Kaldellis Greek, Roman, and Byzantine Studies 49 (2009) 585–616 © 2009 GRBS The Date and Structure of Prokopios’ Secret History and His Projected Work on Church History Anthony Kaldellis The date of the Secret History T WAS Prokopios’ editor, Jakob Haury, who argued (in 1891) that the Secret History was written or completed in 550, and most scholars since then have accepted this conclusion.1 In 1987, Roger Scott argued that two events in the Secret History could be dated to after 550, though he admitted that the case was water-tight for neither one (these were Justinian’s debasement of the coinage, 25.11–12, and his requirement that Jews always celebrate Passover after Easter, 28.16–18).2 Scott’s arguments were convincingly answered by Geoffrey Greatrex and J. A. S. Evans in 1994 and 1996 re- spectively.3 The case seemed closed, but in 2005 Brian Croke, conceding that Scott’s proposals had indeed been answered, argued on different grounds that a later date (viz. 558/9) was still possible.4 The purpose of the present article is to re- establish the traditional (or Haurian) date, this time not merely by (again) putting out the fires lit by the proponents of 558/9 1 J. Haury, Procopiana (Augsburg 1891 = Programm des Königlichen Real- gymnasiums Augsburg für das Studienjahr 1890/91) 9–27. 2 R. Scott, “Justinian’s Coinage and Easter Reforms and the Date of the Secret History,” BMGS 11 (1987) 215–221 (“unfortunately they do not solve all the problems,” 217). 3 G. Greatrex, “The Dates of Procopius’ Works,” BMGS 18 (1994) 101– 114, esp. 103–105; J. A. S. Evans, “The Dates of Procopius’ Works: A Re- capitulation of the Evidence,” GRBS 37 (1996) 301–313, esp. 308–312. 4 B. Croke, “Procopius’ Secret History: Rethinking the Date,” GRBS 25 (2005) 405–431 [hereafter “Croke”]. I 586 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY but by providing new, positive, and hopefully conclusive evi- dence that the Secret History was completed in 550; to urge that Prokopios’ system of cross-references clarifies the relation among his different works (and even different sections of the Secret History); and to argue that the only work that he was planning on writing in 550 was an Ecclesiastical History in the general mode of the Secret History. The argument will also clarify the two different and potentially confusing methods by which Prokopios dated Justinian’s reign—or, to be precise, not so much his reign as the number of years during which he wielded power.5 It must be emphasized at the outset that with one exception (the ambiguous matter of the “32 years”), Croke does not pro- vide positive arguments in favor of the 558/9 date. His argu- ment aims to establish merely that that later date is still possible and it is in this spirit that he suggests that three of the episodes narrated by Prokopios may have occurred after 550; he does not argue on internal or external grounds that they must have done so.6 I will first offer a new interpretation of the 32 years consistent with the 550 date (in fact, that requires that date), effectively removing the need to entertain the possibility of 558/9; and then I will offer positive evidence in favor of 550. At four places in the Secret History, Prokopios states that Jus- tinian had already ruled for 32 years: 18.33, 23.1, 24.29, and 24.33. Now Justinian became emperor in 527, which would date these passages to 558/9. The proponents of 550, begin- ning with Haury, have maintained that Prokopios was includ- ing the reign of Justinian’s uncle Justin (518–527) in the 32- year span, dating the composition of the work to 550, the very year when Prokopios finished the Wars (1–7) and released it to the public.7 Croke concedes that Prokopios did regard Justinian 5 The present author is preparing a new translation of and commentary on the Secret History. All passages quoted here are from that translation. 6 Croke 417–420. The three episodes will be discussed individually below. 7 Those who believe that Prokopios meant that he was writing in the 32nd year (since Justin took power) conclude that it was 549/50, while those who believe that he meant that that 32nd year was completed date the work to ANTHONY KALDELLIS 587 as the power behind his uncle’s throne and that he did include events that took place under Justin within the narrative of his indictment of Justinian.8 However, he adds, Prokopios also makes it clear in the Secret History that the two reigns were distinct by sometimes stating specifically that events occurred in the reign of the one or the other, so he obviously knew that Justinian became basileus only in 527. Moreover (and crucially), in the Wars Prokopios dates various events of Justinian’s reign from 527, exactly as other authors of the period did and as the emperor himself in Novel 47 had required official documents to do. So we should not “accept at face value Prokopios’ slander that Justinian effectively displaced the power of Justin and dominated him.” Moreover, his readers would have known that Justinian’s reign began in 527 and so “to start from some- where else … would be perverse.”9 Whether we, as independent historians, accept the slander at face value or not does not change the fact that much of the invective of the Secret History relies on it: Justinian, Prokopios would have us believe (probably believing it himself ), was re- sponsible for many of the evils that occurred under Justin. This was the outlook that shaped the text and, accordingly, accounts for the 32-year period of power that is attributed to an emperor who, by 550, had reigned for only 24 years. Prokopios makes the distinction between the period of Justinian’s administrative power (under Justin and after) and his kingship (starting in 527) explicit in a passage that provides the key to the riddle, if we attend closely to its vocabulary and then notice that this vocab- ulary is used consistently throughout Prokopios’ works. To my knowledge, this observation has not yet been made. At 18.45, the second section of the Secret History climaxes with a list of natural catastrophes that struck the cities of the empire during the period under review (which included the reign of Justin, as ___ 550/1. I think the latter is probably more correct, but will refer to 550 for simplicity’s sake. 558/9 will stand as it is Croke’s consistent position. 8 In “Justinian under Justin: Reconfiguring a Reign,” BZ 100 (2007) 13– 56, Croke tries to work around the bias imposed by Prokopios’ presentation. 9 Croke 408–416, quotations from 416 (I have corrected the Latinization of Greek names in quotations throughout). 588 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY we shall see): “That was the extent of the destruction of human life that occurred first while Justinian was administering the state on behalf of the Romans and later when he held sole imperial authority (τοσούτων μὲν ἀνθρώπων ἐγένετο φθόρος, Ἰουστινιανοῦ πρότερον ῾Ρωμαίοις διοικουμένου τὴν πολιτείαν καὶ ὕστερον τὴν αὐτοκράτορα ἀρχὴν ἔχοντος).” In other words, the period under review in the Secret History consists both of the time when “Justinian was administering the state,” i.e. the reign of Justin (518–527), plus the time when “he was sole emperor” (527 to the present). As it happens, the 32 years are first mentioned in the Secret History only two pages before this passage, in the lead-up to this climax (18.33), and specifically in connection with the violence of the factions. The proximity of these two temporal markers—the 32 years and the two, care- fully distinguished periods of Justinian’s power—flanking a coherent passage on the destruction caused by this one man, indicates that they are equivalent expressions. In other words, 32 = years of administration + years of sole rule. This conclu- sion is buttressed by the fact, noted by Haury,10 that elsewhere in the same text Prokopios blames Justinian for the violence of the factions that took place during Justin’s reign (Secret History 7 and 9.29–46). Croke attempts to deny this by noting that fac- tional violence in fact occurred later too, i.e. during Justinian’s sole reign, and so “this not need be the case at all” (411). But it absolutely must be the case. Croke himself earlier in his article stated that these incidents of factional violence occurred under Justin (409). Let us review the evidence. In ch. 7 of the Secret History it is clear that Prokopios is dis- cussing the reign of Justin. After all, that chapter is the narrative continuation of ch. 6, which looks at Justin’s early career and reign. While Prokopios does look forward here to one event that happened later (the conquest of North Africa: 6.25), the entire narrative of these two chapters belongs to the reign of Justin. In fact, Justinian is introduced into the narrative at 6.19 with words that are precisely (and, we shall see, tech- nically) consistent with the distinction made at 18.45 between 10 Haury, Procopiana 12–14, 16. ANTHONY KALDELLIS 589 his period of power under Justin and his reign after 527: “his nephew Justinian, who began to administer the entire state while still young, was the cause of disasters for the Romans (ἀδελφιδοῦς δὲ αὐτῷ Ἰουστινιανὸς νέος ὢν ἔτι διῳκεῖτο τὴν ἀρχὴν ξυμπάσαν καὶ γέγονε ῾Ρωμαίοις ξυμφορῶν αἴτιος).” The technical term is διῳκεῖτο, which is what Prokopios at 18.45 says that Justinian did during his uncle’s reign. Thus the fac- tional violence that Prokopios describes in ch. 7 must belong to the reign of Justin, and that violence is included in the 32 years at 18.33. Not only is ch. 6 about Justin, ch. 8 begins by remind- ing us that we are still under Justin, who was “unaware of what was going on around him, even though he was constantly wit- nessing such scenes [i.e., involving the factions] in the hippo- drome” (8.2). There is another reason why the scenes of factional violence described in ch. 7 must belong to the reign of Justin. Prokopios returns to the topic at 9.29–46. Here we are still under Justin because Justinian has recently met Theodora and has not yet married her; moreover, in the narrative itself it emerges clearly that Justinian is not “the emperor.” Prokopios says that when Justinian fell ill “the militants were causing their usual trouble, doing all the things that I explained above, and they killed a certain Hypatios” (9.35). In other words, the murder of Hypa- tios, which without any possible doubt occurred under Justin, happened after “all the things that I explained above,” i.e., the factional violence described in ch. 7. Therefore, to conclude this part of the argument, the first mention of the 32 years (18.33), which refers to the factional violence, must include the reign of Justin, because the overwhelming majority of the material on the factions in the Secret History (if not all of it) be- longs to the reign of Justin. What then of Croke’s objection that in the Wars Prokopios always (and properly) dates events in Justinian’s reign from 527? In fact, even according to many passages of the Secret History Justinian was not basileus until 527.11 So why would Prokopios want to confuse his readers by introducing a differ- 11 Croke 413–415. 590 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY ent system of dating (from 518) when he elsewhere consistently dates events from 527? It is at this point that we must attend closely to his language, for it was not his intention to set up the 32-years as a rival regnal system. Both opponents and defenders of 550 are wrong when they say or imply that in the Secret His- tory Prokopios dated Justinian’s reign from 518.12 There is never any ambiguity about which date Prokopios means whenever he says that something occurred during Justinian’s reign: the latter began in 527. Yet in every passage of the Wars cited by Croke (413 n.27) in which Prokopios dates an event to year x of Jus- tinian’s reign and dates it from 527, Prokopios specifies that this was the year since Justinian took hold of “sole imperial power (τὴν αὐτοκράτορα ἀρχὴν ἔχων)” (12 instances) or the year of his basileia (3 instances). The key passage of the Secret History (18.45), on the other hand, carefully distinguishes be- tween “administering the state” (διοικουμένου τὴν πολιτείαν) and “holding sole imperial command” (τὴν αὐτοκράτορα ἀρ- χὴν ἔχοντος). It is only the latter that Prokopios consistently and correctly dates from 527, both in the Wars and the Secret History. But one did not have to be emperor to “administer the state.” The verb would be used for centuries in Byzantium to designate what we might call the emperor’s “prime minister” or “chief of staff,” the man (or eunuch) who held real power because he was trusted or because the emperor was weak.13 In the Secret History too, someone could “administer the state” (or the basileia or ta pragmata) who was not an emperor, and that someone was Justinian under Justin. In fact, in the Wars and Buildings the combination of the verb dioikeô with any of those objects always refers to the power behind the throne and not exclusively in connection with Justinian (though in his case too, 12 Haury, Procopiana 15–16, quoted by Croke 407 n.11; also Scott, BMGS 11 (1987) 215, 221 n.28; Greatrex, BMGS 18 (1994) 102. This has confused others, e.g. F. Haarer, Anastasius I: Politics and Empire in the Late Roman World (Leeds 2006) 219, who does not know whether to go with 518 or 527. 13 Many instances of the verb (often in conjunction with politeia, as in Secret History 18.45) are cited in I. Christou, Αυτοκρατορική εξουσία και πολιτική πρακτική: Ο ρόλος του παραδυναστεύοντος στη βυζαντινή διοίκηση (Athens 2008). ANTHONY KALDELLIS 591 which proves that Prokopios was consistent in his terminology across works).14 This explains why at 18.37 Justinian is said to have “administered the state,” presumably for the full extent of the just-mentioned 32 years: Prokopios is referring collectively here to all of his years in power, both as emperor and before. The 32 years clearly represented the sum of the two different periods. The other passages in the Secret History where the 32 years are mentioned confirm this conclusion. In none of them does Prokopios state that these are 32 years of Justinian’s basileia (or use any of the expressions that indicate imperial rule). In fact, at 24.29 he makes it clear that the 32 years began “from the moment that this man began to administer the state (ἐξ ὅτου δὲ ἀνὴρ ὅδε διῳκήσατο τὴν πολιτείαν),” the very verb that refers to Justinian’s power under Justin at 6.19 and 18.45. Thus, along with the inclusion of episodes of factional violence under Justin in the period of 32 years, this clinches the argument: the 32 years are not a confusing alternative regnal dating system. It is part of the argument (or “slander”) of the Secret History re- garding Justinian’s power. 14 Wars 3.9.5 (Justinian under Justin οὔπω μὲν ἥκοντι ἐς βασιλείαν, διοι- κουμένῳ δὲ αὐτὴν κατ’ ἐξουσίαν), 5.1.2 (Orestes under Romulus Augustulus τὴν βασιλείαν … διῳκεῖτο), 5.2.3 (Amalasountha under Athalaric τὴν ἀρχὴν διῳκεῖτο), 7.35.17 (Audoin under Waldari τὴν ἀρχὴν διῳκεῖτο); Buildings 1.3.3 (Justinian under Justin τὴν βασιλείαν κατ’ ἐξουσίαν αὐτὸς διῳκεῖτο). There is one passage in the Secret History that is problematic in this regard. At 12.29 one MS. says that a dancing-girl named Makedonia would write letters to Justinian “while Justin was still administering the basileia (Ἰουστι- νιανῷ γράφουσα ἔτι τοῦ Ἰουστίνου διοικουμένου τὴν βασιλείαν)” while two MSS. have “while Justinian was still administering the basileia.” Haury rightly preferred the reading Ἰουστίνου over Ἰουστινιανοῦ, for the ἔτι would other- wise make no sense. He then emended διοικουμένου to διοικουμένῳ on the grounds that διοικέω must refer to Justinian in the context of Justin’s reign: Ἰουστινιανῷ γράφουσα ἔτι τοῦ Ἰουστίνου διοικουμένῳ τὴν βασιλείαν, which is consistent with all the passages cited above. Croke suggested that the original “should probably stand” (414–415 n.28). But Prokopios’ practice in all his other works supports the emendation. What would it mean for an emperor to be “administering his own basileia,” and why did Prokopios not simply use one of his customary phrases for saying that this happened in the reign of Justin? Moreover, it is easy to see how a copyist could have as- similated a dative participle to the genitive noun right before it. 592 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY Without the 32 years, no positive reason remains for be- lieving that the Secret History must (or even may) date to 558/9. There is, however, at least one passage that positively proves that the text could not have been written as late as 558/9, but this passage has been overlooked in the debate because it pre- supposes that one has read the third section of Secret History (chs. 19–30) against the background of Justinian’s edicts. I offered such a reading in my book on Prokopios, where I argued that this section of the work was an innovative historical commentary on Justinian’s laws.15 Specifically, at 23.1 Pro- kopios says that Justinian had never once in 32 years granted a general cancelation of tax arrears to the subjects of the Roman empire. Now, in the preface of Novel 147 (of 553), Justinian stated that he had granted “philanthropy” and tax relief to specific petitioners (attested at Secret History 23.6 and in other sources)16 but admitted that he not yet granted a general can- celation of arrears, which is what he was now proposing to do by this Novel. This proves that Prokopios was writing before 553, which means, accordingly, that the 32 years must have come to an end before then.17 Now Prokopios did rhetorically exaggerate some of his ac- cusations in the Secret History, and in dealing with Justinian’s specific edicts in the third section of the text he may have sometimes twisted his interpretation of their provisions and Justinian’s motivation in issuing them (I say may have because much in the Secret History that we might have rejected on its own is confirmed by independent sources, and also because 15 A. Kaldellis, Procopius of Caesarea: Tyranny, History, and Philosophy at the End of Antiquity (Philadelphia 2004) 150–154, 223–228. 16 E.g., Ioannes Malalas Chron. 18.28 (to Laodikeia after an earthquake), 18.29 (tax remittance to Antioch, Laodikeia, and Seleukeia), etc.; ed. I. Thurn, Ioannis Malalae Chronographia (Berlin/New York 2000); transl. E. Jeffreys et al., The Chronicle of John Malalas (Melbourne 1986). Also Kyrillos of Skythopolis Life of Sabas 69–73; transl. R. M. Price (with notes by J. Binns), Lives of the Monks of Palestine by Cyril of Scythopolis (Kalamazoo 1991) 182–187. 17 This had already been noted by F. Conca, Procopio: Storie segrete (Milan 1999) 281 n.290; J. Signes Codoñer, Procopio de Cesarea: Historia Secreta (Madrid 2000) 292 n.278. ANTHONY KALDELLIS 593 Prokopios knew a lot more about what Justinian was up to than we do; besides, often it is his word against that of the emperor, so we must keep an open mind). For example, at 29.19–20 Pro- kopios says that Justinian issued an edict according to which “when a councilor died without male issue, one fourth of his estate went to the heirs while all the rest went to the imperial treasury and the city council. Yet there was no precedent in all of history for the treasury or an emperor taking a share of a councilor’s property.” This may refer to Novel 38.1 (of 535), which, however, does not assert the claims of the imperial treasury in the matter. Prokopios otherwise has its provisions right, so I will allow that he may have known something about its enforcement that led him to conclude that it was a way to (indirectly perhaps) enrich the imperial treasury. There may have been a lost constitution behind his text. At any rate, I doubt that he completely invented the treasury’s involvement. A recent study has argued that he reports the legal facts cor- rectly here but in a biased way to make Justinian look bad.18 Nowhere in the Secret History does he flat out lie about Jus- tinian’s acts, for example by attributing to him edicts that never existed or denying that he had ever done things that in fact he had. In the case of Priskos the forger of Emesa (28.1–15), Pro- kopios is precise and correct regarding the provisions of CJ 1.2.23 (of 530) and Novel 9 (of 535). Besides, if he were to have asserted after 553 that Justinian had never issued a general cancelation of tax arrears, his text would have lost all credibility among contemporary readers. The latter may not have known for sure whether Justinian wore his head at all times when he roamed the palace at night, but they all knew when they had been given tax relief. Thus, the evidence of Novel 147 is decisive for dating the text to before 553. One of the traditional arguments in favor of 550 is that the Secret History refers to no event that can be dated with certainty or even likelihood after that year. In response to this challenge, Croke has attempted to show three instance where an event or 18 J. Beaucamp, “Le droit successoral relatif aux curiales: Procope et Justinien,” in S. Puliatti and A. Sanguinetti (eds.), Legislazione, cultura giuridica, prassi dell’impero d’oriente in età giustinianea (Milan 2000) 379–395. 594 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY situation in the text could have taken place after 550.19 It must be emphasized that nothing in his argument compels us to so date them, so even if his interpretations are sound that would not necessarily affect the case for 550. Moreover, the odd features of the episodes to which he draws attention in some respects strengthen the case for 550. First, Croke argues that “Prokopios’ characterization of [Jus- tinian’s quaestor] Konstantinos [at 20.20–23] gives the impres- sion that at the time of writing he had been in position more than merely a few months or just a year” (he was appointed in 549). I do not see, however, why this needs to be the case. Prokopios says nothing about Konstantinos other than he was corrupt and aggressively ignored people who he did not think would bring him profit. In fact, Prokopios explicitly says that he “amassed much money in a short time (χρόνου ὀλίγου),” which is an indication that we are dealing with a compressed time-frame. 550 works fine. Moreover, we can compare to this characterization what Prokopios says in his lead-up to the Nika Riots about the corrupt practices of the prefect Ioannes the Kappadokian (Wars 1.24.11–15). He would have us believe that Ioannes’ behavior was so despicable that the rioters were (partly) motivated by grievances against him. Yet Ioannes had been in office for only a few months.20 Here too Prokopios notes that he “he acquired a vast fortune in a short time (χρό- νου ὀλίγου).” In any case, we must accept that this picture is correct because the rioters did in fact demand his dismissal from office (and we know this from other sources too). It was evidently possible to work a lot of corruption in a short time. In fact, Prokopios’ account of Ioannes’ corrupt practices before January 532 is even more intense than what he says about Konstantinos during 549–550. There is no inherent problem, then, in the time-frame required for Prokopios’ denunciation of Konstantinos. The second episode concerns the appointment of Malthanes /Marthanes to some position of authority in Kilikia in the late 19 Croke 417–420. 20 PLRE III 628, Ioannes 11. ANTHONY KALDELLIS 595 540s with a mandate to suppress violence (29.27–38). Probably he was governor, but we cannot be sure. The Acts of the Councils attest that he was in Mopsuestia in June 550 for a local synod, but Croke’s suggestion that this may have been before the events at Tarsos narrated by Prokopios is based on the absence of the office of dux from the titulature that ac- companies his name in the Acts.21 But this is a weak argument (especially when placed before the mass of evidence for 550 that it must now overcome by itself). We cannot, for instance, be sure that the Acts were as “punctilious” in recording the full titulature of every official as Croke states. Moreover, given that Justinian was reforming the governorships in question during those very years, we cannot be sure what titles would have cor- responded to whatever position Marthanes held in the late 540s or in June 550; perhaps the compilers of the Acts did not know themselves. Besides, we do not know the vicissitudes of the man’s career: he may in fact no longer have been governor when he was at Mopsuestia in June 550 (or not held whatever position Justinian had given him in the late 540s). It is possible that when he was recalled to Constantinople after the distur- bances at Tarsos (29.34–35), he lost whatever position he had or was replaced. Perhaps he was sent back to Kilikia with a different mandate for the synod of 550. The third episode concerns the monopoly over the silk in- dustry that was established by Petros Barsymes when he was comes sacrarum largitionum (25.20–25). But not only does this fail to establish that 558/9 is a possible date, it actually positively proves that the Secret History was written before 555. At question are the implications of the following passage (25.22–23): καὶ βασιλεῖ μὲν ἐνθένδε μεγάλα χρήματα ἔφερεν, αὐτὸς δὲ περι- βαλλόμενος πλείω ἐλάνθανεν, ὅπερ ἀπ’ αὐτοῦ ἀρξάμενον ἐς ἀεὶ ἔμεινε. μόνος γὰρ ἐς τόδε τοῦ χρόνου ἔμπορος τε ἀπαρακα- λύπτως καὶ κάπηλος τοῦ ἐμπολήματος τοῦδε καθίσταται. And from this source he made a large profit for the emperor as well, though he also managed to get away with embezzling an even greater sum for himself. This practice, having begun from 21 Croke 419–420; see PLRE III 835–837. 596 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY that, then became permanent. For he openly remains the sole merchant and retailer of these goods to this very day. This, Croke argues, “would seem to imply that Petros is still en- gaged in the dye trade and therefore still comes after 550. Petros ceased to be comes around 555 when he became praetorian prefect for the second time, but Prokopios does not say he was definitely still comes at the time of writing. He may therefore be referring to the period after 555 when Petros was no longer comes but as praetorian prefect still profiting from the monopoly dye business he created as comes” (420). The passage does indeed show that Petros was still lording it over the silk monopoly (not the dye trade, as Croke has it), which is a strong argument that he was still comes and not prefect. To keep Petros on top of this business, Croke has to imply (though he does not quite state) the unlikely scenario that Petros took the silk monopoly with him when he moved from the sacrae largitiones back to the praetorian prefecture in 555. But this was not how the late Roman bureaucracy normally worked. Moreover, the imperial silk and dyework industry had always been under the jurisdiction of the sacrae largitiones,22 which is why the monopoly, when it was instituted, was in- stituted by its comes. To transfer it to the prefecture would have required a massive administrative overhaul, about which we hear nothing in any source. In fact, at the end of his account of the silk monopoly, Prokopios makes it clear that Petros con- tinued to oversee it as comes sacrarum largitionum: “Through all this, the director of the treasury (ὁ τῶν θησαυρῶν ἄρχων) re- mained the sole dealer in these wares, and while he deigned to give a portion of the proceeds to the emperor, as I said, he kept most of it for himself, enriching himself at public loss. That, then, is how this matter turned out” (25.26). Croke must have overlooked this passage. Petros was still comes, so we are before 555. This episode actually indicates again that it is extremely unlikely that Prokopios wrote the Secret History after 550. The problem is not merely that he records no event that can be 22 A. H. M. Jones, The Later Roman Empire (Oxford 1964) I 427–437. ANTHONY KALDELLIS 597 securely dated after 550 but that he records no such event in cases where we would have strongly expected him to do so.23 Petros Barsymes is one of those officials whose careers Pro- kopios tracks relatively closely in the third section of the Secret History. In ch. 22 he had discussed his tenure of the praetorian prefecture (543–546) at length and had done so in the context of offering a history of that office after its tenure by Ioannes the Kappadokian. He notes at the end of that chapter that Petros was subsequently made comes sacrarum largitionum and turns to his activities in that office in ch. 25 (namely in the passage dis- cussed above). It is, then, hard to understand why Prokopios, had he been writing in 558/9, did not anywhere reveal that Petros had been reappointed praetorian prefect in 555 and was still holding that office at the time when the text was being written. That development would surely have been relevant to the long and complex drama that surrounded Petros’ dismissal from that office in 546 (at Secret History 22.21–38: Look, he’s back in that office now!, or the like). Two of Prokopios’ main themes in dealing with Petros and others were how these officials con- tinued to practice their crimes no matter what office they held and also how they were rarely punished for long by Justinian when they were exposed. A second prefecture for this official in particular would have provided the ideal frame for such an ex- position. Conversely, not only does Prokopios give no hint of Petros’ career after 551, he neglects to mention that two other villains of the Secret History, Addaios and Hephaistos, were also ap- pointed to that office in the early 550s.24 Addaios is attacked for the extortionate practices that he instituted as comes of the harbor of the capital (25.7–10), but no mention is made of the fact that he was appointed to the Kappadokian’s old office in 551. Likewise Hephaistos: he is attacked for his activities as prefect of Egypt in the mid-to-late 540s, but no indication is given that he subsequently became a praetorian prefect in Constantinople, in 551–552 (26.35–39); Prokopios takes his 23 For other cases than the ones I mention, see Haury, Procopiana 19–20. 24 See the praetorian fasti in PLRE III 1473. 598 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY leave of him in Egypt. This silence regarding the future of their careers is peculiar, especially as Prokopios does on occasion state that a certain person in his narrative “later” attained a high office in the capital, i.e., later than the events that he is describing (e.g. Longinos at 28.10, who “later” became urban prefect). Prokopios also does not happen to mention at 27.17 that Pelagius, pope Vigilius’ representative at Constantinople, became pope himself in 556, appointed by Justinian. Certainly Prokopios did not have to say this, but he fails to mention too many important things that happened in the 550s that he could have mentioned. To give another example, the monks who brought silk worms to Justinian in 552–553, discussed at Wars 8.17.1–8 (a text finished in 554), are not mentioned in the long discussion of the silk trade at Secret History 25.13–26.25 To conclude this section, Croke’s arguments fail to establish that the Secret History may possibly date to 558/9, though they make the strongest case for that position in the literature to date. Moreover, they are countered by too much positive evi- dence that the text could not have been written after 550, in fact that it was finished in exactly that year (early 550 to early 551). The composition of the Secret History: two phases Many scholars have noted that the Secret History consists of three main sections: chs. 1–5 narrate the sordid marriage of Belisarios and Antonina and the ways in which Antonina and Theodora conspired to humiliate and emasculate Belisarios; chs. 6–18 discuss the origins, personalities, and criminal regime of Justinian and Theodora, and culminate in the climactic 25 The dating of this story is misunderstood by many historians who discuss Prokopios’ account. He has tripped them up by implying that the monks first came to Justinian from India “at this time” (i.e. 552–553), when they informed him that they could obtain the eggs for him and he asked them to do so. In fact this first visit must have occurred long before, in the late 540s; the monks then traveled to China to obtain the eggs at Justinian’s invitation and returned to Byzantion. It was this second return that hap- pened “at this time” (552–553), for their journey must have lasted at least two years. But this confusion does not affect the dating of the Secret History in any way. ANTHONY KALDELLIS 599 crescendo of ch. 18; while chs. 19–30 are a hostile and detailed commentary on Justinian’s administrative, economic, and legal policies, a rare type of analysis in antiquity.26 My concern here will be with the way in which these parts were composed when Prokopios gave the work its final form, that is, in 550. Other scholars (e.g. Geoffrey Greatrex) have made intriguing pro- posals regarding how Prokopios distributed material from the dossier that eventually became the Secret History to the Wars in what we may call the “prehistory” of the texts’ evolution.27 But looking at the Secret History as we have it, the style, language, and technical vocabulary are consistent across the three sec- tions, and there are many internal cross-references between different sections of the text as well as external cross-references to the Wars and to a work of ecclesiastical history that Pro- kopios was planning to write but evidently never did. These cross-references are correct and redeemed by the passages to which they point (with a few exceptions that I discuss below).28 In short, for a work containing such disparate material the Secret History is remarkably homogeneous and consistent. There is no doubt that it acquired its final form in a fairly short period 26 K. Adshead, “The Secret History of Procopius and Its Genesis,” Byzantion 63 (1993) 5–28, radically split these three sections into separate works: the Milesian Tale, the Aetiology, and the Financial Pamphlet; contra, Kaldellis, Procopius 142–159, 260–261 n.75 (proposing Gynecocracy as a title for the first section). The Penguin translation calls them The Tyranny of Women, Jus- tinian and Theodora, and Anatomy of a Regime: Procopius: The Secret History, transl. G. A. Williamson and P. Sarris (Harmondsworth 2007). 27 G. Greatrex, “The Composition of Procopius’ Persian Wars and John the Cappadocian,” Prudentia 27 (1995) 1–13, and “Procopius the Outsider?” in D. C. Smythe (ed.), Strangers to Themselves: The Byzantine Outsider (Burling- ton 2000) 215–228. 28 For the origins of internal cross-references in ancient historiography see S. Hornblower, Greek Historiography (Oxford 1994) 1–72, here 17 n.30. There are few studies of specific authors, e.g. C. I. R. Rubincam, “Cross-Refer- ences in the Bibliotheke Historike of Diodoros,” Phoenix 43 (1989) 39–61 (where Diodoros is compared to Latin authors). As in the case of Prokopios, the problems of interpretation vary from author to author; see for example T. D. Barnes, Ammianus Marcellinus and the Representation of Historical Reality (Cor- nell 1998), esp. 213–217. Comparative studies focusing on late Roman authors would be illuminating. 600 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY of composition and that it was carefully edited, regardless of the length of time during which Prokopios had previously been ac- cumulating the notes and raw material on which it was based. Yet one aspect of the text has troubled readers (including the present author), namely that the fireworks of ch. 18 read like the conclusion of a literary work as a whole and not of only one section of a work that is about to embark on a detailed and more sober scrutiny of administration and economy. That it is more of a conclusion than a transition to the later chapters is indicated by its lack of any forward reference to the material in chs. 19–30 or any indication that more is to come, by its sum- ming up the numbers of Justinian’s victims inside and outside the empire, by the climactic natural catastrophes of the final page, and by the final sentence, which indicates that Prokopios has said all that he has to say on the topic of Justinian’s crimes: “That was the extent of the destruction of human life that oc- curred first while Justinian was governing the state and later when he held sole imperial authority.” This is reinforced by the end of the preface, where he says “I will proceed to relate first all the wretched deeds that were done by Belisarios and then I will also testify to all the wretched deeds that were done by Jus- tinian and Theodora” (1.10). These two passages, focusing on personalities and the deeds done directly by them, seem to cor- respond to chs. 1–5 and 6–18 respectively, which suggests that when he wrote the preface Prokopios intended to write only what is in those chapters (what I called the first two sections of the Secret History), not the “administrative” third section (chs. 19–30). One could, however, subsume the material in chs. 19– 30 under the “wretched deeds that were done by Justinian and Theodora,” though the two parts are structurally different in that the first focuses on personalities while the second takes aim at officials and specific administrative and legal enactments. The following argument will provide evidence from a different direction, one not based on (subjective) literary responses, that the third section was added on later, in the same year to be sure (given the arguments regarding the date offered above).29 29 Three of the four mentions of the 32 years occur in this last part of the text. ANTHONY KALDELLIS 601 It was still an appendix of sorts to what was already in the mind of the author a finished text that, in its original version, focused mostly on the personalities of Belisarios, Antonina, Theodora, and Justinian. The evidence will emerge from a detailed examination of the system of cross-references in each of the two parts of the text, that is, the part composed earlier, chs. 1–18, and the third, “thematic part,” chs. 19–30. (I have up to now referred to the three “sections” of the text, which are in fact thematic divi- sions, but from now on I will refer to the two “parts” of the text, the first of which encompasses the first two thematic sec- tions and was composed earlier, i.e. chs. 1–18, while the second part is the same as the third thematic section, chs. 19–30.) Now it is impossible to be absolutely precise about what constitutes a cross-reference or to divide all of them unambiguously into internal or external ones. There are many places where Pro- kopios clarifies something he is saying with a reference “to what I just said above,” usually a few pages or a few lines earlier (e.g., 5.27 on the death of Theodora refers back to 5.23). These “local,” internal cross-references are used frequently and consistently across the Secret History, giving the work additional coherence and tying many of its episodes together, either in retrospect or in anticipation. But while they should not be taken for granted, they are not, strictly speaking, relevant to the present discussion. In addition, it is sometimes ambiguous whether a cross-reference is internal or external, for there are some events that are mentioned in both the Wars and the Secret History, for example the plague: is 18.44 an internal or an ex- ternal cross-reference? Such ambiguities do not affect the present argument, which is based on the following observations about a fairly large num- ber of unambiguous cases. In the first part of the Secret History (chs. 1–18), external cross-references almost all refer to the other work being cited as a different logos: “as I said in the earlier logoi” or “as I will say in the later logoi” (see below for the quali- fication almost all), whereas internal cross-references never refer to the passage being cited as being a different logos, referring either to some other part “of this logos” (i.e. of the Secret History), or simply “as I said above” or “below.” In other words, when Prokopios was writing chs. 1–18 he used the term logos as a 602 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY technical term by which to refer (a) to the Secret History itself, which was a discrete logos; (b) to another work that he had al- ready written, the Wars, which he designated in the plural as logoi presumably because it contained many books, though the plural logoi could be used to refer to a specific passage (e.g. Secret History 5.28 referring to Wars 4.21); and (c) to the scandalous Ecclesiastical History he was planning to write, which is also re- ferred to in the plural (logoi) possibly because it too was going to consist of more than one book, like the Wars. Thus, in the first part of the Secret History, the presence or absence in a cross- reference of the terms “the earlier logoi” or “the later logoi” safely indicates whether it is an internal or external reference. The number of these cross-references is large, so we can call this a pattern that reflected the author’s understanding of the interrelation among his works, both written and yet unwritten. There are exceptions, but they do not refute the rule: spe- cifically, in some places where Prokopios has recently cross- referenced what he had said “in the earlier logoi,” subsequent (dependent) references to those logoi may simply be to “what I said above.” We do not need to resolve the question of whether he meant by this “what I said in the earlier logoi” or “as I said above when I referred to the earlier logoi.” This is where in- ternal and external cross-references blur into each other, but it does not happen often and is not, in principle, problematic.30 30 External cross-references in Secret History 1–18 (asterisks mark “depen- dent” cases where the reference to the external logos is indirect, via a recent proximate citation of that logos): 1.11 (Antonina in the Wars), 1.14 (Ec- clesiastical History), 1.14 (Wars 1.25), 1.28 (Wars 6.8), 2.15 (Wars 2.16–19 and 1.25), *2.19 and *2.26 (dependent on the previous reference to the Wars at 2.15), 4.1 (Wars 2.22–23), 5.1 (Wars 7.35.1), *5.17 (dependent on the reference to the Wars at 5.1), 5.28 (Wars 4.21), *5.30 (dependent on the reference to the Wars at 5.28), 6.22 (Wars 2.22–23), 7.1 (Wars 1.24), 11.11 (Wars 7), 11.12 (Wars 2), 11.33 (Ecclesiastical History), 12.6 (Wars 2.21.27), 12.12 (Wars 1.24), 16.1 (Wars 5.2), 17.14 (this is a problem case that will be discussed below), 17.38 (Wars 1.25), 18.28 (Wars 2, and not 1.23.1 as some think), 18.38–39 (a problematic case discussed below), 18.44 (this may be in- ternal or external, as Prokopios mentions the plague many times in the Secret History). Internal cross-references in Secret History 1–18: 2.20 (to 1.3), 4.18 (1.14), 9.48 (6.17), 10.15 (a problematic case discussed below), 10.19 (ch. 7), 11.11 (preface), 12.17 (ch. 18), 13.1 (ch. 8), 13.26–27 (11.12), 13.30 (12.20 ANTHONY KALDELLIS 603 By contrast, in the second part (third section) of the Secret History, which takes up 37% of the entire text, no small pro- portion, there are only three external cross-references that explicitly point readers to different logoi, like the cross-refer- ences contained in the first part of the Secret History.31 The prob- lem, however, is with the internal cross-references made in this second part. Those that refer to material in the first part (chs. 1–18) sometimes refer to that material as being in “the earlier logoi” and sometimes only as “what I said above.” Specifically, 19.8, on how Justinian was illegally confiscating money to the treasury and then squandering it during the reign of his uncle “as was said by me in the earlier logoi,” may possibly refer to 8.4–11 but at any rate to material in the first part of the Secret History. Then, 24.23, on how Petros arranged the murder of Amalasountha “as I mentioned also in the earlier logoi,” must refer to Secret History 16.1–5 and not Wars 5.4, for the latter pas- sage says nothing about Petros as a murderer. Finally, 27.13, on how Theodora pretended to oppose the doctrinal politics of Justinian “as was said by me in the earlier logoi,” refers to Secret History 10.15.32 In other words, when Prokopios came to write chs. 19–30 of the Secret History he was thinking of those chapters as a part and continuation of the same text as chs. 1–18 but also as somehow separate; in his mind, and perhaps only un- ___ and 12.27), 13.33 (8.26), 14.15 (14.1), 15.11 (12.20 and 13.28), 17.1 (chs. 3– 5), 18.1 (12.14), 18.14 (refers to various previous points in the text, possibly to the Wars as well), 18.28 (11.12), 18.35 (11.14–30); in addition, 16.25 im- plies knowledge of 4.7, and 17.16 implies knowledge of 9.19. As part of his effort to postdate the text, Scott, BMGS 11 (1987) 216 n.5, offered that 13.26 “may refer” to Wars 8.25.7 (in fact it is an internal reference to 11.12, as indicated also by the absence of a logoi-reference) and that 29.12 “may refer” to Wars 8.13.19 (a mistake for 8.3.19), which was answered by Greatrex, BMGS 18 (1994) 106. Also, some scholars have taken 13.30 as a forward external cross-reference to Buildings 1.7.7 (for the history of the debate see Croke 430 n.67). But its form precludes it from being an external reference; it is adequately explained by 12.20 and 12.27 (and the Buildings had not been written yet). 31 Secret History 20.16 (to Wars 1.24–25), 21.6 (1.24.18), and 26.18 (pre- sumably to the Ecclesiastical History). 32 Non-logoi (internal) references to the first part of the Secret History: 19.12 (12.12), 29.28 (14.16, 17.32). 604 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY consciously, the already-finished parts of the work (chs. 1–18) had acquired an integral unity as a separate logos. In fact, he now referred back to it in the plural (logoi), as if the first part of the Secret History was analogous to the Wars. The simplest explanation for this shift in his previously pre- cise system of cross-referencing is that when he finished chs. 1– 18 he thought that he was finished with what we now call the Secret History, but that he later took up his pen again, maybe weeks maybe months later, and added the second part. This accords with our literary interpretation of the end of ch. 18 as a conclusion to a work as a whole (as it stood at that time) as well as with the programmatic statement at the end of the preface (1.10) which seems also to be fulfilled by the end of ch. 18. In short, the second part of the Secret History was added later by Prokopios to a work that he thought he had finished. He tried to integrate it as best he could, but thematic rifts remained as well as shifts in his method of cross-referencing. And who among us has retained exactly the same system of referencing in the final phase of writing a book that he used when he be- gan?33 The critique of Justinian’s administration and laws was writ- ten at a later time and had a different focus than the rest of the Secret History, at least in part. This would explain, for instance, why that final part of the work does not at any point explicitly correct, refute, or supplement the narrative of the Wars, as Prokopios promises to do in the preface and actually does often in the first part of the work.34 It may also explain why he makes no reference in the second part of the work to divulging the 33 I would add that the first part of the Secret History contains no forward cross-references to the second part, which might indicate that Prokopios was not planning on writing it (or appending it to this work); but there is only one significant forward reference in the entire work as a whole (12.17 to ch. 18), which provides insufficient grounds for a general conclusion. Two other, similiarly phrased, forward references (19.17, 26.26) point to what is directly about to be said and so also do not reflect much prior planning. 10.15 will be discussed separately below. 34 Especially 1.28 (Wars 6.8), 2.15 (Wars 2.16–19 and 1.25), 2.19 (Wars 2.19), 5.28 (Wars 4.1), 11.12 (Wars 2), 12.6 (Wars 2.21.27), 16.1 (Wars 5.2), 17.38 (Wars 1.25). ANTHONY KALDELLIS 605 “causes” of events, which he promises to do in the preface (1.3) and twice explicitly reminds us in the first part of the work that he is in fact doing (2.20, 11.11). The administrative-legal critique came from a different dossier that Prokopios only later decided to append to the finished chs. 1–18. This does not mean, however, that we should dismember the Secret History and treat it as two works. The two parts were written within a relatively short period and display a high level of linguistic and stylistic unity. Moreover, by the time he finished it, Prokopios clearly intended for the work to be treated as a single com- position and that is a choice that we must respect. The later chapters were certainly meant to strengthen the case made in the first part of the work. The present argument aims merely to identify a compositional seam that perfectly matches the shift of focus in the later chapters. It remains to be seen what further conclusions can be based on this seam, but at present I do not believe that its existence detracts from the unity and purpose of the Secret History. That work was not a “history” in a narrative sense and so it could not be extended, like the Wars, by carry- ing the narrative down to a later date. It offered a critique of a regime and, as such, it could be lengthened by adding material that attacked Justinian from a variety of perspectives and structural approaches. Prokopios was planning yet another, also different, approach, one that focused on ecclesiastical his- tory, but he never wrote it (see below). The hypothesis of a two-part composition can, however, ex- plain at least one oddity in a passage of the second part of the text. In ch. 22, when Prokopios is discussing how Petros Bar- symes was dismissed from the praetorian prefecture, he ex- plains that Theodora was vicious (22.23) and Justinian unstable and susceptible to flattery (22.29–30), as though he had not ex- plained these traits in detail in the first part (e.g. 15.17–18). These character-introductions are written in the same language as those in the first part of the work: Prokopios even uses the exact expression for Justinian’s instability (“he could change his position for no apparent reason and become as light as dust that is carried on the wind”) that he had used above (13.10: “his will was lighter than dust and was blown this way or that by those who would take advantage of him”). A time-lag be- tween the composition of the two parts of the work may ex- 606 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY plain why Prokopios felt that he had to explain personalities that he had explained fully earlier; he may have even forgotten that he had used the same proverbial expression (and so, per- haps, the absence of a cross-reference). Reduplication of this kind occurs only once in the Secret History and may be a unique lapse in the otherwise tight economy of the work.35 Finally, in a separate study I pointed out that Prokopios’ Secret History (and the passages of the Wars that are most like it in outlook) reverberate with Ioannes Lydos’ treatise On the Magistracies of the Roman State, which was completed around 553–554, and concluded that Lydos may have been among the original readers of the secret work. He and Prokopios had iden- tical opinions of recent emperors and high officials and many of their complaints against Justinian’s regime are identical, sometimes even in language and expression.36 We can add to that hypothesis now the observation that almost all the points of convergence between the two works focus on the second part of the Secret History. That is, when Prokopios decided to expand the work with the addition of the legal and administrative ma- terial, he did so with potential readers like Ioannes Lydos in mind. Prokopios’ projected work on church history and the references to it in the Secret History Now that we have a clearer understanding of the system be- hind Prokopios’ cross-references, it would be useful to discuss anew the “unredeemed” cross-references in the Secret History, which have troubled scholars in the past but which have also, in some cases, given rise to shaky assumptions regarding the relationship among his works. Specifically, more can be said about the cross-references that point to an ecclesiastical history of sorts that Prokopios promised but never wrote; about the possible contents of that work; and about whether the Secret 35 For a case of unnecessary repetition in the Wars implying a later ad- dition, see Greatrex, Prudentia 27 (1995) 6. 36 A. Kaldellis, “Identifying Dissident Circles in Sixth-Century Byzan- tium: The Friendship of Prokopios and Ioannes Lydos,” Florilegium 21 (2004) 1–17. ANTHONY KALDELLIS 607 History ever refers to the Buildings. This will certainly be the most speculative part of the present contribution, for we do not actually have the ultimate referents of the cross-references in question, as the work was apparently never written; also, our understanding of the evolution of Prokopios’ corpus is still tentative.37 But a new examination may clear up some old problems. There are three secure forward (external) cross-references in the Secret History to logoi that would have constituted a scan- dalous ecclesiastical history. For the sake of convenience, I will call it an Ecclesiastical History, though its exact title is not attested in Prokopios’ forward references to it. This work would have explained how Antonina engineered the deposition of pope Silverius to please Theodora (1.14); Justinian and Theodora’s policies toward the Christians (11.33); and Justinian’s treatment of priests (26.18). The context of the last reference implies that the future work would expose how Justinian mistreated priests. Two other passages in the Secret History that do not contain cross-references also suggest that priests were as terrified of Theodora as was everyone else (2.26, 10.7), with an implicit criticism of them on Prokopios’ part (namely, that they were cowards). On the other hand, 13.4–7 implies that Justinian al- lowed and facilitated priests’ injustices toward their neighbors, which suggests that they were less the victims and more the accomplices and beneficiaries of his tyranny. Be that as it may, these forward cross-references to the unwritten ecclesiastical history occur in both parts of the Secret History, demonstrating that Prokopios had not given up his intention to write this work after writing the first part of that work (as he must have after- wards: see below). From these cross-references it is legitimate to deduce that the tenor of the work would have been more like that of the Secret History than of the books of the Wars (1–7) completed by that time, for Prokopios refers to it only in the Secret History and not in those books of the Wars. In other words, by 551 only those who had access to the Secret History would have known that an Ecclesiastical History was in the works; it was 37 The most recent reconstruction is by W. Treadgold, The Early Byzantine Historians (New York 2007) 184–192. 608 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY part of the dissident culture. Its approach, moreover, would probably have been unconventional for the genre. After all, this was a historian who could criticize an emperor for persecuting heretics on the grounds that persecution cost many subjects of the empire their jobs (11.19–21), and who could state in so many words that an emperor’s “faith in Christ” was causing people to be murdered or to lose their property (13.4–8). There is one more passage which suggests that Prokopios was planning to write the Ecclesiastical History in the spirit of the Secret History. At 27.32, which exposes the hypocrisy of Jus- tinian’s claim to be the champion of the Christians against their oppressors, Prokopios notes that “I have not said much re- garding the manner in which he chose to defend the rights of Christians, but what I have said, however brief, provides suffi- cient evidence.” This indicates continued self-awareness on the historian’s part that he has not treated this theme fully within the Secret History, and is probably to be taken in conjunction with his repeated promise to write a separate work on that topic. It comes, then, as a bit of a surprise when Prokopios openly announces his intention to write an Ecclesiastical History in Book 8 of the Wars, which he finished around 554. Prokopios is discussing the diversion of a Roman army, sent to aid the Lombards, to the city of Ulpiana in Illyricum “as a civil war had arisen there among the inhabitants on account of those things about which the Christians fight among themselves, as will be written by me in the logoi about this matter” (8.25.13). So as late as 554 Prokopios was still planning to write this work and it still seems to have had the same emphasis on shameful conflict. But there was now a major difference: the work was announced publicly. In the preface to Book 8 of the Wars, Pro- kopios boasted that the first installment had been read in all regions of the Roman empire (8.1.1), and so, presumably, he expected the same for Book 8. So now everyone could know that Prokopios was planning to write an Ecclesiastical History, which would have placed him in a difficult position. He could no longer write it in the spirit of the Secret History, which per- haps meant that he could not write it at all, because most of the material that he had been gathering for so many years was of a critical nature—unless Justinian died first. The emperor’s re- ANTHONY KALDELLIS 609 ligious policies were so unpopular, even with many on the so- called Orthodox side (whichever one so designates), that an attack on them would be welcome (in fact, Prokopios may have hoped to eventually go public with the Secret History). But having committed himself to a public work meant that Jus- tinian had to die first, and there was good reason to think in 554 that his death was not far off; but Justinian did not die until 565 at an unreasonably old age. This may explain why the Ec- clesiastical History was never written.38 To return to the Secret History, there is one ambiguous cross- reference regarding the coverage of religious matters. At 10.15 Prokopios says that “the first thing they [Justinian and Theo- dora] did was to set the Christians against each other and di- vide them all into rival factions by pretending to take opposite paths in the controversies, as I will explain shortly (ὥσπερ μοι λελέξεται οὐ πολλῷ ὕστερον).” As there is no explicit mention here of later logoi, this probably refers to a later chapter of the Secret History itself—probably 27.13, which refers to Theodora opposing her husband in the Monophysite controversy. Yet the Secret History cannot be said to give a satisfactory treatment of this policy of “divide-and-conquer.” It is likely that Prokopios was planning to say much more along these lines in the Ec- clesiastical History. The latter book would probably not have included material on the persecution of still-surviving ancient heretical groups, e.g. Arians and Montanists, who are discussed in the Secret History (11.14–23). It would have focused on the contemporary controversy over Chalcedon. The latter appears only once in the text, precisely in the aforementioned ch. 27, regarding the instructions given by Justinian to the patriarch Paulos of Alexandria to bring the Alexandrians into line with Chalcedon. But the events are not covered there from the standpoint of religious history; Prokopios is still focused on Jus- tinian’s duplicity, hypocrisy, lawlessness, and greed, the themes of the Secret History. 38 Prokopios seems to have waited for Theodora to die (in 548) before writing the Secret History: Treadgold, The Early Byzantine Historians 187–188. He may also have been waiting for Justinian to die to integrate the material of the Secret History into the Wars: Greatrex, in Strangers to Themselves 215–228. 610 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY Next, there are unredeemed cross-references whose referent is unclear. At 17.14 Prokopios promises to tell in other logoi about the crimes committed in office by two vulgar men whom Theodora married to two well-born widows. He does not name them and we never hear about them again. The manuscripts read as follows: “these things will be told by me in earlier logoi,” which makes temporal nonsense (given that Prokopios always refers to his earlier logoi, i.e. the Wars, in the past tense). Braun changed “earlier” to “later” (ἔμπροσθεν to ὄπισθεν). If the emendation is allowed to stand, this would probably be another reference to the Ecclesiastical History, the only “later logoi” to which Prokopios refers in the Secret History (at least in cross- references that are not textually corrupt; see below for one more). These two men may, then, have played a vicious role in Church politics. It is unlikely that the solution is to correct the tense of the verb and make this into a back reference to the Wars, for it would be impossible to identify these men in the rich prosopography of that narrative. Prokopios is precise and clear in all of his other back references to the Wars, so this must be a forward reference. We come, finally, to the most problematic external cross- reference in the Secret History, the one that has stirred the most controversy. At 18.38 Prokopios refers to the flooding of the Skirtos river in 525 and the damage that it caused to the city of Edessa, ὥς μοι ἐν τοῖς ἔμπροσθεν λόγοις γεγράψεται (“as will have been written by me in the earlier logoi”). As with the pas- sage discussed above, this makes little temporal sense. Haury first suggested that Prokopios may have discussed the flood at Wars 2.12.29, where there is a lacuna of about nine lines in a discussion of Edessa (γεγράψεται could, then, be changed to γέγραπται). Scholars have viewed this suggestion skeptically, and even Haury himself soon changed his mind.39 In the Wars passage, Prokopios is discussing the authenticity of a promise made in an apocryphal letter sent by Jesus to Abgar, the king of Edessa, and the lost text in the lacuna seems to have concerned a brief period of Parthian rule over the city, not an appropriate 39 See the discussions of the textual history of this passage by Evans, GRBS 37 (1996) 310–311; Croke 429–430. ANTHONY KALDELLIS 611 context to mention a sixth-century flood.40 As it happens, Pro- kopios offers a long account of this flood at Buildings 2.7.2–16, where he praises Justinian for his preventative and restorative works in Edessa. Believing correctly that the Secret History was written before the Buildings, Haury simply changed ἔμπροσθεν to ὄπισθεν. Later scholars, lacking an alternative, have ac- cordingly assumed that this was in fact a forward reference to the Buildings.41 Whatever solution we adopt to the textual problem, it is un- likely that the Secret History contains a reference to the Buildings at this point, or, indeed, at any other point. The Buildings would not be written for another four years at least. It was, moreover, a work whose ideology and purpose (to praise Justinian) was so at odds with the Secret History that it would make little or no sense for Prokopios to casually cite it here as though its outlook naturally complemented or meshed with the Secret History. Let us not forget that the whole purpose of mentioning the flood at this point was to attack Justinian as an evil power, whereas in the Buildings passage it would be to praise his solicitude for his subjects’ welfare. All the external cross-references in the Secret History are, moreover, carefully written to ensure ideological consistency, as we have repeatedly seen. Prokopios refers to the Wars to buttress or give a context to what he is saying in the Secret History, or to explicitly correct what he had misreported there because of fear. His forward references to the Ecclesiastical History are likewise always in complete conformity with the outlook of the Secret History. In short, Prokopios uses his cross- references for the same reason that any scholar would: to support what he is saying or to defer a discussion that would 40 See below for a closer discussion of this passage. 41 E.g. A. Palmer, “Procopius and Edessa,” AntTard 8 (2000) 127–136, here 128 (see below for Averil Cameron). Palmer usefully summarizes the sources for the 525 flood and debates the evidence for reconstruction by Justin versus Justinian (see also below for another aspect of his argument). Croke (429–430), wanting the Buildings to date before the Secret History, ac- cepted the reference but objected to the emendation, though his chrono- logical grounds have been answered above: a date of 558/9 will not work for the Secret History. This does not mean, of course, that the emendation must stand. 612 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY likewise support what he is saying, not to undermine what he is saying. It would literally refute the Secret History to assume that he was, in ch. 18 no less, referring to a panegyrical work such as the Buildings would turn out to be. There is, moreover, so much material in the Buildings that is topically relevant to so many chapters of the Secret History that we must wonder why he would refer to it only in this one instance, and so casually at that. Setting aside Hagia Sophia, which is mentioned three times in the Secret History (both before and after the Nika Riots: 3.24, 9.35, 17.9), why did Prokopios not refer forward to his account in the Buildings of the flooding of the Kydnos river at Tarsos (5.5.14–18), an event that he mentions in the Secret History only two sentences after the flooding of the Skirtos (18.40)? Or how are we to understand the constant attacks in the Secret History on Justinian’s “stupid buildings” and on his failure to build anything useful for the cities (8.7–8, 11.4, 19.15, 26.23, 26.33), attacks that seem to be almost designed to pre- emptively undermine the Buildings? (In fact, they only aimed to undermine Justinian’s image as a builder, as Prokopios prob- ably had no plans yet to write anything like the Buildings.)42 The corruption of the text at this point is probably an in- superable obstacle; we will never be able to resolve the prob- lem. But we have seen that in the entire Secret History, Prokopios refers back to only one set of logoi, namely the Wars, and refers forward to only one other, the Ecclesiastical History. Let us con- sider these two possibilities. If we grant Haury’s emendation (to 42 This one forward reference was cautiously cited by Averil Cameron to support her view that the real Prokopios was to be found less in the Secret History and more in the Buildings: Procopius and the Sixth Century (London 1985) 106–107: “we can see how closely the three works are linked, and how the Buildings slots into Procopius’ writing as a whole.” But her reading of the Buildings as a sincere work expressing Prokopios’ ideology (cf. 11: his “ac- ceptance of the basic assumptions of the Justinianic regime”) and her prob- lematic dismissal of the Secret History (which makes him look too dissident and skeptical) as “personal rather than political” (55) have been answered in Kaldellis, Procopius 45–61 (it is in fact in the Buildings that Prokopios deploys classical allusions to undermine the panegyrical message of that text). Pro- kopios attacked the regime on fully political (and therefore ideological) grounds. ANTHONY KALDELLIS 613 “the later logoi”), it would seem that Prokopios planned to discuss the 525 Skirtos flood in connection with ecclesiastical affairs. Edessa was a major city and could well have been the site of controversies and events that he planned to discuss. The leading Monophysite missionary of the period and a client of the empress Theodora, Jacob Baradaios (after whom the Jac- obite Church is named), was the bishop of Edessa after 542/3. But we do not need to speculate about events that could have taken place in the same city at a different time on the assump- tion that Prokopios might have been moved, in the course of narrating those events, to digress on the 525 flood. Syriac sources tell us that Monophysites were being persecuted in Edessa by the bishop on the very day that the flood occurred, making it virtually certain that Prokopios intended to discuss the 525 flood in the Ecclesiastical History. “While Asklepios was bishop there and using violence to constrain the faithful to agree to the iniquitous synod of Chalcedon; and had arrested twenty solitary monks and was torturing them pitilessly to make them agree to the accursed synod … that very same night there occurred a flood…”43 Our notices continue by narrating the course of the flood in connection with the ongoing religious persecution. This is exactly the material that Prokopios in- tended to place in the Ecclesiastical History—“those things about which the Christians fight among themselves,” as he would put it at Wars 8.25.13. On the evidence of the Secret History, then, it is much more likely that Prokopios had that work in mind than the Buildings when he referred to a future discussion of the 525 flood. We should not fall into the trap of thinking exclusively within the set of surviving works. If, on the other hand, Haury was wrong and this is a back reference, we may have to reconsider his original proposal 43 This is translated from the chronicle of Michael the Syrian, Jacobite patriarch of Antioch in the late twelfth century, who used sixth-century sources, in Palmer, AntTard 8 (2000) 131. See also the account, ultimately based on the chronicle of Yuhannan of Amida (i.e. John of Ephesos), in W. Witakowski, Pseudo-Dionysius of Tel-Mahre, Chronicle, Part III (Liverpool 1996) 41–44. For the historical context see V. L. Menze, Justinian and the Making of the Syrian Orthodox Church (Oxford 2008) 111–120. 614 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY about the lacuna at Wars 2.12.29. Even though a forward refer- ence to the Ecclesiastical History is the most plausible solution to this problem, let us revisit the Wars passage because it is, in fact, not as unlikely a context for a discussion of the 525 flood as scholars have assumed, especially in light of what other sources report about the event. At Wars 2.12, Prokopios is ex- pressing skepticism regarding the Edessenes’ belief that Jesus had promised their king Abgar that the city would never be captured by enemies. He points out that the city did in fact fall under “Persian” power not long after, only not by being cap- tured. One of Abgar’s sons apparently turned it over to them but his people expelled the garrison and gave their city to the Romans. Then there is a lacuna, a suspicious one given the context, and when the text resumes we find Prokopios develop- ing the tortured (and probably tongue-in-cheek) position that, while the promise of inviolability was inauthentic, God had de- cided to uphold it anyway so as not to give people grounds for skepticism.44 Why might Prokopios have mentioned the 525 flood in this context? Two reasons can be offered. First, Andrew Palmer has suggested on the basis of a contemporary Syriac source, the Chronicle of Edessa, that some Edessenes were appealing to Justinian for aid in coping with the floods: insisting on the damage caused by them “is a round- about way of saying that this promise of immunity, though effective only against the Persians, is of no avail against the floodwaters, the consequences of which are almost as deva- stating as capture by the enemy … For those [such as Pro- kopios] who entertained doubts [about the immunity], the fact that Edessa was vulnerable to flooding meant that it was also vulnerable to siege, especially after a major flood.”45 We do not need to discuss here Palmer’s further proposal that Prokopios’ 44 See a discussion in A. Kaldellis, “Prokopios’ Persian War: A Thematic and Literary Analysis,” in R. Macrides (ed.), Byzantine History as Literature (forthcoming). Euagrios (HE 4.27) condensed Prokopios’ account in Wars 2.12 but gave no hint as to what the lacuna might have contained; cf. M. Whitby, The Ecclesiastical History of Evagrius Scholasticus (Liverpool 2000) 225 and the commentary in n.70. 45 Palmer, AntTard 8 (2000) 133. ANTHONY KALDELLIS 615 narrative of Edessene history in Wars 2.12 was based on the text of an Edessene plea to Justinian for help in coping with the floods. It suffices to know that, in the 540s, the question of immunity could be linked to the damage caused by the 525 flood. “The bitter irony in saying that the Blessed City [Edessa] has four times been destroyed and her children drowned by floods since the Ascension comes right at the end of the chron- icle and seems designed to lead up to the question why.”46 Prokopios’ argument in Wars 2.12 is all about irony and asking why. In other words, it is not fanciful to think that the lacuna might have contained a brief mention of the 525 flood. Second, the contemporary chronicler Malalas (17.15) reports that the people of Edessa were surprised at the destructiveness of the 525 flood, having, apparently, grown complacent at the smaller scale of the event in the past. He adds that after the flood a tablet was found which predicted a terrible flooding of the Skirtos. This too fits nicely with the context of Prokopios’ skeptical discussion in Wars 2.12 of the prophesy regarding Edessa’s immunity from capture. In light of Malalas’ testimony, then, it is easy to see how he might have digressed from his discussion of the apocryphal prophesy of the city’s immunity to this curious prediction regarding the flooding of the river. This solution is less likely, in my view, than the alternative proposed above, but the context still fits. In short, we have three sce- narios that make more sense than an otherwise incompre- hensible reference to the Buildings at Secret History 18.38.47 46 Palmer, AntTard 8 (2000) 135. 47 One minor problem remains. At Secret History 18.39, i.e. immediately after his mention of the Skirtos flood, Prokopios refers to his “earlier” account of the Nile flood that would not subside (Wars 7.29.6–8): ἅπερ μοι καὶ πρότερον δεδιήγηται. While this is an external cross-reference, it lacks the logoi-qualifier because it comes soon after another external cross-refer- ence (namely to the account of the Skirtos flood: see n.30 above). In all other cases where external cross-references follow each other, the reference is to the same text. Does this support the ἔμπροσθεν reading for 18.38? It might depend on how we take the emphasis in καὶ πρότερον. I am un- decided. 616 PROKOPIOS’ SECRET HISTORY AND CHURCH HISTORY Conclusion By attending closely to the language that Prokopios uses when he refers to the 32 years of Justinian’s power in the Secret History, and by bringing to the discussion the evidence of the emperor’s edicts and the careers of his officials, it is possible to lay to rest the question of the text’s date: that date is 550/1. Second, the odd thematic structure of the Secret History, in combination with a close examination of the system of internal cross-references employed throughout the text, indicates that chs. 19–30 were written somewhat later (albeit in the same year) and do not fully share in the text’s original narrative objectives. Lastly, the article considered the tenor and scope of his planned but not written Ecclesiastical History and attempted to identify the referents of unredeemed cross-references in the Secret History, arguing in particular that 18.38 does not refer to the Buildings.48 May, 2009 Dept. of Greek and Latin The Ohio State University Columbus, OH 43210-1319 kaldellis.1@osu.edu 48 The author is grateful to Brian Croke for providing thoughtful criti- cism, corrections, and advice that substantially improved this argument. His impartiality, open-mindedness, and generosity set a high standard for us all.