Apparently not missing the tricky communication, Cyrus cuts open the hare and reads the letter. As a response, he considers the most cunning way in which he might convince the Persians to revolt (125.1). He decides it would be best to write in a letter what he wants to have happen (125.2). He then gathers an assembly of the Persians, and after unfolding the letter and reading it, he states that Astyages has appointed him leader of the Persians (125.2). Herodotus shifts to oratio recta for the announcement that they should come, each one with a sickle (125.2). The revolt proves to be successful (128.1): Astyages is taken prisoner (128.3), and Harpagus appears, taunting him (129.1). Astyages chides Harpagus for thinking that what Cyrus has done is his work (129.2). Harpagus responds that, since he wrote the letter, it is rightly his work (129.2).

At least in this episode, Herod’s problem, as far as Josephus’ narratives are concerned, is that the letter stands ironically against claims Herod makes in other parts of the episode. The scene points up the problems that epistolary material can present if managed improperly by the internal characters, if its message publicly says something those characters do not want it to say. A narrator could face a slightly different kind of problem if a letter fails to fit with other elements of the text, such as theme or characterization.

This chapter will explore two ways in which letters are used for attestation. First, I will examine ways in which letters appear within episodes in Josephus’ narratives as proof, often in the context of trials. In this “episodic” use of letters, characters within the textual world employ letters to substantiate certain claims. These claims appear in a chronological sequence with the other events in the plot.

This chapter will also examine ways in which Josephus employs letters as evidence for his own interpretation of events. This is an “extra-episodic” use of letters in the sense that the narrator is presenting a letter outside the order of events in the plot, which will involve texts that demonstrate subtle narrative techniques by which Josephus uses a letter that appears to be evidence in a trial or other situation, but in fact is a piece of evidence for building the narrator’s interpretation. These uses of the letter get to the heart of the meaning of letters in public and private life, including how they were perceived and the authority attributed to them.

This sequential-logical distinction is important for at least two reasons. It demonstrates that letters are not only useful for a historian when they fall within the immediate story line, they are also important for the author/narrator’s task of giving a plausible presentation of a series of events. Although these events may not immediately involve a letter, an extra-episodic letter reveals that an author has reflected upon the utility and function of letters. The point is important by itself for those who may consider an author’s use of embedded letters simply {165|166} a necessity, given that they appear in the sequential order of events in his source.

Circumventing considerations of an errant messenger, an epistolary document itself could be falsified. Forgery, the manipulation of epistolary material, is a significant topic in any epistolographic study. But the way in which forgeries are carried out—the motivations and often subtly nuanced effects—are inextricably linked with other portions of text and with the whole text, and this structural relatedness is precisely what intratextuality allows us to trace. Thus the relevance of “forgery” here is not primarily as a technical issue of handwriting analysis or archive mischief, but a matter of playing on identity, and how such play affects the situations in which letters are prominent.

The author’s identity in many instances of forgery in Josephus is not as clear as Diodorus makes it in the example above. A first example comes from Josephus’ Vita. When Josephus is describing cities as either for or against Rome, he discusses how Gamala remained loyal to Rome in AD 66 (Γάμαλα δὲ πίστει τῇ πρὸς Ῥωμαίους ἐνέμεινε δι ̓ αἰτίαν τοιαύτην, Vita 46). Here Philip is king Agrippa’s attendant. Babylonians in Jerusalem save Philip’s life after he escapes a siege of the palace by Menahem and some brigands (λῃσταί). He is hidden in Jerusalem for four or five days until he flees, wearing a wig, to a loyal town. He sends (πέμπει) for reinforcements, writing a letter (γράψας ἐπιστολάς) to Agrippa II and Berenice and sending it through one of his freedmen, who is to deliver the letter to Varus, Agrippa’s legate (48). Varus suppresses the letter (γράμματα), which tells of Philip’s escape, because it threatens his own power (50). And so to protect himself, Varus brings the letter bearer (τὸν τὰς ἐπίστολας κομίσαντα) to the people (εἰς τὸ πλῆθος), accuses him of forgery (πλαστογραφίαν ἐπικαλέσας, 50) and lying about Philip fighting against the Romans, and puts him to death. When the messenger does not return, Philip sends another messenger to deliver letters with the same message and to discover what has happened to the first messenger. Varus put him to death also (52). The episode reveals not merely the plausibility of a charge of forgery but also its gravity, given that death is deemed an appropriate punishment. The charge may not have been just, since Varus is seeking to consolidate power for himself, but the fact that he chooses to charge the {171|172} messenger with forgery demonstrates the importance of authenticity for the conveyance of a letter: to discredit the messenger is to discredit the message. The felicity of communication depends upon the security of the material, as can be seen with regard to forgery and to the manipulation of epistolary media to protect the messenger.

One of the most well-known references to forgery, and the second of my two examples, is in Bellum Judaicum 1.528–529 and its parallel at Antiquitates Judaicae 16.317–319. However, the reader must be careful with the narrator’s assertion of forgery. This discussion will proceed with a detailed comparison of these two texts to show that reports of forgery can be as shrouded in mystery as the materials that are deceptively created—they require the reader to question identity as much as may be expected of the internal recipient.

Both texts refer to a document: in Bellum Judaicum 1.528–529 the terms are ἐπιστολή and γράμματα, and in Antiquitates Judaicae 16.318, γράμματα. This nuance does not seem to make too much difference. It is possible that Josephus changes them simply for variatio, or because he is following Nicolaus to expand his material for the Antiquitates Judaicae; in any case, Josephus remains consistent with his term in the Antiquitates Judaicae. He later refers to the evidence with the diminutive γραμματίδιον ‘little tablet’, probably placing more stress on the letter as a physical artifact, thereby emphasizing its materiality. Furthermore, the variation of terms between the Bellum Judaicum and Antiquitates Judaicae might be explained by the fact that the letter occurs in indirect speech spoken by Alexander: “Alexander said the secretary imitated …” If Herod does not accept Alexander’s argument, perhaps he will accept his diminution of the evidence in question. In narratological terms, γραμματίδιον is focalized by Alexander—it reflects his interpretation of the document. In the Antiquitates Judaicae, the letter seems never to satisfy the reader as to its usefulness. An accusation is presented, and a letter is introduced as evidence. Here the reader may speculate that the truth will come to light. However, the letter “proves” only half of the {173|175} accusation: the commander receives Alexander and Aristobulus, but the provision of money from the royal coffers is never corroborated. This seems consistent with Alexander’s diminution of the evidence. Thus it seems that a tension is being created in the Antiquitates Judaicae account between the accusation and the quality of the evidence to support that accusation. The Bellum Judaicum is even less clear about how straightforward the evidence is: the text of the letter is presented, and then Alexander claims that it is a secretary’s forgery. Receiving the letter first, the reader may perhaps be convinced by its contents before learning that it could be a forgery. In the Antiquitates Judaicae, Josephus hints to the reader that it is “presumably” in Alexander’s hand.

In both texts, the reader response, given the ambiguity concerning the “forgery” at episode’s end, mimics the uncertainties of the internal audience. The lack of clarity suggested here as to the truth of the letter is supported by a development of tension in the plot. In Antiquitates Judaicae 16.319, an antithesis is created with a μέν … δέ construction, which forms the tension between Herod’s inner thoughts and the external action of Alexander that propels the narrative forward: having heard the letter, Herod has no doubt about his sons’ plot against him. The external reality of the following δέ clause conflicts with the thought presented in the μέν clause—Alexander still does not confess. This is different from the Bellum Judaicum narrative, where the same construction is used, but the μέν clause is the Antiquitates Judaicae’s δέ clause, and Bellum Judaicum’s δέ clause develops Alexander’s case (Bellum Judaicum 1.529). Here we have indirect discourse, as in Antiquitates Judaicae (16.319). In both texts we get an explanation of who the secretary Diophantus is, that he is known to be clever at imitating handwriting, and so we are led to think that Alexander may well be correct. The Antiquitates Judaicae keeps the reader from reaching a clear verdict: “it seemed (ἐδόκει) Diophantus was clever at such things.” Like the Antiquitates Judaicae, the Bellum Judaicum informs us that Diophantus is later put to death for a similar crime, but with the uncertainty of Herod’s justice, this is not the most convincing evidence.

With this framework for the reliability of embedded letters in mind, let us turn to Josephus’ use of embedded letters as evidence within episodes.

One important example of an embedded letter being used as evidence within an episode involves the accusations of Alexander and Aristobulus. Recall that Herod named three successors, Antipater, {175|176} Alexander, and Aristobulus (Bellum Judaicum 1.467–468; Antiquitates Judaicae 16.244–453), which temporarily calmed his turbulent domestic situation (Bellum Judaicum 1.458–466; Antiquitates Judaicae 16.133). These intrigues augment those provided by other members of Herod’s family (Bellum Judaicum 1.476–487; Antiquitates Judaicae 16.194–219). The intrigues of Herod’s eldest son Antipater in his father’s court form a set piece (Bellum Judaicum 1.495; Antiquitates Judaicae16.244-60), and eventually implicate his half-brothers Alexander and Aristobulus.

Both the Antiquitates Judaicae and Bellum Judaicum accounts involve accusations, but they differ slightly regarding the charge that is being denied. This is significant if a letter is to substantiate a charge in either account. In the Bellum Judaicum, Herod’s servants Jucundus and Tyrannus deny accusations made by Herod’s eldest son Antipater of conspiring with Alexander and Aristobulus against Herod. Herod tortures Jucundus and Tyrannus, but they confess to none of the accusations. In the Antiquitates Judaicae, the commander of the Alexandrion fortress denies the charge that he promised to receive Alexander and Aristobulus into the citadel and to furnish them with money from the royal treasuries.

Recall that both texts employ a letter to attest to the princes’ guilt: in the Bellum Judaicum, it is referred to as ἐπιστολή, while in the Antiquitates Judaicae it is γράμματα. Alexander is supposed to have been involved with the document in both narratives. In the Antiquitates Judaicae, the letter is written “presumably in Alexander’s hand” (16.319). The Bellum Judaicum simply says that the letter is addressed by Alexander to the commander of the Alexandrion citadel. Josephus gives the letter in oratio obliqua in the Bellum Judaicum and says that Alexander requested that the commander should receive him and his brother Aristobulus in the citadel after they had killed their father. {176|177} Josephus quotes the letter directly in the Antiquitates Judaicae, which has Alexander stating, “… just as you promised to receive us …”

The extent to which the letter substantiates the charges is problematic in both texts. The information that the quotation of the letter provides in the Antiquitates Judaicae is not very substantial, though the fact that it is quoted directly adds some color and dramatic flair to the account. In fact, the quoted letter may strike the reader as rather odd because it does not explicitly measure up to the charge. Interestingly, the weakness (ἀσθενεῖς) of the letter as evidence is noted in the Bellum Judaicum (1.530) but not in the Antiquitates Judaicae. As discussed above, a tension exists between the accusation and the material available to substantiate the charges. In the Bellum Judaicum, the narrator is not as explicit regarding the real machinations of the investigation. {177|178}

The assumption seems to have been made that the evidence is true: after the letters are exhibited, the Volscians confess their orders. They make no charges of forgery or trickery of any kind. The material record of letters presented publicly seems to have settled the case as to their true mission. Furthermore, some importance seems to attach to reading the evidence aloud. Merely making reference to intercepted letters and what they contain does not seem to be enough. At the same time, public reading seems a bit strange in that it raises the anger of the troops, and Postumius must calm them so that he can handle the situation in a tactically expedient way.

In summary, this episode reveals letters being used as evidence that the Volscians are not Roman allies and, by implication, that {180|181} members of the delegation are spies guilty of attempting to gather intelligence in a Roman camp under the guise of ambassadors. Only the letters are offered to attest to their guilt. And letters seem to have been enough in this text to elicit a confession and to raise the ire of the Roman troops. However, despite this affirmation of their reliability as evidence, letters prove vulnerable as a means of communication—they are intercepted—which would seem to inflect letters themselves with some ambiguity in terms of how readers might assess them. Yet letters can cut two ways: if interception does not cause the characters to be suspicious, it can give them the confidence of having caught the epistolary plotters “in the act,” as it were. In this case, although the characters do not seem to be at all suspicious of the epistolary evidence, Postumius appears not to think it will stand up against “universal law.”

The treatment of letters as at least partially reliable in the trial of the Volscian “spies” is similar to a narrative in which Xenophon has characters presenting a letter as evidence in a trial scene. In his Historia Graeca, after the battle of Arginusae, the victorious Athenians have not collected their survivors. The generals leave to provide reinforcements for Conon, who is besieged at Mytilene, while the captains Theramenes and Thrasybulus and company are to gather those whose vessels have been damaged in the battle (1.6.35). The narrator explains that they desire to do these things (ταῦτα δὲ βουλομένους ποῖειν, 1.6.35), but a wind and a great storm prevent them (ἄνεμος καὶ χειμὼν διεκώλυσεν αὐτοὺς μέγας γενόμενος, 1.6.35). {181|182}

In response, the Athenians depose the generals, with the exception of Conon, and they install Adeimantus and Philocles to serve with him. Archedemus, leader of the popular party, brings a charge against one of the generals in the battle, Erasinides, saying that he has stolen money and accusing him of misconduct as a general (κατηγόρει δὲ καὶ περὶ τῆς στρατηγίας, 7.2). The generals defend him by citing the greatness of the storm, and Timocrates moves that they be imprisoned and tried (7.3).

The outcome of this trial reveals the power of epistolary evidence. We may surmise that, because the letter does not actually read in the way that Euryptolemus suggests the generals had originally intended it would, they are found guilty. The fact that the letter is the focus of the rhetoric regarding the generals’ culpability (apart from the legality of the trial, which is the initial focus of their defense) reveals that letters are considered to be reliable and, perhaps, impartial pieces of evidence in this text. The reliability of this epistolary evidence would be undercut if the external audience were to consider letters superfluous: it may have been that the storm was to blame for the disaster, a simple explanation that would not require further substantiation.

Even in the texts from Dionysius and Xenophon, where letters are treated as reliable evidence and thus contrast with the Josephus text discussed above, the reliability of letters can subtly be called into question by other features in the narrative. In the Dionysius text, it is the narrator’s remark that Postumius is unwilling to punish the Volscian spies because he does not want to appear to be violating “universal law,” from which the epistolary evidence would presumably have released them. In Xenophon, it is the seeming simplicity of the generals’ defense, the unavoidable effect of a storm, that would obviate the need for epistolary evidence in the first place. While these are matters left largely ambiguous, which readers are left to work out, they are not unlike those in Josephus, where the nature of the epistolary evidence is not always as straightforward as it appears to be. The quality of the “Diophantus letter” is not so clear-cut, nor is the relevance of its message to the charges being reviewed in the Alexander-Aristobulus trial.

Having examined Josephus’ use of letters as evidence within episodes, let us turn now to that second use of attestational letters in Josephus’ texts, letters as evidence outside of the chronological sequence of events, or extra-episodic attestations. {183|184}

Letters are involved with a trial in the parallel texts found in Bellum Judaicum 2.23–25 and Antiquitates Judaicae 17.227–229, though they do not explicitly appear in the trial itself. I have excerpted the parts of the texts that involve letters from an account of a trial before Augustus, which attempt to determine the successor to Herod. I will briefly summarize the context of the letters before discussing them in detail. Five days before Herod dies (Bellum Judaicum 1.665), he orders the execution of his son Antipater, whom he had originally chosen to succeed himself (1.664). Because of this development, he again amends his wills (τὰς διαθήκας) and names Archelaus as his successor, with Antipas as tetrarch (1.664).

As Archelaus speeds to Rome, the procurator of Syria, Sabinus, meets him and his companions in Caesarea while en route to Jerusalem to take charge of Herod’s estate. The narrative intimates that he has intentions of taking advantage of disturbances at Jerusalem and the instability of Herod’s household by noting that his superior, Varus, appears at the same place and attempts to take control of Herod’s troops and fortresses (Bellum Judaicum 2.16–19). To further complicate matters, Antipas apparently sees the unrest as an opportunity to make {184|185} his own claim on the throne, based on the sixth will, and so departs for Rome himself (Bellum Judaicum 2.20; Antiquitates Judaicae 17.224).

After each makes his preliminary case in his own way, speeches are given, one against Archelaus by Herod’s nephew and another by Nicolaus, Herod’s courtier. Varus meets Augustus’ indecision with letters regarding an uprising in Jerusalem, of which Sabinus finds himself in the middle. The rest of the country, too, is in disorder, Josephus says (Bellum Judaicum 2.55–56; Antiquitates Judaicae 17.269–270), and several other usurpers appear. After Varus has put down the rebellions and Augustus has called a new council of advisors, Augustus finally decides that half of the kingdom will go to Archelaus as ethnarch and the rest will be divided between Antipas and Philip as tetrarchs (Bellum Judaicum 2.93–100).

The first character to appear in these texts is Sabinus. In both the Bellum Judaicum and Antiquitates Judaicae, he accuses Archelaus, who appears immediately previous to this episode at Bellum Judaicum 2.22. In the Bellum Judaicum, Sabinus’ activity is connected to the plot in that all of Archelaus’ previous enemies form the antecendent of αὐτοῖς. Sabinus has authority to intervene (if at all) because he is ἐπίτροπος of Syria (2.16), apparently subordinate to Varus, legatus of Syria. Josephus first introduces Varus in Bellum Judaicum 1.617 when he materializes at Herod’s court as the imperial representative in Antipater’s trial.

In this particular episode, then, Josephus does not have Augustus use the letters in the trial scenes themselves, but as preparation for the trial. Archelaus, Varus, and Sabinus write letters to Augustus to provide background and supporting evidence for his decision. Thus letters here are not called as “witnesses” or used by witnesses in the trial, but are {185|187} considered as part of Augustus’ decision, in addition to live witnesses, since some of the important players are absent from the trial.

Another “trial” for which letters are important takes place at Bellum Judaicum 1.620–640 and Antiquitates Judaicae 17.93–133. In this episode, though Josephus appears to have a letter serve as evidence in a trial, he actually uses the letter to support his own interpretation of the events presented, which differs slightly from the previous example of an extra-episodic embedded letter. In fact, this narrative also contains an episodic letter, a feature that will throw the extra-episodic letter into further relief. The conflict between the letters involved in this trial is made even more dramatic by the dubiousness of letters as a regular feature of Josephus’ texts in general, particularly those concerning the Roman period. I will give the background to the text before discussing the function of the embedded letters.

The assumption of Antipater’s guilt is not a new theme in the Bellum Judaicum narrative. Herod does banish him (1.433), preferring to promote Alexander and Aristobulus, his younger sons by Mariamme I. When those two begin scheming against Herod (Bellum Judaicum 1.446; Antiquitates Judaicae 16.67), he recalls Antipater. Then Antipater’s schemes begin. The narrator’s interventions assert his guilt in an episode in which Herod gives benefactions to the children of the murdered princes Alexander and Aristobulus, aged five and four respectively. Herod shows pity with resolution: πειρῶμαι δ᾿, εἰ καὶ πατὴρ ἐγενόμην ἀτυχέστατος, πάππος γοῦν γενέσθαι κηδεμονικώτερος (“I will make an attempt, if I have been a most unfortunate father, to be at any rate a more provident grandfather,” Bellum Judaicum 1.556). Within a few lines of Herod arranging the marriages of his grandchildren, Antipater, a member of the audience, immediately becomes “cold and hard” (ἐπαχνώθη δ᾿ εὐθὺς, 1.559), and “it was clear to everybody {188|189} that he was suffering pain” (δῆλος ἦν ἅπασιν ὀδυνώμενος, 1.559). These statements serve to confirm the narrator’s assertion about what “everybody knew”: πάντων ἐπισταμένων, ὅτι τὰς διαβολὰς τοῖς ἀδελφοῖς πάσας ἐπισυντάξειεν οὗτος (1.552). The narrator focalizes Antipater’s mind (with ὑπολαμβάνω, συλλογίζομαι, γιγνώσκω) to unveil his schemes (1.559–560).

Herod suspects Antipater’s guilt (1.564), and the Antiquitates Judaicae parallel maintains in essence the same story (17.12–18), but with two important differences. First, the narrator presents a second rationale for Herod’s betrothing of his orphaned grandchildren—so that Antipater will feel more pity on them because of their marital connections. And, second, at this stage, the audience has not yet learned that Herod is suspicious of him. In fact, the audience learns (17.18) that Herod is ignorant of Antipater’s wiles. However, an inconsistency appears at Antiquitates Judaicae 17.40, where Herod’s sister Salome reports to him Antipater’s and Pheroras’ intrigues, and the narrator says, “he already had put many things together for himself” (τῷ ἀδελφῷ ἀνέφερεν συνιέντι μὲν ἤδη τὰ πολλὰ καὶ δι᾿ ἑαυτου). Regardless of the timing of Herod’s suspicions, in both texts the narrator apparently intends to leave the narratee without doubt about Antipater’s designs.

A thorough summary of the trial is essential before discussing the letters because the placement and function of the embedded letters frame the entire trial. In both the Bellum Judaicum and Antiquitates Judaicae, a council is called for Antipater’s trial. In Bellum Judaicum 1.620, Herod gathers his relatives and friends and presides with Varus. In Antiquitates Judaicae 17.93, together Varus and Herod call a council. In the immediately preceding narrative (Bellum Judaicum 1.617), Antipater is the focus; having returned from Rome, he enters his father’s presence and Varus, the legatus of Syria, is at the time within the palace (ἔτυχεν δ᾿ ἔνδον ὢν Οὔαρος ὁ τῆς Συρίας ἡγεμών). In the Antiquitates Judaicae, Josephus has given more detail regarding Varus’ identity, with a fuller name and his order in succession: as legatus of Syria, he follows Saturninus. Saturninus appears in the trial of two other Herodian sons, Alexander and Aristobulus, in both parallel accounts (Bellum Judaicum 1.541; Antiquitates Judaicae 16.369). The repetition of detail is probably explained by the fact that mention of the legatus in the present narrative is the first in Book 17; in the Bellum Judaicum, the “current” legate is simply part of the long narrative of Book 1.

After Varus’ introduction, witnesses are introduced. The letter that will be discussed in due course appears when its messengers are introduced as slaves of Antipater’s family (belonging to his mother [Bellum Judaicum 1.620] or perhaps to Antipater [Antiquitates Judaicae 17.93]). The letter, addressed by Antipater’s mother to Antipater, warns Antipater to stay away from his father because Herod has discovered everything. In the Bellum Judaicum, Antipater is brought in with these slaves (1.621); in the Antiquitates Judaicae Antipater never “enters,” but his presence is made known when he falls before his father and begs for a fair opportunity to prove his innocence (17.94; cf. Bellum Judaicum 1.621).

The form of each letter seems to fit the internal logic of the particular narrative in which it appears. In short, the majority of the speech material in the Bellum Judaicum is presented in oratio recta, so the letter is presented in oratio recta. In Antiquitates Judaicae, most speech material is given in indirect discourse, and the letter is presented the same way.

By contrast, Nicolaus appears in Antiquitates Judaicae on at least 18 other occasions, including parallels to the Bellum Judaicum but also obviously in unique circumstances. This seems to be because Josephus shows an interest in Nicolaus’ craftiness in carrying out assignments. For example, in Antiquitates Judaicae 16, Nicolaus goes to Rome to reconcile Herod and Augustus, who have a falling-out over Herod’s attack on Arabia. The Arab king Obadas’ executive Syllaeus (16.220) takes advantage of an uprising in Trachonitis while Herod is away in Rome (16.271). The inhabitants of Trachonitis, thinking that Herod has died, begin their habit of brigandage once again. In response, Herod’s generals put down the uprising. Forty leaders from Trachonitis flee to Arabia and are received by Syllaeus, who provides them with resources. When Herod arrives back in Judaea and hears what has happened, he goes to war against the Arabs to set right the interference of Syllaeus in his affairs, which is allowing brigands to raid the country. Herod also demands the repayment of a loan given to Obadas. Syllaeus flees to Rome to accuse Herod before Augustus (16.282–292), leading to Augustus’ displeasure with Herod.

Herod dispatches Nicolaus to Rome to mend relations (16.299). While in Rome, to ensure his success at setting matters straight between Herod and Augustus, Nicolaus takes the opportunity to accuse Syllaeus before Augustus. This builds confidence in Nicolaus’ own testimony on Herod’s behalf (16.335–356)—a detail that does not appear in the Bellum Judaicum. So Nicolaus presents various accusations against Syllaeus, which include causing the death of Obadas and other Arabs, borrowing money for criminal acts, and committing adultery in Arabia and in Rome (16.338–340). He then proceeds to tell Augustus that Syllaeus has lied to him about Herod’s activities (16.340). With this context, Augustus is obviously much less willing to trust Syllaeus’ testimony, and eventually Nicolaus wins the day for Herod. Thus Josephus reveals himself to be very aware of Nicolaus’ “spin” on events in Herod’s favor. He says as much about Nicolaus as a historian, citing this as a reason to distrust his accounts of the Herodian court (16.183–186), though he relies on them anyway. And because Josephus even acknowledges that Nicolaus had contrived charges for Herod to justify the murder of his wife Mariamme I and her sons (16.185), could we not also suspect that Nicolaus might have made a speech in which he contrived charges {196|197} against Antipater? With this in mind, the expanded speech that Josephus presents as Nicolaus’ may be fitted with this greater awareness of his ability to construct clever arguments and to be persuasive in a court of law.

To give a further example, Nicolaus appears in the Antiquitates Judaicae after the trial of Alexander and Aristobulus (16.370–372). Herod meets Nicolaus sailing into the port at Tyre from Rome and gives him a report of the trial at Berytus, asking for some indication of the attitude toward his sons at Rome. Nicolaus proceeds to give Herod an assessment, apparently in line with Roman opinion, of how his sons should be punished. Josephus gives the speech, as in our text, in oratio recta. However, neither the speech nor the speaker appears in the Bellum Judaicum (cf. Bellum Judaicum 1.543). Thus it seems that Nicolaus’ Antiquitates Judaicae speech fits into Antiquitates Judaicae’s characterization of Nicolaus, just as his Bellum Judaicum speech matches his lower profile in that work.

Running through both accounts of the trial, there is a tension between two styles of argumentation: emotiveness—it might be called apoplexy—and the employment of competing forms of evidence. Herod’s tirade so overwhelms him that he is apparently unable to finish. Antipater then begins his own defense by offering what are presented as two reliable pieces of evidence, a letter from Augustus and a testing of his own testimony under the pressure of torture. He apparently is fairly certain of his innocence, and a letter, particularly one with the credibility that the name of Augustus would inevitably carry, would seem to be as reliable as evidence can get. Both accounts note that Herod is moved to some extent by Antipater’s evidence. Nicolaus then offers, in his long, reasoned speech, the testimony of many witnesses against Antipater. As has already been discussed, Nicolaus is a regular character in this part of the Antiquitates Judaicae, and his other appearances seem to tell against his consistent trustworthiness. These details within the narrative put the reader in a difficult position: the first, extra-episodic letter appears to demonstrate Antipater’s guilt outright, but the accumulating evidence within the trial seems to cut the other way.

Augustus’ letter is not quoted in either narrative. Antipater refers to it in the course of his speech to support his claims that he had been faithful to his father and could not have plotted against him. Antipater possesses the letter in the Bellum Judaicum, and he offers it to Herod. In the Antiquitates Judaicae he refers to the letter that had been sent. In both cases, Antipater argues that Augustus’ letter is more reliable than the other testimony that has been offered, and, it must be noted, comparing the testimony of the princeps to that of some slaves and other tortured informants is certainly a clever argument.

The juxtaposition of the two letters in the trial presents a conundrum to the reader, and that conundrum needs further exploration. The internal audience hears of (and in the Bellum Judaicum probably sees) {198|199} the letter from Augustus, and this, along with the rest of Antipater’s speech, moves them to compassion (Bellum Judaicum 1.636; Antiquitates Judaicae 17.106, although no explicit comment is made about the impact of the “documentary evidence”). However, the reader has read the initial letter, which appears to incriminate Antipater, and now a second document presents a different story. It seems that Josephus has deliberately given his reader more to analyze than has been provided to the internal audience. Admittedly, the letter is closely related to the trial. But it is strange that Herod makes no mention of it in his speech, as he had in the previous trial, nor does Nicolaus mention the letter. By excluding the participants, among them Nicolaus, who is both involved in the events directly and provides source material for Josephus, the historian seems to be claiming for himself a greater authority than the participants themselves have from their eye-witness perspective. He reserves for the narrator’s voice the introduction of evidence that bears on the reader’s analysis of the trial, and, although the initial letter would seem to be essential information to mention in the trial itself, the characters fail to do so.

The narrative that follows is not reassuring about the straightforwardness of letters: it refers to a forged letter by Antipater, purportedly from Herod’s sister Salome, which slanders Herod. Antipater allegedly bribes Acme, a servant of Augustus’ wife Livia, to say that she has found the letters among Livia’s documents and is forwarding them to Herod because of her goodwill (Bellum Judaicum 1.641–643; Antiquitates Judaicae 17.134–141). Herod is then even more convinced of Antipater’s guilt, so he begins further action against him and at the same time is overtaken (mysteriously) by a serious illness (Bellum Judaicum 1.645; cf. Antiquitates Judaicae 17.144, where he is “deeply pained”). Nevertheless (μέντοι), he writes a letter to Augustus regarding the discoveries.

We have seen that embedded letters can be used in a trial as direct evidence to substantiate a charge. Embedded letters can also be used as relevant to a trial, while not necessarily being included as part of that trial. These two uses I have called episodic and extra-episodic. In Josephus, especially in the Herod narratives, letters used as evidence can be considered as being of dubious quality, although in other authors, such as Xenophon and Dionysius, letters functioning in this way seem to be regarded as more reliable. This is not a strict generalization. Josephus has Antipater appeal to letters from Augustus at {204|205} Bellum Judaicum 1.633; because of their apparent source, these letters lend an air of credibility to Antipater’s case, though Josephus’ narrative shows that Antipater’s guilt is all but certain by the time of his plea before Herod. In any case, letters in trials in Josephus are generally to be regarded with suspicion and to be considered with extreme care.

However, the extra-episodic quality of letters appears to be largely unique in the genre of history. This may well be because most other historians, Thucydides, for example, use letters largely as documentary evidence. Regarding an epistolary poetics in Josephus, then, both broad categories of attestational epistolary intratexts, episodic and extra-episodic, avoid such neat generalizations as “deceptive” or “reliable.” Instead, they require that the reader take them intratextually, i.e. in the context in which they appear as portions of text related to other portions of text, and, as the worked examples have shown, as related to the whole text(s). {205|}