II. The epigraphical comparanda

The clauses describing Areus’ letter and reporting its delivery are likelier to be archival, given some epigraphical comparanda that have not, with one notable exception,Footnote ¹³ been noticed in this connection. These comparanda can be found in contemporary inscriptions from five stelae.

Inscription no. i, now at the Louvre, contains two documents: a letter of 182 BCE written by Eumenes I of Pergamum inviting participation in a new festival, the Nicephoria, and a decree of some Greek city (probably Iasos) in response.Footnote ¹⁴ Between the two documents is a short note reporting the delivery of Eumenes’ letter and its date. In synopsis:

This short delivery notice is clearly not part of either document: Eumenes could not have known the specific day on which his letter would be delivered, and the civic decree that follows begins with its own formulary prescript, which included a date. Megon’s delivery and its date is its own separate datum. The action that is described must have been common (all documents received by a city had to be delivered by someone) so, too, it seems, was the recording of such deliveries. What is unusual in this inscription is that this detail was included when these documents were being prepared for public display. In this specific case, a motive might be guessed: in his letter, Eumenes had named two couriers who were to deliver his letter, Megon of Ephesus and Kalas of Pergamum; only Megon had in fact completed his assignment, and the discontinuity between what Eumenes had described and what actually happened motivated someone to take the unusual step of naming Megon as the sole courier.

Inscription no. ii contains a decree of an unidentified city honouring a foreign judge (δικάστης) who had come from Cos to adjudicate some dispute there. Once he had completed this task, a decree was passed in his honour and forwarded to Cos, where it was inscribed. Again, several supplementary clauses were added (the superscript letters are added to facilitate discussion):Footnote ¹⁵

Three details are included here. The first (‘a’) reports the names of envoys who had been appointed in a process prescribed by the decree (lines 10–21). Once that selection process had been completed and the individuals chosen, an addendum recording their names was appended. Such appointment riders seem to have been standard practice,Footnote ¹⁶ and this one was part of what was sent to Cos. The next two clauses reflect what happened after it arrived there: ‘b’ records the date by month and day (but not the year) on which the aforementioned envoys delivered their decree, which is reminiscent of Megon’s delivery in no. i. In ‘c’ the public seal of the sending city is described. (More on this below.) Again, this note was added in Cos and reflects a Coan process.

Inscription no. iii was found at Astypalaia in the Dodecanese. It contains two decrees of Smyrna dating to the second century BCE; the first honours Astypalaia and the judges (δικασταί) whom it had sent to Smyrna; the second decree honours an Astypalaian official who had facilitated the matter:Footnote ¹⁷

Both of these decrees had authorized the appointment of an envoy, just as in no. ii, and to each decree is appended its own separate supplementary clause noting the appointment of Leukios, son of Dionysius (‘a’ and ‘b’). The delivery itself is recorded at ‘c’ followed in ‘d’ by a description of the seal that had been applied to them. Again, the clauses dating the delivery and describing the seal are Astypalaian, not Smyrnian.

Inscriptions nos iv, v and vi are from the late third century BCE, part of a large dossier of some 50 responses to an invitation issued by Cos to recognize its sanctuary of Asclepius as ‘inviolable’.Footnote ¹⁸ The royal letters and civic decrees received in response were displayed on a series of marble stelae, a handful on each stele, each document introduced with a title identifying the sending city with a genitive plural governing an assumed ψήφισμα. Thus no. iv introduces a decree of Pella with the title (ψήφισμα) Πϵλλαίων, and then a pair of decrees from Naples and Elea (Velia in Latin):Footnote ¹⁹

Almost all the decrees in this dossier are presented like the one for Pella, with a simple title followed by its text. The decrees of Naples and Elea are introduced differently, apparently to lend more prominence to the theōroi who had undertaken the mission to these distant cities. Each is then inscribed with a title that both identifies it and describes the emblem that appeared on the seal (‘c’ and, possibly, ‘e’).

Inscription no. v presents the decrees of five cities. It is laid out thus:Footnote ²⁰

Note the titles introducing the decrees of Corcyra (‘d’) and (if the supplement is correct) Leukas (‘e’), which again describe the seals. These, together with the titles in no. iv, are the only descriptions of seals amongst the roughly 50 documents related to asulia that have survived at Cos.

Inscription no. vi is a recently published stele, part of the same series as iv and v, that presents the decrees of four cities and two royal letters.Footnote ²¹ Each of the royal letters describes the monarch’s seal.

III. Extra-documentary details as archival metadata

Especially noteworthy in these inscriptions are clauses that have been added by a receiving city: for example, the titles added at the top of the documents in nos iii, iv, v and vi that identify a text as ‘Decree of [city name]’. Cities do not use such titles for their own decrees; they are ubiquitous, however, for foreign documents, for the obvious reason that no one would know a foreign document’s origin without such a title. Much rarer are reports of delivery such as those found in nos i, ii and iii,Footnote ²² and extremely rare are the descriptions of seals such as those found in nos ii, iii, iv, v and vi. The rarity of delivery reports and seal descriptions invites explanation, not least because all documents sent from one city to another were delivered by someone (obviously) and sealing is a well-attested practice.Footnote ²³ Nonetheless, deliveries were rarely reported and seals, almost never. Indeed, the seven examples cited above in nos ii–vi are the only seal descriptions known to me, and it is striking that six of them are from a single city (Cos), all but one of which were part of the same diplomatic project.

But why ever make a record of an incoming document’s seal? It is worth stepping back and considering this from the archival perspective. The details that we have been considering (the identity of couriers, the dates of delivery, the descriptions of seals) are not part of their documents but about those documents. In modern terms, these are metadata, data about data, and various kinds of metadata can be found in Greek inscriptions: headings and titles, dates of processing, notes recording delivery, names or titles of officials involved in their reception. As we have seen, such details are sometimes appended below their documents (as in the Josephan letter of Areus and nos ii and iii), but sometimes are added above them either in self-standing fashion (nos iv and v) or included within a ‘docket’.Footnote ²⁴ Each city, one assumes, had its own practices regarding what kinds of details to preserve within its archive and under what circumstances one or more of them might be included when a document was prepared for dissemination or display.

But why keep a record of such details at all? Here it is important to remember that there was a practical side to this. A detail such as an envoy’s name would be lost if not recorded. So, too, the appearance of its seal, which will probably have been destroyed during processing. In private contexts, of course, correspondents knew one another’s seals at a glance. Between institutional actors, however, such knowledge would have to be bureaucratized and such details kept ‘on file’, that is, in some piece of metadata. What we find in these documents is a vestige of that.

IV. The epigraphic comparanda and the clauses in AJ 12.227

So, what are the final clauses found at the end of Josephus’ version of Areus’ letter? That question has been debated for centuries; some scholars have asserted that they are part of Areus’ letter and others that they are something that Josephus added.Footnote ²⁵ In fact, they are neither. When such clauses are found in the epigraphical record, they are archival. The note recording Demoteles’ delivery (§227b) finds parallels in the final clauses of nos i, ii and iii, above, and in the heading of no. iv. The same point can be made about the descriptions of seals, as is best illustrated by presenting them together in series:

• seal of Areus’ letter: ἡ σφραγίς ἐστιν ἀϵτὸς δράκοντος ἐπϵιλημμένος (AJ 12.227d);

• seal of unknown city: ἔχον ἐπίσαμον | ζῴδιον γυναικϵῖον (no. ii, ll. 25–26);

• seal of Smyrna: ἁ σφραγὶς το[ῦ δήμου τοῦ Σμυρναίων(?)] (no. iii, l. 102);

• seal of Naples: ἔχον ἐπίσαμον ζώιδιον ἀνδρϵῖον (no. iv, l. 18);

• seal of Elea: ἔχον ἐπίσαμον ζῶιον γυναικέον καθήμϵνον (no. iv, l. 28);

• seal of Corcyra: ἔχον ἐπίσαμον γρῦπα (no. v, l. 56);

• seal of Ziaelas: ἔχουσαι ἐπίσαμον ἱππῆ (no. vi, l. 67);

• seal of Laodike: ἔχουσαν ἐπίσαμον ἄνκυραν· (no. vi, l. 76).

The image of each city’s seal is different, of course, that is their entire point, but it is striking how similar these descriptions are to one another formally: each is offered paratactically without subordination or conjunction. This is, of course, consistent with what might be expected of archival practice, where no artistry is needed or appropriate, only the efficient recording of the relevant facts. We should suppose, then, that the clauses at the end of Josephus’ version of Areus’ letter are not part of Areus’ letter, nor Josephus’ additions; rather, they are either archival vestiges or clauses forged in imitation of them.

But what is to be made of ‘what is written is square’ (§227c: τὰ γϵγραμμένα ἐστὶ τϵτράγωνα)? This, too, has provoked some puzzlement. Given, however, that the naming of the courier and the description of the letter’s seal are archival, the description of its script (which appears between them) should be, too, and therein lies a possible explanation. During the course of the third century BCE, Greek letter-forms began to move away from ‘epigraphic’ letter styles, which were upright and square, and towards rounder ones: four-bar sigmas (Σ), for example, gave way to lunate ones (ϲ); epsilons evolved from ε to ϵ; omegas, from Ω to ω.Footnote ²⁶ In this context, one might reasonably guess that an archivist (or someone imitating one) took note of the old-fashioned ‘square’ lettering of an incoming document, in this case, perhaps, reflecting (or imagining) a Spartan conservatism in orthography that is not (to my knowledge) otherwise attested. If, as seems likely, the archiving of documents typically involved retranscription, this aspect of the physical appearance of a document would be lost if not noted in the archive’s records. In this case, it might also imply that whoever was responsible for these clauses, presumably archivists in the temple in Jerusalem, knew what contemporary lettering should look like. In itself, this should probably cause no great surprise. Any interaction with either the Ptolemies or Seleucids will have occurred in Greek, and it has been widely accepted since Martin Hengel that many within the upper echelons of Judaean society were highly Hellenized.Footnote ²⁷

V. Why include archival details with Areus’ letter?

Even if all this is correct, however, why append these details to the foot of Areus’ letter now? Some scholars have invoked a concern for security, suggesting that Areus had included these details to allow Onias, the recipient, to check whether someone had broken the seal and replaced it with another.Footnote ²⁸ Such a security measure, however, would be comically easy to circumvent by breaking the seal, rewriting the letter, and describing the new, fake seal. A seal only guarantees security if it is already known or delivered separately (as Pliny the Younger seems to have done at least once).Footnote ²⁹ Neither seems a realistic possibility here.

So, why are these details included with Areus’ letter? Again, the specific context is important. As noted above, Areus’ missive to Onias had originally been appended to Jonathan’s letter, but Josephus had moved it for his own reasons to his narrative of Onias III. This was, as several scholars have noted, a mistake. Jonathan (or whoever composed this letter in his name) seems, however, to have made a similar misidentification.Footnote ³⁰ That, at least, seems the best way to explain his excuse for the delay in answering Areus’ letter, which invokes recent wars and tribulations and a Judaean reluctance to burden their kin with their own troubles. Such an excuse more plausibly alludes to the quarter century from the time of Onias III (whose assassination in 175 BCE marked the beginning of the turmoil that is alluded to) than to the century-plus that had passed since the time of Onias II (or I). In the light of such an assumption, describing the letter’s seal, characterizing its script and naming its courier makes some sense, since these details offer the Spartans a means of authentication: it is reporting to them things about the letter that its sender would recognize. Areus’ seal might be remembered; the old-fashioned script might resonate; and Demoteles’ diplomatic mission might be recalled. Indeed, given Jonathan’s erroneous assumption that only a few decades had passed since Demoteles’ mission, he might have supposed that, even if no longer alive, he might be remembered by his countrymen. The archival clauses at the end of Areus’ letter, therefore, were added by Jonathan (whether real or fictional) to allow the Spartans to authenticate it.

VI. The formula ualetudinis at AJ 13.166

The clauses that appear at the end of the Josephan version of Areus’ letter, then, are archival, or crafted in imitation of such. This requires, as several scholars have already deduced, that the ultimate source for Josephus’ version must have been archival,Footnote ³¹ which also requires that it is independent of 1 Maccabees. This also explains the presence of a formula ualetudinis in the Josephan version of Jonathan’s letter, but not Areus’.Footnote ³² Again, some have suggested that Josephus had added this to make his text more believable,Footnote ³³ which is less convincing, especially given that Josephus does not seem to add such clauses to other letters. The Letter of Aristeias, for example, quotes three documents in its narrative, and in his rewriting of that narrative Josephus includes all three. The first does not include a wish of good health (8.29), but neither does Josephus (AJ 12.36). The second (Aristeias 9.35) has a short one in the form of χαίρϵιν καὶ ἐρρῶσθαι, which Josephus trims to χαίρϵιν (AJ 12.45). The third (Aristeas 11.41) contains a rather elaborate wish, which Josephus paraphrases fairly closely.Footnote ³⁴ Similarly, among the Roman letters quoted by Josephus, no additions can be found. Caesar’s letter of 47 BCE to Sidon, for example, contains an elaborate wish of good health, which is closely paralleled in two epigraphically preserved letters of Caesar.Footnote ³⁵ Similarly, the formula found in Antony’s letter to Hyrcanus finds an almost identical match in a roughly contemporary letter of (probably) Octavian to Aphrodisias.Footnote ³⁶ The Augustan letters quoted by Josephus do not include such a clause, but to judge from contemporary epigraphy, usage had changed and such wishes were no longer expected.

None of the evidence suggests, then, that Josephus was inclined to change much, which implies that the formula ualetudinis appears in the Josephan version of Jonathan’s letter because he found it in his copy, and not in Areus’ letter because he did not, both of which would be consistent with the fact that wishes of good health seem to have been a standard feature of official letters of the second century BCE, but not of the mid-third.Footnote ³⁷

VII. The Josephan versions do not show signs of translation or retroversion

If Josephus acquired these letters from an independent source, the obvious question is whether they, like the Maccabean versions, had been translated from Greek to Hebrew and back to Greek. Almost certainly not, at least to judge from the texts as we find them in the Antiquities. The very act of translating a text, it is important to remember, has certain predictable effects that are obvious in 1 Maccabees but absent from Josephus’ versions. One such effect is lexical simplification, which is so common in translations that some theorists include it among a short list of ‘universals of translation’, or features found in translations regardless of the source or target language.Footnote ³⁸ In the case of the Maccabean versions of these letters, which seem to have been translated at least twice (Greek-to-Hebrew-to-Greek and possibly Greek-to-Aramaic-to-Hebrew-to-Greek),Footnote ³⁹ simplification is fairly evident. The verb ἀποστέλλϵιν, for example, appears six times in the Maccabean versions, which are listed here with their Josephan parallels, where they exist:

The difference in verb choice in (e) and (f) (ἀποστέλλϵιν vs πέμπϵιν) is trivial. Nor is much more to be made of (b), where 1 Maccabees has ‘the sent man’ (τὸν ἄνδρα τὸν ἀπϵσταλμένον), probably in order to avoid naming Demoteles. (Simplification may again have been a consideration, or a narrator’s understandable disinclination to name a character who plays no further role in the narrative.) In (c) and (d), however, 1 Maccabees uses ἀποστέλλϵιν in describing written communication, and in doing so probably reflects its Hebrew stage, likely using the verb שלח (šlḥ, literally ‘send’, that is ‘send word’ or ‘write’),Footnote ⁴⁰ which seems a textbook example of interference, another translation universal.

In the Antiquities, by contrast, greater lexical variety is evident in these letters. It uses κομίζϵιν in (a), which is very common in diplomatic contexts and perfectly idiomatic here.Footnote ⁴¹ At (d), instead of the Hebraism ἀποστέλλϵιν (‘send [word]’ or ‘write’) the Josephan version uses a much less common word, ἀναπολϵῖν (‘remind’, lit. ‘turn up the ground again’; hence, ‘go over again, repeat’, LSJ, q.v.), which is unlikely to be the product of a double translation: from a continuum of lexical options, translations typically opt for a midpoint.Footnote ⁴² Whatever one makes of each of these points individually, taken together it looks like the author of 1 Maccabees has used a single Hebrew verb where his source texts had used five or six different ones, and that the Greek translator had little choice but to render them all with ἀποστέλλϵιν in his retroversion.Footnote ⁴³ Once the lexical bridge had been burnt, the translator had no path back. The Josephan versions, however, are consistent with what we would expect of a non-translated text.

Similar flattening can be seen elsewhere in the Maccabean versions. Kinship terms, for example, are limited there to ἀδϵλφός (‘brother’) and ἀδϵλφότης (‘brotherhood’); in the Josephan versions we find not only ἀδϵλφός,Footnote ⁴⁴ but also συγγένϵια (‘kinship’: 13.167, 169) and οἰκϵιότης (‘familial bond’: 12.226, 13.168).Footnote ⁴⁵ A similar observation could be made about how Spartans are referred to. The Maccabean version of Jonathan’s letter (1 Macc. 12:6) is addressed to the Spartiates (Σπαρτιάταις), and Areus refers to himself as the ‘King of the Spartiates’ (βασιλϵὺς Σπαρτιατῶν: 1 Macc. 12:226). The term ‘Spartiate’, however, referred to the elite warrior class that dominated Sparta, which included other (subordinate) classes;Footnote ⁴⁶ in inscriptions, it is always ‘Lacedaemonians’.Footnote ⁴⁷ This is another case, one assumes, of lexical flattening. The Josephan versions use the correct term, Lacedaemonians (AJ 12.226, 13.166). It is also noteworthy that the Josephan version correctly calls the chief magistrates of Sparta ‘ephors’, which is more impressive than it seems at first glance. Letters to Greek cities were typically addressed to some combination of a city’s ‘magistrates, council and people’ (ἄρχουσιν καὶ βουλῇ καὶ δήμῳ); when magistrates are included, they are almost always referred to generically as ἄρχοντϵς (‘magistrates’, here), even in cities that had no specific magistracy called that. Although this practice is standard almost everywhere else, Sparta was one of a handful of cities where this was not the case. Extant letters to Sparta are addressed only to its ἔφοροι.Footnote ⁴⁸ In short, the Greek translator of 1 Maccabees seems to have assumed that Jonathan had addressed his letter using the generic ἄρχοντϵς (‘magistrates’), which would have been the correct guess for almost every other Greek city. The version presented by Josephus avoids this mistake. Why was Josephus able to get this right? Because he was copying what was in front of him, and it had never been translated.

VIII. Some conclusions and implications

Our findings can be summarized thus: the clauses found at the bottom of the Josephan version of Areus’ letter have parallels in Greek epigraphy that are clearly archival (sections II–III), which implies that they are, too (section IV). A motive for including them thus emerges: they had been included when Areus’ letter was sent back to Sparta so that the Spartans could authenticate it (section V). Given that the Josephan version had included such archival material, and that his version of Jonathan’s letter includes a formula ualetudinis, these documents must be independent of those found in 1 Maccabees (section VI), which explains why they betray none of the typical artefacts of translation that are so apparent in the Maccabean versions (section VII).

These findings have several implications, the most obvious of which pertains to their authenticity, a question that has both dominated previous scholarship and produced a bewildering array of answers.Footnote ⁴⁹ Little would be gained by rehearsing all the arguments on this question here, and a full exploration of them would lengthen an article that is possibly too long already. I will limit myself to a few observations. First, it is important to acknowledge from the outset that, strictly speaking, authenticity can never be proven for any document, ancient or modern: the best that can ever be said is that evidence of forgery is absent. Because of this, the burden of proof lies with those who allege forgery. Second, any future argument about authenticity needs to concentrate on the Josephan versions of the letters of Areus and Jonathan, which are not only independent of the Maccabean ones, but also (and more importantly) better than them (unsurprisingly, given that the Maccabean versions were translated at least twice and, to judge from the result, not well either time). Third, although a degree of doubt will inevitably linger about these letters, not least because the discovery of ‘lost’ documents was a tactic in the ancient world for introducing new material into a discourse and elevating its significance,Footnote ⁵⁰ any argument against authenticity needs to explain where the archival details appended to the foot of the Josephan version came from and why these details were included here. This significantly complicates the argument against authenticity, perhaps beyond the point that it can be reasonably sustained.