nothing but first names. We would not expect to find family connections spelled out, as in §47, but the text of record had to include more specific identification. Most of the names are commonplace, and even an Alcibiades must sometimes be distinguished by deme, as in §65. If Andocides himself had copied these documents into the speech for circulation, he would surely have included patronymics, demotics or other identification, as he did in §47.
From these features it seems reasonable to treat the three denunciations as largely reliable but not close copies of any official text. If the editor had the records in hand, he was not likely to omit the crucial details by which these men were identified. It seems rather more likely that these μηνύσεις were constructed from some other summary of the events; perhaps the additional names, those that Andocides does not mention, were jotted in the margin alongside those that he does name, and a later editor brought them into the text under headings that he adapted from Andocides' own wording. However they were introduced, the lists we have are not authentic documents in the strict sense but products of some second-hand composition.
With that model in mind, let us turn to the three decrees for which substantial documents have been inserted in the speech, taking them up in the order of presentation. Also included in the legal argument are those early restoration laws (§§85–8) which the orator summarizes and for which the slim inserts often seem quite contrived, but let us focus on the decrees that frame that litany of laws. For the three decrees, the inserted documents are not essential to the train of thought; the main text tells the readers what they need to know to follow the argument, and Andocides had no reason to copy these statutes into the version for circulation, when they were still familiar and easily accessible. It was evidently a later editor who found or fabricated what he thought would fit the description, when the text of record was no longer available.
II. PATROCLIDES' DECREE (P) AND ANDOCIDES' DESCRIPTION
On formal details in P, Hansen has given a point-by-point rebuttal to Canevaro and Harris and, in my view, shows that many of their objections are hypercorrect and misleading.Footnote 9 But, on balance, the disparities are sufficient to suggest that P is not a close copy of the decree. Some of the points in dispute are reconsidered in this section, but I focus on the major incongruities for which neither side has given a convincing explanation. Let us begin with the very purpose of this emergency measure in 405.
Andocides introduces Patroclides' decree in §73 as a measure to restore citizens to their rights, ἀτίμους ἐπιτίμους ποιῆσαι, and he promptly proceeds with an explanation of various types of ἀτιμία.Footnote 10 The document P is inserted after the primer on ἀτιμία, beginning §77. As for its purpose, P says only ‘for the people to decree the same measures as were taken in the Persian Wars to the advantage of the Athenians'. Later, in amplification (§§106–9), Andocides alludes to the archaic precedent, and, as Canevaro and Harris suggest, the ‘forger’ probably relied on that comment to invent the one clause in the document that characterizes the scope of the amnesty (§77), ψηφίσασθαι τὸν δῆμον ταὐτὰ ἅπερ ὅτε ἦν τὰ Μηδικά.Footnote 11
For P itself does not quote the essentials of the amnesty;Footnote 12 instead, it proceeds directly to list the records to be deleted, without any provisions describing the cancellation of old liabilities per se.Footnote 13 That shortcut has never been clearly explained. Yet, if this is to be read as the full text of so critical a change in the status of so many Athenians, it seems inadequate not to say more straightforwardly what liabilities are cancelled. Instead, P begins, ‘Since the Athenians have voted immunity (for anyone) to speak regarding … those indebted …’, calls for ‘the same measures as (approved) in the Persian Wars', and then turns to the tally of official records, ‘concerning those listed with the πράκτορες …’, and so on. Thus the document merely alludes to the patriotic model and then proceeds directly to the long list of records to be deleted. That treatment seems to me more likely to derive from an abridged version in an historical account that emphasized the adaptive strategy of erasing documentation. For, whether there was any such reprieve in 490 or 480, disposing of the accumulated record of debts and penalties was not so burdensome a task then as it would have to be in 405.
Of course, the archaic precedent in P (‘the same measures as … in the Persian Wars') may have been added by an editor or compositor of the document, either from Andocides' amplification (§107) or from an historian's explanation. For the only part of P that reverts to long-standing rules is the exception (§78) regarding those condemned for homicide and other unpardonable crimes.Footnote 14 Here again, rather than define the crime, P focusses on the public record, ὁπόσα ἐν στήλαις. Those excluded are the ones proscribed on stelai or condemned by some court or tribunal; those records are not to be deleted. ‘All others … and any copy anywhere’ are to be erased:Footnote 15 no one is allowed to keep a copy even privately; anyone who does so is subject to exile. In fact, the main content of P is all about deleting records, a task that loomed large in an era of ever greater documentation.Footnote 16 And that feature of the decree was bound to be of particular interest to any historian whose work tracked the Athenians' documentary habit; in the course of compiling major decrees, he could be expected to note the arrangements for deleting relevant records, along with the very exception that preserved the decrees against Antiphon and Phrynichus (treated by Craterus, FGrHist F 5 and F 17).
Another complication emerges just as P broadens the scope (at the end of §77) to those who are indebted and disfranchised because of recent or pending decisions; here P includes a group whose penalties are not yet copied into official records, ‘up until the (current) council leaves office, when Callias was archon’, εἴ τις μὴ ἐξεγράφη, μέχρι τῆς ἐξελθούσης βουλῆς ἐφ’ ἧς Καλλίας ἦρχεν. Of course, the archon date rings true, and, as Hansen points out, ‘until council leaves office’ has parallels in the epigraphic evidence.Footnote 17 This way of indicating the time period, distinguishing between the council year and the archon year, would have been sensible and important in the situation addressed by this measure. But we might have expected the archon date and similar programmatic details in the prescript. Instead, it looks as though the historian introduced the measure with a comment on its timing, and the editor wrote that into the document, however awkwardly.
So I suggest that P is reconstructed from a source that summarized the context and effect of the measure while focussing on the historic adaptation, the wholesale deletion of official records; he may have quoted excerpts but did not provide a full copy of the decree. What is missing points to a version of this type. Thus in P we find nothing corresponding to Andocides' second category of ἄτιμοι (§74), those whose persons were fair game (τὰ μὲν σώματα ἄτιμα) but whose property remained secure, such as those guilty of wrongdoing in office (κλοπῆς ἢ δώρων) or in military service (ἀστρατείας ἢ δειλίας, etc.). The source historian may have considered such persons obvious candidates for amnesty, needing no particular notice. So, too, where Andocides gives ample scope to targeted ἀτιμία (κατὰ πρόσταξιν), barring particular rights or remedies for those who served under ‘the tyrants’ (§75), P mentions only the ‘names of the Four Hundred inscribed or anything recorded anywhere regarding events in the oligarchy’ (§78). Conversely, in regard to the critical exception (πλὴν ὁπόσα ἐν στήλαις γέγραπται τῶν μὴ ἐνθάδε μεινάντων, etc.), Andocides says merely (§80) that exiles were not allowed to return. In all, P’s version seems adapted from another summary, quite independent of Andocides' description; and that summary gave particular emphasis to the cancellation of records, even including ordinary fines and debts to the polis that had nothing to do with civil conflict.Footnote 18
That emphasis may explain why P begins by acknowledging the immunity (ἄδεια) for proposals regarding ‘those indebted’ (περὶ τῶν ὀφειλόντων). In modern editions Sauppe's supplement, περὶ <τῶν ἀτίμων καί>, has been largely taken for granted: permission for proposals regarding ἄτιμοι seems essential, and so, if the document is a close copy, that phrase must have been simply lost in transmission.Footnote 19 But the focus of P is the cancellation of all relevant records of debts and penalties—not just the decrees against those disfranchised for their role in the oligarchy which Andocides emphasizes there (§§76, 80) and thereafter (§§106–9). Of course, those indebted were also disfranchised unless and until the debt was paid, and that is implicit in the further catalogue of erasures. The first category to be expunged (§77)—those listed with πράκτορες, with ταμίαι of the gods or with the ἄρχων βασιλεύς—belong to a group whose ἀτιμία derives from debt or fine; and the second group (§78)—‘those under judgment from accountings’, etc.—also face ἀτιμία for penalties unpaid.
Thus P—the document as we have it—seems to rely on a source in which the historic adaptation was emphasized: most notable is the scope of the cancellation, not just (as we might expect) erasing the decrees against those complicit in the oligarchy but even against those who might be prosecuted because of old debts or outstanding penalties of any sort. Such a proposal almost certainly required immunity to address the status of ἄτιμοι generally and not just debtors, but the compositor of P reconstructed the clause about ἄδεια from the particular emphasis of his source.
III. THE DECREE OF TISAMENUS (T) AND THE ‘SCRUTINY’ OF LAWS
The editor who inserted T (§§83–4) probably construed it as an introduction to the restoration laws that follow (§§85–8), despite the disparity between it and the decree that Andocides described. Those fragments from the restoration laws were obviously framed to match the orator's references, not presented as close and complete copies of the law in question. For example, in §87 the compositor has lumped together two disparate provisions: the rule that decisions from private suits and arbitrations under democracy remain valid, alongside the statute of limitations that Andocides relied upon, the rule that criminal procedures (γραφαί, ἐνδείξεις, and so on, listed in §88) apply ‘from Euclides’ archonship’. A more likely corollary to the rule that decisions under democracy remain valid is adduced in Demosthenes 24 (Timocrates) 56: all decisions under the Thirty are invalid.
For T, the measure that precedes those garbled laws, MacDowell offered a detailed defence; Canevaro and Harris reject it for defects both of form and of content; but Hansen has now offered an attractive theory to account for most of the disparities.Footnote 20 Yet, whatever we make of it, T is almost certainly the wrong material, not the same provisions that Andocides invoked.Footnote 21
Andocides introduces T in evidence of the confirmation or ‘scrutiny’ (δοκιμασία) of old statutes that seems vital to his defence; for, in conclusion to this scrutiny of old laws, a statutory limit was put upon them, and it is that limitation that Andocides relies upon. It is not enough to show that Isotimides' decree against him is no longer valid (for reasons that may seem sophistic), or that he is innocent of the original charges (a plea that is certainly disingenuous); he has to prove that he cannot be prosecuted for the original offence because of the rule that criminal procedures—including the ἔνδειξις against him—cannot reach back to events before the year of Euclides (§§88–9). To put that rule in its proper context, he describes how ‘legislators’ (νομοθέται) were chosen to carry on the reconstructive process and only then was it recognized that traditional remedies would have to be limited ‘from Euclides’, if the amnesty was to be a reality.Footnote 22
Instead of that scrutiny of old laws, T describes a procedure to draft and approve ‘whatever is needed in addition’ to the Solonian corpus.Footnote 23 Andocides alludes to such legislation, but that is not the process that interests him here. He tells how the transitional committee of the Twenty was formed to manage the city, while or ‘until laws were enacted’; one might suppose that the laws-to-be-enacted were fundamental rules recognized by the amnesty agreement, the very laws that he outlines after the reading of T (§§85–8). But their mandate also included the restoration of old laws that had been abrogated or altered under the Thirty, and it was realized in the course of that reconstructive process that the old laws for public remedies would pose a threat to many citizens; therefore, rather than amend them piecemeal, the amnesty law was introduced.
Thus, as Andocides tells it, the need for scrutiny of old laws was not recognized in the original mandate for legislation but was ‘discovered’ by the legislators. Rather than relinquish the document that seems to contradict him, scholars have routinely discounted Andocides' argument.Footnote 24 But he is not likely to have invented that turn of events. At trial, if his ‘scrutiny’ were not plainly indicated in the decree he called for the clerk to read, it would gravely weaken his case; for this point of law is the very crux of his defence. And yet, we know that he won acquittal, though resentment remained strong against him. So it seems to me more reasonable to recognize that the document is not a good match. The editor went looking for something to represent the decree that Andocides called for, the one that set in motion a δοκιμασία of old laws; but he could only find a version of the legislation law that provided for addenda to the Solonian laws and mentioned a ‘scrutiny’ of that pending legislation.
On any straightforward reading, T is not part of the process to restore old laws; it prescribes a new legislative protocol for drafting and approving ‘whatever is needed in addition’.Footnote 25 But, in order to show that T is not only the wrong document but also a fraud, Canevaro and Harris insist that Andocides' ‘legislators’, the νομοθέται whose finding led to the ‘scrutiny’, were not the expert law-writers that MacDowell and most commentators have supposed but a preliminary session of citizen lawmakers, who convened before the legislation law was passed.Footnote 26 By their reading it must be this body of Athenians whom Andocides addresses in second person; so it was this assembly of the people that ‘discovered’ whatever problem led to the scrutiny of laws. Thus Andocides' explanation is simply ‘tendentious’, and what he describes as a ‘scrutiny’ of old laws is nothing other than the examination of the corpus that had been ongoing since 410. But Andocides regularly credited ‘you (Athenians)’ when referring to decisions of court or council rather than the assembly.Footnote 27 So MacDowell's reading seems at least as plausible: those who ‘discovered’ the need to survey old criminal laws and set limits on them were the ἀναγραφεῖς, newly reappointed by council, whom Lysias also refers to (however ironically) as νομοθέται; among those law-writers were Nicomachus and Tisamenus.
Now, T calls for a second body of νομοθέται, who enacted legislation that the first νομοθέται had drafted; we might expect Andocides to mention that second group as well, if they had anything to do with the review of old legislation that he calls a ‘scrutiny’. But there is nothing in Andocides' account to suggest the two distinct bodies of νομοθέται that T recognizes. Instead, Andocides calls for a decree requiring that ‘all the laws be scrutinized’, and that those confirmed in this process be inscribed in the stoa (§82); and thereafter he refers to this scrutiny as a reauthorization of the laws for public remedies (ἔνδειξις, etc.) subject to the limitation ‘from Euclides’ (§§88–9). But T begins (§83) by reaffirming that traditional laws and legal standards remain valid, only to introduce the procedure for ‘whatever is needed in addition’.Footnote 28 Those addenda are to be drafted by the law-writing νομοθέται who have been chosen for that duty by council (and whose names were listed in the original, if the MS οἵδε is right); their drafts are posted for public notice and given to the officers to enforce. So far we have nothing that bears upon the ‘scrutiny’ at the focus of Andocides' description.
Then, in §84, the detail is introduced that gives the document its tenuous relevance:
τοὺς δὲ παραδιδομένους νόμους δοκιμασάτω πρότερον ἡ βουλὴ καὶ οἱ νομοθέται οἱ πεντακόσιοι, οὓς οἱ δημόται εἵλοντο, ἐπειδὴ ὀμωμόκασιν. ἐξεῖναι δὲ καὶ ἰδιώτῃ τῷ βουλομένῳ, εἰσιόντι εἰς τὴν βουλὴν συμβουλεύειν ὅ τι ἂν ἀγαθὸν ἔχῃ περὶ τῶν νόμων. ἐπειδὰν δὲ τεθῶσιν οἱ νόμοι, ἐπιμελείσθω ἡ βουλὴ ἡ ἐξ Ἀρείου πάγου τῶν νόμων, ὅπως ἂν αἱ ἀρχαὶ τοῖς κειμένοις νόμοις χρῶνται. τοὺς δὲ κυρουμένους τῶν νόμων ἀναγράφειν εἰς τὸν τοῖχον, ἵνα περ πρότερον ἀνεγράφησαν, σκοπεῖν τῷ βουλομένῳ.
The laws ‘being submitted’ to the officers (τοὺς δὲ παραδιδομένους νόμους) shall first be ‘scrutinized’. That δοκιμασία is then explained with details that prove problematic: along with the council, this vetting involves another body of νομοθέται, whom the manuscript identifies as ‘the five hundred whom the demesmen chose’, and who are to serve ‘when they have sworn’ (ἐπειδὴ ὀμωμόκασιν). Then T empowers the Areopagus council to assure that the current officers apply the laws recognized in this process. Finally, the laws approved in scrutiny are to be inscribed or written up ‘on the wall, where they were written up earlier, for anyone willing to examine’ (ἵνα περ πρότερον ἀνεγράφησαν, σκοπεῖν τῷ βουλομένῳ). The first and the last provisions (scrutiny and republication) suggest a match with Andocides' description, but they are obviously not fabricated from it.
The ‘scrutiny’ that T describes in §84 seems rather reminiscent of the old legislation law that Demosthenes invokes against Leptines (20.90–3). In the 350s Demosthenes has to explain an obsolescent procedure in which new laws must be confirmed in two stages, and any conflict with standing law is decided by those ‘who have sworn’ the dikastic oath.Footnote 29 To justify this complication, Demosthenes points to the parallel scrutiny of θεσμοθέται; for (in his day) they were examined for fitness first in council and then in court.Footnote 30 Indeed, the legislation law that Demosthenes invoked probably referred to the two-stage legislative process as a δοκιμασία, and that term prompted him to draw the obvious analogy: judging the fitness of laws is much like the scrutiny of officers. Variations on that theme lie behind both the scrutiny that Andocides envisioned and the scrutiny reconstructed in the latter part of T. But, where Andocides' scrutiny envisioned a confirmation of old laws without trial, T drew upon a source that described an important adaptation for new laws or those involving substantive amendments: in order to settle any conflict, these bills would be confirmed or rejected by a jury of (or including) those who had sworn the dicastic oath.Footnote 31
The scholia show that the old legislation law caused some confusion, but the more discerning commentators understood Demosthenes' analogy: judicial review of laws was parallel to the two-stage scrutiny of archons. That is, after examination in council, there is a hearing in court. And in this connection, as the lexica and better-informed scholia explain, there were two sets of personnel called νομοθέται: those that introduce the laws and those that judge or decide upon them.Footnote 32 Similarly, the compositor of T seems to have found in his source an explanation along the lines of Demosthenes' old legislation law. And so that compositor would naturally identify the second body in T as νομοθέται, aptly enough, though the term itself may be a bit premature.Footnote 33
But the document T could not be simply fabricated from Demosthenes' description, any more than it matches Andocides'. None of the procedure in T matches the δοκιμασία that Andocides described; his ‘scrutiny’ was an ad hoc procedure for reauthorizing old laws, not a template for new legislation ‘needed in addition’. And that disconnect helps to explain the last disparity: where Andocides says that the laws confirmed in scrutiny were inscribed at the stoa (§82), T says that laws in the process of ‘being authorized’ (κυρουμένους) shall be written up ‘on the wall where they were written up previously’ (§84). The latter probably refers to temporary publication, pending permanent inscription.Footnote 34
IV. DEMOPHANTUS' DECREE (D) AND ANDOCIDES' ‘LAW OF SOLON’
With D we have a more complicated problem: again we have to weigh the document against Andocides' description, but we also have testimony from Lycurgus nearly seventy years after Andocides' trial (Against Leocrates 124–7). Canevaro and Harris suggest that Lycurgus' description both inspired D and exposes it as a forgery. Sommerstein argues persuasively that D could not have been concocted from what Lycurgus says, but he has to agree that D is the wrong document, not the measure that Andocides called for at this point in his argument. Now, Hansen defends both its authenticity and its relevance: D is based on reinscription, after the fall of the Thirty. Some of the most awkward difficulties can be resolved in this way, but, it seems to me, there are still three critical disparities, where D is substantially at odds with the orators' comments. These features have not been straightforwardly explained but may yet tell us something about the editorial process that brought this document into the manuscript: (1) in D we find a ‘starting date’ that seems oddly placed and pointless; (2) Andocides seems to invoke old law, whereas D and Lycurgus refer to measures of the decade before Andocides' trial; and (3) Lycurgus drew upon provisions that encompassed ‘treason’ as well as subversion, with important implications for pre-empting such plots, none of which is indicated in D. Let us leave the first crux (the starting date) for last, until we have some context for it, and begin with the basic problem of old law in new settings.
Andocides refers to a ‘law of Solon’, which stood in front of the council chamber, for the rule that any collaborator who serves in office after democracy is overthrown can be slain with impunity.Footnote 35 The document D is a decree drafted by Demophantus and datable from the prescript to the first prytany of the restored democracy in 410. The first provision corresponds to Andocides' point of emphasis (in italics):Footnote 36
Anyone who overthrows the democracy at Athens, or holds office when the democracy has been overthrown, shall be an enemy of the Athenians, to be slain with impunity. His property shall be confiscated with a tenth to the goddess. Both he that kills the offender and he that conspires [to kill him] shall be blameless and untainted.
The bulk of the text then prescribes an oath for every Athenian to swear, ‘convening by tribe and by deme’ (κατὰ φυλὰς καὶ κατὰ δήμους),Footnote 37 against anyone who acted or collaborated in the overthrow of democracy. The oath includes provisions for disposing of property and rewarding defenders of democracy (points of emphasis and disparity are italicized):Footnote 38
(§97) ‘I shall kill by word and deed, by ballot and by my own hand, if I am able, anyone who overthrows the democracy at Athens, or anyone who holds office when the democracy has been overthrown hereafter; and [likewise] if anyone should set up a tyranny or collaborate with the tyrant. And if another kill [the offender], I shall hold him blameless before gods and divinities, as the killer of an enemy to the Athenians; and the property of the slain [offender], I shall repay half to the killer, disbursing it all and withholding nothing.
(§98) But if anyone die in killing such an offender, or in the attempt, I shall reward him and his sons, just as Harmodius and Aristogiton and their descendants [are rewarded]. All oaths sworn against the Athenian dēmos, at Athens or in camp or elsewhere, I rescind and renounce.’
Now, for Lycurgus, in the case against Leocrates, the treatment of traitors is the point of emphasis. Contrary to D and to Andocides' comment, he calls for a reading ‘from the stele in the council chamber’, not in front of it, ‘regarding traitors as well as those who overthrow the dēmos';Footnote 39 he dates that measure after the fall of the Thirty (in 403/402, not 410); and he evidently found in it an oath to hold guiltless anyone who kills a traitor as well as the killer of a would-be tyrant or subversive. It is only then, after recapping this material from the stele (§126), that Lycurgus refers to the decree of Demophantus by name and insists that here ‘you (Athenians) have sworn … to kill the one who betrays the fatherland, by word and deed, by hand or ballot’ (§127).
In D Demophantus is identified as one who ‘composed’ or drafted the text (συνέγραψεν); that description of the task suggests that he was chosen as συγγραφεύς to draw up a new formulation based on old prototypes.Footnote 40 That would be the most likely way to frame such a measure adapting old rules to the situation at hand: rather than leave it to anyone willing, the council appoints or approves draftsmen with special competence. If so, the material on the stele would include old ‘Solonian’ rules that were probably acknowledged as such, before the new prescription that was formulated on that basis. From this perspective there is, indeed, no conflict between Andocides' reference to Solon's law and the measures that D and Lycurgus attribute to Demophantus: the text on the stele that Andocides read in 400 was part of a συγγραφή that recognized the Solonian antecedents.
But Lycurgus focussed upon ‘treason’, προδοσία, betraying the city, the fatherland or the dēmos, and that crime is nowhere addressed in D (nor in Andocides' comment).Footnote 41 Of course, Lycurgus may be taking liberties with the law in order to make it fit his rather dubious case against Leocrates,Footnote 42 but he is not making up treason out of whole cloth. That particular focus was an almost inevitable feature of the anti-subversion measures of 410 (such as D represents): for the charges against the leaders of the Four Hundred (Antiphon et al.) had to be framed in terms of treason rather than tyranny. Conversely, the main strand of anti-subversion legislation does not specify προδοσία: thus Eucrates' law, barring the Areopagus council from collaborating (Rhodes and Osborne 79), begins by restating the rule that anyone who joins in subverting the democracy can be killed with impunity, and the main provision reads like a corollary to the rule that Andocides emphasized (against anyone who serves in office after democracy is overthrown), but there is no mention of προδοσία.Footnote 43
Moreover, Lycurgus puts his greatest emphasis upon pre-emptive action, which is not recognized in D. Now, as we noted, the later orator, in dating his document to the aftermath of the Thirty, may have been simply misled by a decree for re-inscription in the year of Euclides. But it is at least as likely that the restored democracy, in that year or thereafter, framed its own adaptation of the old commitments against tyranny, treason and subversion. For Lycurgus puts a particular emphasis on material that seems to have stood at the top of his readings from the stele, a text that recognized a pre-emptive licence, to act upon suspicion, before the plot proceeds:
(§125) They voted and swore, if anyone should try to set up a tyranny or betray the city or overthrow the dēmos, one who perceives it (τὸν αἰσθανόμενον) and kills [the traitor or subversive] is guiltless (καθαρός). … They thought that citizens must live their lives so as to come under no suspicion of these crimes. Read me the decree. (§126) These provisions they wrote on the stele and set it up in the council chamber, as a reminder to those who convene there each day and deliberate for the fatherland, how they must act toward such men. Therefore, they swore, if anyone perceive those who are merely about to do any of these acts, to kill them, for good reason. For against all other crimes the retribution must be assessed after the fact, but for treason and overthrow of the dēmos [it must be] earlier: if you lose this critical moment at which they are about to commit some infraction against the fatherland, it is not possible afterward to exact justice from those who commit the crime; for they are already stronger than retribution from those who are wronged.
Lycurgus seems to make rather much of the patriotic principle, that good citizens should get the jump on insidious characters, but the document that was read to the court must have allowed for that interpretation. We might at least expect some wording that corresponds to the element of discovery (τὸν αἰσθανόμενον, 125; ἄν τι αἴσθηται μόνον μέλλοντας, 126).Footnote 44 In the document D, of course, there is nothing to that effect, and in view of that disparity it seems to me all the more certain that Lycurgus' reading from the stele, with its emphasis on προδοσία, and the document in Andocides' speech, which never mentions it, cannot be the same material.
There is perhaps another sign of that mismatch in the section that follows (§127). For it is only here that Lycurgus identifies Demophantus’ decree, and he may be turning to another measure in the series. In the earlier sections (§§124–6) he was describing an oath, prescribed by decree, that stood in the βουλευτήριον for the guidance of those who convene there; now (§127) he speaks more generally of the ‘reminders and paradigms that you have before you, defined in the decrees regarding the wrongdoers’. It is only in that broader context that he cites Demophantus’ decree and summarizes the oath ‘to kill one who betrays the fatherland, by word and deed, by hand and ballot’. And he gives that commitment a rather different interpretation than he put upon the reading from the stele in §§125–6 (in the council chamber, for the guidance of those who deliberate there). Here (§127) he insists that judges and fellow citizens have inherited the obligation that their forebears swore to: ‘Don't think that you are heirs only to the property that ancestors bequeathed and not also to the oaths and the pledge that your fathers gave the gods as surety for the city's shared prosperity.’
In any event, whether it refers to the same oath or another variation on the same theme, there is nothing in Lycurgus’ version to match the provision in D that anyone who holds office after democracy is overthrown is subject to the same outlawry as those who stage the coup. That was the whole point of the citation for Andocides, and indeed it is duly repeated in D, in the initial statement of the rules (§96) and again in the oath (§97) to enforce them ‘by word and deed, by ballot and by my own hand’. Canevaro and Harris are right to be sceptical; but, rather than suppose that the document is an utter fabrication, we can better account for the confusion by supposing that the editor who brought this document into the manuscript drew upon an excerpt or abridged version, and he inserted that clause against collaborators into the oath, where it may not belong.
It was probably Lycurgus’ testimony that prompted the editor of the speeches to search for a decree of Demophantus (whom Andocides does not mention). The text that he found may have been simply an excerpt focussing on the oath, and it would not be surprising to find that the editor simply added what the context seemed to require. For we have a suggestive parallel in the same collection of speeches (Burney 95 = codex Crippsianus), where Lycurgus called apparently for the ‘Oath of Plataea’ (Leocrates §81).Footnote 45 Here the editor evidently took the oath as quoted in Ephorus (Diod. Sic. 11.29) and added a vow to dedicate a tenth of the plunder, which he found in the framing narrative (Diod. Sic. 11.3.3 and 33.2). Andocides’ Mysteries and Lycurgus’ Leocrates are the only speeches in this codex where substantial documents have been inserted;Footnote 46 and, it seems reasonable to suppose, the editor(s) processed these two patriotic documents in the same way.
Now, let us turn back to the ἄρχει χρόνος clause, the ‘starting date’ in D for the rules against subversives and collaborators. Lycurgus says that his reading from the stele was actually an enactment of 403, but D begins with an elaborate prescript that dates the measure emphatically to the first restoration in 410.Footnote 47
It was decided by the council and the people; Aiantis was in prytany, Cligenes was secretary, and Boethus presided; these provisions Demophantus drafted. The time for this decree begins (ἄρχει χρόνος) with the council of five hundred chosen by lot, for whom Cligenes was first secretary.
As a statutory limit the ἄρχει χρόνος clause seems pointless, and the secretary's name is misspelled in the manuscript (both times, Κλεογένης for Κλειγένης). For Canevaro and Harris, this muddling is further proof that D cannot be genuine. But, in my view, the misspelling counts for little and the process required to produce a complete ‘forgery’ seems rather unlikely: a researcher of some sophistication patched together what Lycurgus dated to 403 with an historical reference to that first prytany of the democratic regime in 410. It seems more economical to suppose that the editor found the document in its proper context but not in one piece: the source historian provided the prescript and noted that a starting date was added when the law was reauthorized. That is in line with Hansen's conclusion: D is drawn from the version of Demophantus’ decree that was inscribed in 403/402, to replace what the Thirty had deleted. The starting date, to specify that the measure applies from that first prytany of the new regime in 410, was added with reinscription.Footnote 48 In the prescript we are told that Demophantus composed or drafted the measure (συνέγραψεν rather than εἶπεν), and that would aptly describe the bill of 410 (where συγγράφειν means drawing on old models for new enactment). Those features, the law-writing process and the re-inscription, were bound to be of interest to any historian who focussed on the publication and preservation of the people's decrees; so it should not be surprising that a source who gave only the movers’ names in the other instances gave the full prescript for D.
But the epigraphic parallels are not reassuring, and Canevaro and Harris are right to object. Ordinarily a starting date (χρόνος ἄρχει) is closely linked to particular provisions, in leases and the like, specifying the time frame for particular obligations (when to pay, take possession, and so on).Footnote 49 There is no instance where the starting date comes after the prescript and governs the whole decision, as it seems to do in D. But I think that the disparity between the epigraphic evidence and the awkward formulation in D provides another clue to the way in which the documentary insert was framed.
As a statutory limit upon the full content of D, both the initial restatement and the oath based upon it, the starting date would have no legal or practical effect at re-inscription: the old rules against tyranny and subversion were bound by the comprehensive amnesty law, limiting legal remedies ‘from Euclides’ archonship’ (just as Andocides insists). In other words, even without the starting date, the rule against collaborators was unenforceable. It is all the more pointless with regard to the real burden of D: the new formulation of old rules in an oath that all Athenians were to swear on that singular occasion in 410, convening by deme and tribe in a commitment to the restored democracy.
By contrast, for many decrees the secretary is noted at the top of the text, before the prescript (where he is named again), and sometimes the archon is added to that heading.Footnote 50 And we learn from Diocles’ law (Dem. 24.42) that a systematic notation was added as the new regime put its law books in order after 403: the secretary is to specify, for old laws and new, whether they are valid from enactment or from some starting date.Footnote 51 That bit of book-keeping would explain how and why an otherwise dubious detail was added. The problem is its placement: it should stand at the top of the text. So again, I suggest, the source historian may have commented on that complication in dating this measure, and the enterprising editor put that detail in the document where he thought it belonged.
In sum, it seems reasonable to conclude with Sommerstein, in agreement with Canevaro and Harris on this point, that D is the wrong document, either the wrong decree altogether or not the excerpt that Andocides called for. Of course, Hansen's solution is also on the right track: Andocides may, indeed, refer to Demophantus’ decree as a law of Solon, and Lycurgus’ confusion on the dating can be easily explained (if he was confused) on the assumption that Demophantus’ decree was re-inscribed in 403/402. But the divergent content cannot be so easily dismissed: neither D nor Andocides gives any indication that treason, προδοσία, was among the grounds for outlawry, yet in a decree of 410 it seems almost inevitable, as it was the one viable charge against the oligarchic leaders.Footnote 52 Again, neither in D nor in Andocides is there any sign of the pre-emptive licence that Lycurgus found in his reading from the stele. In fact, it seems reasonable to question whether the one provision that Andocides fixed upon—the old rule against officers who serve when democracy is overthrown—was included in the oath prescribed by Demophantus. If the document that Andocides cited did contain the oath as we have it in D, we would expect the orator to make something of it: not only would Epichares be fair game for anyone to kill with impunity (if not for the limitation ‘from Euclides’), but also every Athenian who swore that oath would be bound to do so. Instead, I suggest, the editor who reconstructed D inserted that clause in the oath, just as he or his colleague augmented the ‘Oath of Plataea’ in Leocrates.
By this reading, D is perhaps a ‘forgery’ in the minimal sense of ‘a document pretending to be something it is not’;Footnote 53 it is not the material that Andocides called for. But, call it what we will, D involved some editing: for the ἄρχει χρόνος clause comes awkwardly at the end of the prescript, whereas, if the limitation was added later, in accordance with Diocles’ Law (Dem. 24.42), we would expect it to be simply an annotation at the top of the text. If Craterus or another historian drew upon the later version, rather than quote the full text verbatim, he probably introduced it with a note explaining that complication, that it was enacted immediately after the first restoration of democracy and was later subject to limitation after the second restoration. It was then the ‘forger’ or enterprising editor who reconstructed the text rather awkwardly, putting the time limit after the prescript as though it belonged to the original decree.
V. CRATERUS’ ON DECREES AND ‘LATENT FRAGMENTS’
In contrast to the reconstructive model that I have outlined in the previous sections, the prevailing documentary model assumes that these three decrees are presented as close copies from the text of record: some researcher undertook or a forger pretended to transcribe those decrees from a documentary source, from an historian who preserved the text if not from the archives or inscriptions. That documentary model is supported elsewhere, in a few instances where the author or editor did indeed endeavour to copy closely a substantial text that matches the official record. Such, for instance, are the excerpts from the homicide law in Demosthenes 23 (Aristocrates). But in that case the authentic documents indicated in the stichometry may have been copied from a crucial attachment, the set of laws in conflict with the targeted decree (παραγεγραμμένοι) that were written up on a placard (σανίς), for visual reference in court, and probably preserved with early copies of the speech.Footnote 54 A better comparandum is perhaps to be found in Demosthenes 43 (Macartatus) 57, where an excerpt of Draco's law was added in the later, post-stichometric tradition. But in that case, as well, the source law was well known and readily accessible, whereas the documents in Andocides were drawn from more obscure materials.
For the decrees in Mysteries, the presumptive source is Craterus’ work on decrees, as some of the testimonia suggest that his work was a collection of texts (ψηφισμάτων συναγωγή), copied in full, and one of the fragments seems to confirm that testimony. On that basis, scholars of the nineteenth century readily supposed that Craterus was responsible for many of the more reliable excerpts that found their way into speeches and histories without attribution: such were the fragmenta latentia that Krech collected.Footnote 55 Jacoby discounted most of those latent fragments, but he regarded Craterus with much the same reverence; and in more recent work the picture of a meticulous archivist often seems unquestioned—whether we are to assume that he relied on inscriptions or on the archive in the Mētroon.Footnote 56
But aside from a few grand pronouncements in late testimonia, there is little to indicate that Craterus dutifully copied official records complete and verbatim. One third of the fragments contain nothing more than a few names from the outer reaches of the tribute lists.Footnote 57 In all, about half of the fragments cite Craterus only for particular details, often at odds with other sources. Notably in F 16, on the decree(s) against Diagoras, Craterus is cited only for context, while Melanthius (FGrHist/BNJ 326 F 3) is said to have presented an actual copy (ἀντίγραφον); apparently Craterus did not replicate the decree but explained the grounds for it.Footnote 58 In most of the major fragments he is credited or faulted for particular, often divergent, details. Thus in F 10, on the decree to grant citizenship to the Plataeans, Craterus is cited only for his silence on whether or how they were represented at the Stoa Poikile. In F 14, on the decree against Arthmius, Craterus is cited only for Cimon's authorship. Again on the decree of Cannonus (F 15), Craterus is cited merely for a procedural note (use of the water-clock). Moreover, in F 17, Craterus is cited along with Didymus by a scholiast explaining that the riddle in Ar. Lys. 333 refers to Phrynichus (‘Who among the generals on Samos would lend a hand with the firewood?’). The comment on comedy was probably added by Didymus, but the point of reference conforms to the pattern: whatever version Craterus gave for the decree against Phrynichus, already in the first century b.c., Craterus is cited for salient features, not for close and complete copies.
One might yet suppose that Craterus himself provided full text and a later epitomizer singled out the salient details that feature in our fragments. But Plutarch seems to have had access to the original work, and his ambivalent treatment confirms what the other fragments suggest, that Craterus usually provided text from each decree, but not always, and he also referred to other sources and supporting materials.Footnote 59 Thus, from Plutarch's comment and the preponderance of the fragments we should conclude that Craterus’ work On Decrees was not a collection of transcripts.
Craterus seems to have followed in the path of the Peripatetics who wrote on legal and constitutional matters, and in their published work they regularly summarized and excerpted documentary material.Footnote 60 Thus, Theophrastus in Book 4 of On Laws (F 2 Szegedy-Maszak = 636B Fortenbaugh) gave a summary of the grounds for εἰσαγγελία in phrasing that seems largely authentic, and Craterus in his account of the case against Themistocles (F 11a) evidently concurred with Theophrastus.Footnote 61 That is, Craterus seems to have acknowledged the list of offences that his predecessor had given and identified the crimes of Themistocles among them. Craterus, of course, also identified the mover of the decree against Themistocles (Leobotes, son of Alcmeon). But it was the link with Theophrastus that made his testimony worthy of note for the lexicographer: he preserved this information precisely because Craterus and his predecessor agreed, contradicting Caecilius’ view that εἰσαγγελία applied ‘against new and unwritten wrongs’ (κατὰ καινῶν καὶ ἀγράφων ἀδικημάτων).
Craterus probably made the same sort of contribution in the one fragment that supposedly conveys close and complete copy: F 5, the decree and verdict against Antiphon et al. For the (apparently) full text that ps.-Plutarch preserves is itself a ‘latent fragment’, not attibuted to Craterus by name. To be sure, in F 5a Harpocration cites Craterus but only for naming the mover of the decree (Andron); then in F 5b ps.-Plutarch provides the complete text (or nearly complete), for which he cites Caecilius (F 2 Woerther). The usual assumption is that Craterus gave the original text in full and later tradition left the lexicographer with only the mover's name; but the biographer indicates quite plainly that Caecilius is his source for this and other full-text documents.Footnote 62
To be clear: there should be no question but that Craterus gave this document (F 5) fairly substantial treatment; he seems to have treated the events of 411–410 in particular detail. Presumably this text was preserved by the exception in Patroclides’ decree (as represented in Andoc. 1.78): most of the punitive decrees prior to 405 were removed from the public record and all copies, official and privately held, were to be deleted; but the inscribed decrees against Antiphon, Phrynichus and others were spared by the provision regarding unpardonable crimes. Yet, if Craterus provided the full text of the inscription, it is surprising that ps.-Plutarch cites only Caecilius and not also Craterus. Indeed, Caecilius probably would not have bothered to produce the text at such length, if he were not correcting or improving upon a predecessor whose version he found inadequate.
Inspired by the likes of Caecilius, perhaps, the editor or ‘forger’ who brought the three decrees into Mysteries endeavoured to improve upon the abridged documents in his main source. In this light, his aims and methods seem rather different from those of the forgers who padded Demosthenes’ masterpiece On the Crown. In the latter it often looks as though documents were concocted in rhetorical schools for exercises where it mattered very little whether the inserts were at all accurate.Footnote 63 But in Andocides’ speech the documents seem to be included precisely for historical detail, not for rhetorical effect or for an entertaining rendition.
To satisfy that historical interest, evidently, the editor of Mysteries drew upon sources such as Craterus’ On Decrees, and he tried to reconstruct the formalities (as in the prescripts and instructions for publication) from the context or summary that the historian gave. These inserted materials were not props or fillers to give a sense of how documents punctuated the proceedings, and they do not much resemble the laws and decrees that were framed for hypothetical cases for students of rhetoric to wrestle with. The inserts in Mysteries would rather appeal to readers who took an interest in the background to this peculiar case and would appreciate documentary evidence that bears upon it. That process is perhaps akin to the ‘forgery’ that Canevaro and Harris find, but it seems quite unlike what we normally mean by that pejorative term. After all, ‘forgery’ involves fraud, but this material often lacks what an editor could easily have invented, if his aim was to convince a credulous clientele.
Thus, in the denunciations early in Mysteries (§§13, 15, 35) the editor seems to have reconstructed what he could find, without any attempt at recreating authentic-looking documents. He has the first names of persons implicated in 415 whom Andocides does not name, and, without inventing patronymics or plausible formalities, he frames them to fit the context (ὁ δεῖνα ἐμήνυσε).
The same method may account for infelicities in the major documents, but in these there are also substantial disparities between the material that Andocides called for and what we find inserted in the text. For Patroclides’ decree, the document P glosses over essential features of the amnesty to focus on the list of records deleted. For Tisamenus’ decree, there is nothing in the document T corresponding to the reauthorization of old laws, subject only to the limitation ‘from Euclides’, that Andocides relied on in his argument; T prescribes a procedure for new measures or substantial amendments. For Demophantus’ decree, we find in D a restatement of the law against tyranny and subversion as preface to an oath that Andocides does not mention and probably would not have read to the court. All three inserts show a particular interest in problems that would loom large for any historian trying to describe the evolving body of decrees by which Athens and her empire were governed in the fifth century. And thus the most significant defects seem to emerge from a process just the reverse of what is often assumed: each decree is not a close copy that deteriorated in transmission but a reconstruction from the excerpts and context that Craterus or another source historian supplied.
