Tragedy, Authority, and Trickery: The Poetics of Embedded Letters in Josephus

  Olson, Ryan Scott. 2010. Tragedy, Authority, and Trickery: The Poetics of Embedded Letters in Josephus. Hellenic Studies Series 42. Washington, DC: Center for Hellenic Studies.

Chapter 5. Conclusion

Now that we have explored various aspects of the poetics of embedded letters in Josephus and in other Greek authors, we have left to ask, why does Josephus use all of these letters? Such a question is another important element of epistolary poetics, since it can help us to understand Josephus’ literary and cultural aims. This final chapter will attempt to answer this question and will conclude by noting implications of this study for our understanding of Josephus and of the broader field of ancient epistolography.

Epistolary Refracting and Signaling

In general, Josephus followed his sources both directly in parallel episodes and indirectly in style. Recall the analysis of Josephus’ texts about David and Cambyses, and their parallels in Greek translations of the Hebrew Bible, as well as the Letter of Aristeas and Nicolaus. Josephus wrestled with the same questions of reliability that Thucydides and Euripides did before him, repeatedly positing the slipperiness of letters, which will be discussed further below. Josephus reflected a world that could be understood more fully by including the extensive use of letters as historians before him had included. These influential Greek historians included Diodorus, Dionysius, and Polybius, as well as Herodotus and Thucydides, to name a few.

Josephus’ Jewish cultural heritage also exerted some influence, as several of the works discussed above illustrate. This heritage reaches far back into a form of “scribal culture” that produced the Hebrew Bible over many centuries, but draws most directly from Greek translations of that sacred literary tradition. Josephus also changed his sacred source significantly, even though he claims to stick closely to it in the first half of the Antiquitates Judaicae. Thus the Greek epistolary embedding practices that Josephus employed refract the more functional, spare use of letters in Hellenistic Jewish literature seen in Greek {207|208} translations of the Hebrew Bible and in Maccabeean literature. Yet Josephus also refracted Graeco-Roman literary practices, innovatively appealing to his readers at a meta-textual level (recall the Herod narratives) and thoroughly developing characters (recall Cambyses) and plot (recall Gaius and the Temple). Described in the detailed analysis above, the resulting picture in Josephus, wherein he aspires to and builds on the Greek tradition before him, is unique.

Why, then, did Josephus use letters? In part for reasons already discussed at length: because he found them in sources, such as Nicolaus and the Letter of Aristeas, in literary forebears like Diodorus and the Maccabeean books, and in his cultural milieux, represented by documentary letters and such distant authors as Thucydides. His submission of letters to readers on a meta-textual level opened the door for letters to be considered as having some additional function than developing the plot or the characters or internal evidence for a trial. Josephus laid evidence before his readers, in a few cases even inviting them to check background and archives themselves, and allowed them to draw their own conclusions from the information he had marshaled.

A modern way of framing Josephus’ extensive epistolary usage is to focus on incentives. Consider Josephus’ historical position. He had fought in and survived a war with the known world’s suzerain. He had surrendered and curried imperial favor. The center of the Jewish world, the Temple, lay in ruins, as did Jerusalem. And, as evidenced by Josephus’ speaking of this central institution as if it still existed even after it had been destroyed, Judaea’s future remained unclear (Contra Apionem 2.193–198). Thus Josephus had to convince part of his audience, the Roman elite, that Judaea still had some strength. Perhaps Josephus, looking ahead, was surmising that the battle with the Romans may not be over. Perhaps he was attempting to impress readers with his and Judaea’s success at resisting Roman conquest for so long (Bellum Judaicum 1.8). But more importantly, failing the ability to demonstrate a convincing show of military might, Josephus was choosing to portray the Jews as important cultural collaborators, raising awareness of the cost of their destruction. He promoted peace, long a quality of Jewish history, as shown in the numerous epistolary exchanges already described, by demonstrating the Jews’ quality as a partner in cultural exchange. Contra Apionem is the clearest statement of this, but the thread of embedded letters throughout Josephus—with the exception of the court scenes in the Herod narratives save for Josephus’ literary sophistication in handling Herod’s tempestuous court—coalesces to arrive at the same point. As it turned out, Josephus was right to worry; {208|209} unrest between Rome and Judaea, though by no means inevitable, would flare up at least two more times in the next several decades: the uprising under Trajan (113–115) and the Bar Kochba revolt under Hadrian in 132. [1]

Given this historical position, what was Josephus doing with letters? A helpful explanatory model can be borrowed from microeconomics—that of signaling theory, which presents signals as perceivable indicators of qualities that are not directly observable. [2] Developed in 1973 by Nobel laureate A. Michael Spence, the theory explains how agents in a market behave when information is asymmetrical. [3] The concept is part of the field of information economics and has been applied by biologists to explain signals exchanged by animals and how they relate to the signalers’ qualities. [4] Spence focuses particularly on the job market, in which an individual is attempting to sell his or her labor to a potential employer, the buyer. In general, buyers are willing to pay more for better labor, so a seller desires to communicate his or her potential value to the employer. Because the employer does not know the potential value, or productivity, of the seller, but the seller does know his or her own value, information in the job market is said to be asymmetrical. To close this knowledge gap between the buyer, who desires to know, and the seller, who desires to communicate, the seller can produce a signal to indicate his productivity. One signal that Spence focuses on is education. An advanced or professional degree is a signal that a person is able to enter and successfully exit a difficult regimen of instruction, research, and examination. Thus education is a proxy for qualities that are otherwise not visible, such as self-discipline and intelligence. As such, educational attainment can be useful to a buyer in the labor market (employer) who wants to know whether a seller (job applicant) will be a productive employee. [5] Signals can be costly to acquire, as is the case with education, and costly signals are difficult to send for one not actually in possession of them (i.e. costly signals are difficult to fabricate).

Similarly, embedded letters in Josephus are a signal intended to communicate the Jews’ cultural productivity. Josephus recognized asymmetrical knowledge in the cultural market, as it were, between {209|210} Graeco-Roman understanding of Jewish history and culture and his own knowledge. Having survived the war with Rome and secured imperial favor, he likely saw an uncertain future for Judaea, which had lived at peace with its neighbors, including Rome, for centuries. He therefore attempted to signal Jewish cultural parity or even superiority. This must have been difficult: Josephus was competing with Greek and Roman literary achievements and antiquity. He had to demonstrate the sophistication of Jewish tradition, as well as the talent that could improve upon the best Graeco-Roman conventions. By doing so, he would have been able to show that Judaea was a fitting cultural partner for Rome, to be protected and nurtured, not treated as alien and rebellious. One way in which he did this was by drawing upon and innovating a common Greek literary device, the embedding of letters in narratives.

This means that Josephus was not using letters and practicing historiography because it was predetermined by his cultural heritage. Rather, it was in part the result of a judgment that, by using sophisticated Greek literary style to demonstrate cultural equity or superiority with the Romans, he could secure the Jews’ peaceful future, one that was aligned with the epistolary evidence of peaceful interaction over a long history of foreign relations. This signaling also had an internal audience: embedded letters reminded the Jews of their placid history with foreign powers, and the form itself demonstrated that Jews could be culturally similar, or even superior, to the broader cultural context in which they found themselves in Judaea and throughout the Diaspora.

To all audiences, Josephus has been able to signal with letters his own authorial meticulousness and the sophisticated ways by which he engages his readers. Josephus has researched his epistolary material, found and weighed the evidence, reached clear conclusions, and embedded it deftly; and in some cases he has paid his reader the compliment of trusting them to work out for themselves the epistolary puzzle, doubtless using the procedures of his epistolary poetics as a model that they have increasingly come to trust as they proceed through his works.

Josephus as Source for Epistolary Culture

All four pieces together provide only the briefest and most partial of introductions to the topic of embedded letters and their literary and rhetorical uses, which is a far larger one than can be illustrated by brief excerpts in an anthology such as this. Much depends, for instance, in drama and (especially) the novel on the relationship of the letter and its contents with the surrounding narrative context, which can only properly be explored in an unexcerpted text.

This, of course, is quite right.

It may be tempting to treat an embedded letter as a kind of archive to be explored. And it can be an archive of sorts, in the sense that an (ostensibly) ancient letter has been recorded and transmitted by scribes and copyists in manuscripts through the centuries. But, then again, it is also not an archive, since it becomes intratextual when it is written out, summarized, or referenced in a text; an embedded letter becomes related to a larger narrative. Ceasing to have a life purely of its own and remaining part of a whole, the letter itself, as well as the whole text, lose something.

Take the example of Cambyses’ letter. It is significant, both for Josephus’ text and for the letter, that it is presented as from Cambyses and not another dynast, as Ezra and 1 Esdras present it. In light of Cambyses’ name and the malevolent content of the letter, one may recall Herodotus’ “mad” Cambyses, which lends a natural explanation for his bad behavior and his (anomalous) treatment of a normally well-liked people. And Cambyses as epistolary author, which is revealed in the embedding narrative and not the address formula quotation, makes good sense for the letter, too, since the sender can be deemed capable {211|212} of such an “obvious” mischaracterization of his subjects. Josephus thereby undercuts the letter’s contents.

At the same time, embedded letters can affect the narrative with their materiality, as evidenced by the example of Herod’s exaggeration of a letter inaccessible to the internal audience or of the self-description of the Spartan Areus’ letter to the Jewish high priest Onias. Josephus apparently knows that materiality can affect his readers’ perception of letters’ reliability, so their materiality can be important for shaping or coloring his narrative. Letters’ materiality highlights more directly the notion of epistolary archivability, which obtrudes in Josephus’ narratives as discussed above.

However, letters as epistolary intratexts are also different from an archive of letters in their potential to play on a meta-level. Letters at court, whether extra-episodic or episodic, help to create a meta-textual court, where audience members weigh evidence, under the influence of the narrator to some extent, but by their own estimations as well. An incriminating letter in an archive, without context, may tell quite a lot about the author, the recipient, and even the circumstances, but without an embedding narrative it remains flat. Even if an archive explorer were to find two letters about the same event, an original and a reply, we still would not know about the glances, the expressions that reveal the prosecutor’s change of mind, the gasp of the council, the outcry of the witness. These narrative elements all draw from the courtroom letters and affect our reading of them.

The implication of the interaction of letters with their embedding texts is that embeddedness ought not be ignored at any time. For historians interested in what embedded letters tell us about the world outside of the text, a narrative can give important clues. Did Josephus dig Antiquitates Judaicae 14 and 16 letters out of archives in the receiving cities? He might have. [7] But the narrative, however spare it may be at those points, tells us that Josephus thought his readers would find the material convincing. If we were to find one of the decrees in an excavation, we might be able to tell something of its value by the fact that it was inscribed on a stele. We might be able to tell that it was significant by its proximity to an important area of the site. We might be able to find another like it that would tell us it was not unique. However, by itself, it would tell us little of the reaction to the decree; it would tell us little of its typicality or its value to its intended recipients. But in a narrative, the epistolary decree can be introduced, placed in proximity {212|213} to other narrative elements, illuminated with commentary, and its value can be reflected back through rhetoric that the author must have supposed the audience would have found persuasive. True, it can also be distorted and manipulated. But even then, we would learn something of the audience it was intended to impress, of the way it might be considered impressive, the way the embedded letter may relate to a documentary letter. To take an ideal but telling example, Claudius’ epigraphic speech may be useful as a “control” for Tacitus’ version, but Tacitus’ version tells us something that the inscription alone cannot.

Stressing embeddedness raises again the issue of the nature of the embedding text and, specifically, generic expectations. As this study has shown to a limited extent, letters can transcend generic boundaries; they can be found in historiographical narrative, in tragedy, in “autobiography,” or in “biography.” [8] I have not stressed the generic boundaries simply because it has not been necessary. There do not appear to be major differences in how letters look or in what they do in particular settings. Josephus’ use of letters in the Vita (e.g. 62) looks much like it does in the Bellum Judaicum or Antiquitates Judaicae. True, there is a personal element to the use of letters in the Vita—Josephus receives a letter from his father (204); he is the focus of attention in several operations involving epistolary exchanges (e.g. 186, 245, 269, 310)—but he also sends collaborative letters in the Bellum Judaicum when discussing his generalship (3.140, 191). The Vita is most notably different because there are more letters to, from, and about Josephus than anywhere in his strictly “historiographical” writings, though characters use letters as evidence in his other narratives, and in the Vita Josephus as character uses letters in this way (254, 260). (But then again, so does Josephus as author in Books 14 and 16 of the Antiquitates Judaicae.) This personal use in the Vita should not be surprising, given the extent to which it is, of course, natural for him to focus on his own activities in that work.

Crossing generic boundaries, Josephus presents himself as a master of letters in his historical narratives and in his autobiography. In the Vita, he outsmarts his opponents epistolarily, just as the Greeks outsmarted the Persians. He also uses letters to defend the veracity of his history of the Jewish War against his critics, citing the approval of Flavian emperors and quoting letters from Agrippa. In the Antiquitates Judaicae, he cites epistolary evidence to prove the esteem in which rulers had held the Jews historically. This fits well with Josephus’ theme in Contra Apionem of the antiquity and cultural parity, if not superiority, of the Jews.

Implications for Josephus

One theme that emerges from the analysis is the slipperiness of letters. Narratives that appear on the surface to be quite straightforward in terms of the intended effect or import of letters often contain subtleties that bring the nature of the epistolary intratexts into question. Texts that idealize epistolary communication, such as the Cyropaedia, sharply contrast with other narratives where certain elements can and do go wrong.

But Josephus evidently determines that the weighing needs to be done, since, as we have seen, the evidence does not always fit neatly into assertions of guilt or innocence. This does not mean that Josephus shies away from attempting to use letters to attest truth, or at least {215|216} his version of the truth. Though he presents forgeries and covert letters sewn in the tunics of denying messengers, Josephus himself marshals letters as evidence, apparently assuming that his audience will consider them convincing. Josephus presents Herod’s approach to letters in the same way when Herod passionately reads a letter before the council at his sons’ trial; the letter could probably be seen in his hands, at least in the Antiquitates Judaicae account. Moreover, the “writtenness,” the materiality, of letters is essential to convey authority and avoid the foibles of incompetent messengers (recall Herod and Nicias, too). But then again, writtenness also causes trouble—interception, forgery, exploitation of normal epistolary practice (recall Aristobulus’ commands in his own hand), exaggerating or lying messengers, denial of access due to the inflexibility of the recorded message. Thus the weighing of letters has to be done one text at a time, and though they are generally slippery, they are not always intended to be so. Attention to subtlety is important, for the answers are often elusive.

The elusiveness and trickery of letters in Josephus, paralleled in other authors’ works, seems to fit quite well with the perennial problem of how to take Josephus’ self-presentation. [14] How did Josephus want to be portrayed as an author? As Graeco-Roman? Jewish? Josephus has more recently been thought to have considered himself in a multi-faceted way as Jewish, Greek, and Roman, though defining that mix more precisely is fraught with problems. [15] This study of embedded letters will most likely frustrate attempts to set the self-presentation answer too firmly. While he may at times have included an element in the Antiquitates Judaicae that seems Graeco-Roman, such as using a letter to put rulers on speaking terms, the biblical parallel to the Antiquitates {216|217} Judaicae text has a similar letter, and such a convention has its roots, not only in the LXX, but it is also traceable, perhaps only through the LXX, to the MT. Yet Josephus is not content to leave the letter buried in a Hebrew text, which might be overlooked. He can contemporize letters, and does so in two different texts concerned with the same letters, by referring to and inviting his readers to consult an archive in Tyre that still contains the letters from the Hebrew and Tyrian kings. [16] Josephus’ act of “preserving” the archive in his narrative brings the presence of collaboration by means of letters in Jewish texts in line with Graeco-Roman practices and gives his readers a multifaceted sense of his identity, as Roman, Greek, and Jewish.

But the identity of that some of the audience remains an open question. This study has referred to audience expectations of embedded letters, an idea that in the first chapter linked to generic horizons, that is, what a reader of Josephus’ works may have expected when reading a work of history or autobiography. [20] Remembering a few of Josephus’ texts should clarify the point. Jewish readers, who I argued were part of Josephus’ audience, [21] would have recognized the letters that had been important in the biblical texts with which they were undoubtedly familiar. David’s letter with the order to station Uriah on the front lines of battle, carried by Uriah himself, would no doubt have been recognized by Jewish readers, who could recall the original, biblical story, perhaps in Hebrew. That same episode would have struck some Graeco-Roman (and some Jewish?) readers as a Homeric echo or an anticipation of Bellerophon. The epistolary exchanges around the Cambyses incident would have been recognized as having their provenance in an important time in Jewish history, and Jewish readers who may have been familiar with the Hebrew/Aramaic text would most likely have been interested to see Josephus’ “corrections” from another source, which they may well have known. They probably also noticed the slight changes, the simplification of “tributes” and omission of the “copy,” which they would have found familiar if they were aware of other parts of Josephus’ texts. Likewise, upon mention of Popilius, readers—perhaps Graeco-Roman—may have recalled one of a number of intertexts, remembering through the “trigger” παραγγειλάντων something of the circumstances described by Polybius or the apophthegm, regardless of whether Josephus had recounted the episode earlier. As to letters, then, there appears to have been “something for everyone,” as Josephus refracted Greek practices and signaled Jews’ cultural sophistication and value.

This universality makes Josephus, as I argued at the outset, an important author for the study of embedded letters. With a clearer understanding of the poetics of embedded letters in Josephus, it {218|219} becomes easier to study other Graeco-Roman authors’ epistolary material on its own terms. Such a poetics should take into account the range of epistolary terminology, the relationship between letters and speech, letters’ presentation style in the narrative, the presence (or lack) of an archive concept, letters’ documentary nature and effect, their basic functions, the extent of their reliability, and, ultimately, their purpose.

Such subtlety in narrative as can be seen in Josephus’ handling of letters belies the idea of the author as Ἰώσηπος Ἀντιγραφεύς. We have here not a historian who has put on blinders, trading out and copying sources with no personal creativity or vision; rather, we have a historian whose life experiences as Jewish provincial aristocrat, anti-Roman Galilean general, Greek author, Roman citizen, and recipient of Flavian beneficence, insofar as he reveals these experiences, dovetail remarkably with his literary style, particularly his use of letters, which facilitates the drawing together of various literary traditions and involves a fair amount of finessing. Josephus’ works deserve and repay subtle readings by literary scholars and historians interested in the long century for which Josephus is our best and only Roman Near East correspondent. {219|}


[ back ] 1. For a summary of sources and literature, see Goodman 2000b:669–674.

[ back ] 2. Donath forthcoming provides a non-technical summary.

[ back ] 3. Spence 1973.

[ back ] 4. Donath forthcoming; see e.g. Cowen 2007:79–112.

[ back ] 5. For a technical description, see

[ back ] 6. Trapp 2003:311; Rosenmeyer 2001 presents detailed analysis highlighting the importance of the narrative as well: see p29.

[ back ] 7. See Rajak 2005:90.

[ back ] 8. Costa 2001 shows this even more with the letters he has excerpted, particularly from novels. On biography, see Pelling 1999:329n14; on autobiography, the introduction to Mason 2001. See the qualification of Josephus’ use of the autobiographical genre in Rajak 2002:12n4.

[ back ] 9. Chapman 1998:51–106 describes the cannibalism committed by a mother at the sacking of the Temple: “In delivering this account of cannibalism, Josephus faces a real historiographical challenge: how can he relate the horrible deed without losing the sympathy of his readership for the general Jewish populace, whom he consistently portrays in his history as innocent? He cannot omit it, since, as he informs us, the Roman army learned of the cannibalism. In fact, he tells us in his introduction that the readers for his Bellum included the Flavian family, Roman veterans of the war, and prominent Hellenized Jews. And how can he deflect blame from the Roman commander, Titus, and his army? Josephus finds his answer in Greek tragedy. Josephus purposely uses the language of tragedy to encourage his audience to pity Mary. Tragic madness and desperation dominate Josephus’ account of the famine. For the episode of Mary, the historian purposely evokes Euripides’ descriptions of the desperate circumstances of mothers such as Medea and Agave in the Bacchae in order to increase the pathos” (52–53); see also Chapman 2005:302. Note also Landau 2003 (= Landau 2006) on the Herod narratives.

[ back ] 10. Pelling 1997b notes that tragedy can provide historical evidence, “illustration of what it means to be a citizen, what a citizen would feel uneasy about” (234); for general conclusions to the essays in Pelling 1997c about the importance of real life to the interpretation of tragedy, see 234–235.

[ back ] 11. Cf. the discussion in Marincola 1997:318–319 of the “history” included in Tacitus Agricola, a work “usually classified as a biography (not a history proper),” with details of Britain’s history included to an extent that is perhaps “more than is needed for a biography” (319); Marincola also points out (310–311) that Xenophon in his Hellenica acknowledges that he is continuing Thucydides’ history, but he actually produces “contemporary history with a different focus, purpose, and orientation” from Thucydides’ history (311), which is underlined by Pelling 1999:327–328; Marincola argues, with other examples as well, that Graeco–Roman historiographical genres are not “static categories in which one writer merely followed all or most of the aspects of his predecessors …” (320).

[ back ] 12. See p39; Pelling 1999: “it is evident that any genre of historiography will be a shifting and cumulative thing, with the audience’s expectations developing as the store of past works grew greater” (328).

[ back ] 13. For example, recall Mason’s comment about the letters in the Vita, p23 above.

[ back ] 14. The slipperiness of letters also goes with other slippery problems, such as how to interpret Josephus’ Masada narrative at the end of the Bellum Judaicum; Rajak 2000:xv puts it well: “How are we to explain why Josephus highlighted and appeared to eulogise the fate of the last of the detested rebels whom he had up to that moment not ceased to denounce? … The jury remains out on the Masada story, a symbol, perhaps, of Josephus’ ability to elude us just when he is at his most vocal.”

[ back ] 15. See the discussion of Rajak 2002: “We now understand much better than we did [at the time of the appearance of Rajak 1983] that his [Josephus’] enterprise was part of a long process of integration of the cultures, not the combination of two distinct and contrasting systems [i.e. Jewish and Greek] … At the same time, it is clearer than ever that, even in the Jewish War, composed when the author was not long out of Jerusalem, Josephus’ writing does not represent a pure Palestinian Judaism clothed in Greek dress … Josephus’ own identity as a Roman at Rome remains, however, shrouded in mystery…” (xiii). Note also the much stronger statement in Schwartz 2010: “Despite the unattractive toadying, Josephus, at least in his writing, never identified himself as a Roman” (84).

[ back ] 16. See p107.

[ back ] 17. As Barclay 2005 points out, Josephus criticizes Greek historiography, but “he still wants, and perhaps needs, the support of Greek witness to both the antiquity and the cultural attraction of the Judean people” (40).

[ back ] 18. See the above discussion of readers, p38; for the possibilities of different receptions by different types of audience members, see Pelling 1997:220–221; Pelling 2000b:197–200.

[ back ] 19. Mason 2005a.

[ back ] 20. See p37.

[ back ] 21. See pp40–44.

[ back ] 22. See e.g. Jones 2005:204, 207.