I URBAN RIOTING AND IMPERIAL RESPONSE IN 341–2

The winter of 341 to 342 witnessed grave unrest at Constantinople. Eusebius of Nicomedia, a sympathizer of Arius, had been installed in 337 as bishop of Constantinople to replace the deposed Paul. When Eusebius died in 341, his supporters fielded Macedonius as successor; but Paul, newly returned from Pontic exile, was put forward again by those who prized him as a true adherent of the Nicene Creed. Rioting broke out towards the end of 341. Hermogenes, one of Constantius II's generals, was instructed to deal with the problem. But in early 342, probably January, after apparently trying to use soldiers to eject Paul from church, Hermogenes was lynched by an angry mob. The situation was now a crisis. Constantius, who was occupied at the time on the eastern frontier, made a rapid trip in winter weather across the Anatolian mountains to Constantinople to settle the problem in person, before making almost as rapid a trip back to Antioch. Remarkably, however, he reportedly spared the death penalty. Instead, among other measures, he expelled Paul from Constantinople and halved the city's grain dole. His round-trip brought him back to Antioch by April.Footnote ¹

The episode is narrated by Socrates and Sozomen, while Hermogenes' part is also recorded in the Historia acephala, Jerome's Chronicle, and the Consularia Constantinopolitana. Jerome dates Hermogenes' death to Constantius' fifth year, namely 341–2, while Socrates and the Consularia Constantinopolitana give a consular date of 342.Footnote ² In addition to these rather later sources, Libanius' Oration 59 — which dates from the 340s — refers at a significant juncture to the episode. In the form of a series of rhetorical questions, Libanius goes on to attest Constantius' journey from the eastern front, in winter and in haste, to deal with the crisis. He also records Constantius' decision to spare the death penalty.Footnote ³ In the process, however, he indicates that the episode caused controversy between the senate of Constantinople and the emperor:Footnote ⁴

Significantly, the structure of Libanius' rhetorical questions suggests that the subject of controversy was the decision to exercise clemency toward the rioters by replacing the death penalty with other measures: Libanius immediately juxtaposes Constantius' clemency with his appearance in the senate to give a speech and exchange in dialogue, with no indication of any other source of controversy between them. That the emperor's clemency toward the rioters was indeed a source of contention need hardly be surprising, particularly in view of the halving of the city's daily bread dole from eighty thousand to forty thousand modii. It is worth keeping in mind that the bread dole was not necessarily a welfare system for the urban poor. In Rome and Italy, the qualification for receipt of free bread had long been the possession of a token (tessera) and such tesserae were frequently issued, not to the urban poor, but to well-to-do citizens. To a significant degree, it was, in short, a form of civic privilege.Footnote ⁵ If the same was true for Constantinople, it follows that Constantius was substantially cutting the scope for tokens to be distributed — hence cutting back a system of patronage by which more powerful figures could channel the distribution of tokens to their preferred clients.Footnote ⁶

Whose idea this clemency toward the rioters was must remain an open question. It was in the nature of the genre of panegyric that the credit should be attributed to Constantius directly. In reality, he might have done no more than respond to the lobbying of interested parties. But as the emperor's urgent winter journey makes plain, the episode was a major early case of the advance, under Christian rule, of Christian differences onto the front stage of political life in the eastern empire. The violence intruded on the rhythms of urban life and might well have provided the first encounter by the eastern senate, as a corporate body, with Christian ecclesiastical problems. It is unsurprising that it should have proven a testing incident for the senate and Constantius alike.

The present paper will argue that Themistius' Oration 1 also alludes directly to this episode and to the controversy it generated about the exercise of imperial clemency (II); and furthermore that, although the speech is usually dated to either 347–8 or 350–1, there are good grounds to date it to the immediate aftermath of the rioting, in 342 itself (III–IV). In conclusion, it will explore the implications of this reading and chronology both for the trajectory of Themistius' early career and for the way in which it intersected with imperial ideology and policy (V–VI).Footnote ⁷

II ORATION 1 BY THEMISTIUS

Let us now juxtapose with the crisis of 341–2 a reading of Oration 1 by Themistius, taking account both of its overall shape and some points of detail. As John Vanderspoel has noted, none of Themistius' surviving panegyrics follow the structural precepts associated with Menander Rhetor, which were usually adhered to by others in the fourth century. Although some of the relevant material might be used, the traditional sections on personal background, education and accomplishments are dropped in favour of a philosophical structure, in which ideal qualities and concrete illustrations are integrated.Footnote ⁸

It is this approach that Themistius takes in his Oration 1. The philosopher wastes no time in making plain that his subject is the emperor's own soul (Or. 1.2b–c). The theme of the emperor's love of mankind is soon introduced and defined in opposition to anger and to other manifestations of a lack of self-restraint, which are cast as the hallmarks of a tyrant (ibid., 4a–7d). The importance of love of mankind as a distinctively kingly and divine virtue, and the peculiar destructiveness of the anger of a king, are emphasized (ibid., 6a–7c). The good king therefore tries to imitate God through love of mankind, and God promotes or disposes of kings accordingly (ibid., 8d–9c). The king who loves mankind finds that he rules over friends; to rule by fear is to rule people who cower, and only a tyrant, not a king, is happy to be better off than the wretched. Hence Shapur is not a king at all (ibid., 10c–11c). Wickedness hates virtue, and what defeats Shapur is not the battlefield but the virtue of his neighbour Constantius (ibid., 12a–c).

Vanderspoel observes that the ‘shift from abstract to concrete is subtle’;Footnote ⁹ and indeed, Constantius and Shapur are slotted quite elegantly into an otherwise highly generic set of philosophical notions about kingship.Footnote ¹⁰ But the element of specificity that is most striking about Oration 1 is surely not the contrast of Constantius and Shapur, which is still a rather obvious direction for the speech to go, especially against the background of the emperor's eastern campaigning through the 340s.

It is, rather, somewhat further on in his text that the sequence of deeply traditional philosophic propositions about kingship becomes distinctive through his choice of specific illustration. Themistius tells his audience that to love mankind is also to hold mankind in great reverence. Kings give rewards but the task of punishment is passed to the executioner. The dignity of a king owes more to reward than to punishment — the former adding something good for mankind while the latter merely removes something bad (Or. 1.12c–13d). It is better, therefore, for a king to lean away from punishment and vengeance, which do not improve the wrongdoer but simply kill him for the benefit of others (ibid., 13c–14b).

Turning again from the generic to the specific, Themistius observes that it is surely for this reason that Constantius removed the death penalty (ibid., 14b–c, quoting 14b):Footnote ¹¹

For ancient law, Themistius argues, aimed at severity and an undue uniformity of punishment by death, whereas the king who loves mankind sees the inexactitude of written law and sets aside its harshness (ibid., 14c–15c).

Themistius continues with another conjunction of abstract and specific material. Justice, he says, should distinguish between calculated wrongdoing, error born of emotional turbulence, and sheer misfortune born of some alien fault (ibid., 15c–16a). Themistius illustrates his point by saying that he alludes to actual events for an example. What he picks for his argument is the observation that the killing of a man is something that can arise in different circumstances (ibid., 15d–16a):Footnote ¹²

Themistius goes on fleetingly to mention the story of Adrastus killing the son of his host, the Lydian king, while out hunting. What is striking is not the mythic illustration itself but rather the choices it reflects. Themistius is opting to focus his discussion of clemency on a scenario in which a man dies in violent but accidental circumstances. Love of mankind, Themistius says, strives for fairness and does not punish random happenings (ibid., 15d–16a).

He draws to a close by observing that Constantius has shown precisely such gentleness and proven, as a result, that love of mankind causes wickedness to shrivel rather than flourish — achievements especially remarkable in a young king (ibid., 16b–c). His parting remarks are that the lover of mankind is also a lover of his friends because he knows how precious and necessary friends are; and unlike the friends of a tyrant, the friends of a king know they are safe in his company (ibid., 17b–18a).

Viewed as a whole, the persistent subsidiary theme of Oration 1 is the particular manifestation of the kingly love of mankind in the form of his restraint of his own anger. On Themistius' account, this restraint is basic to the distinction of king from tyrant; and it is, by implication, a central part of what makes his rule an image of divine rule. This much is true to well-established Hellenistic notions of kingship. But as the speech progresses, there is a notable progression in specificity, with the theme of the restraint of anger taking more concentrated form in the shift of focus to the moderation of punishment. It is in this context of clemency that Themistius makes his two most striking assertions of fact: that Constantius had removed the death penalty; and that the killing of a man illustrates from real life the differences between wrongdoing, error, and misfortune.

These assertions present us with an interpretative puzzle. As regards a cancellation of the death penalty, Constantius II plainly did not do any such thing in a generalized sense. He ruled — as Ramsay MacMullen charted, in a sad and celebrated essay — in an age in which an ever wider range of offences came to meet with capital punishment.Footnote ¹³ One need only scan the pages of the Theodosian Code to be sure that Constantius was no exception. Faced with this potential gulf between rhetoric and reality, our best way to understand Themistius is to regard Constantius' removal of the death penalty as a decision specific to a particular context. The puzzle is to identify that context. It was something which loomed large enough that Themistius had no need to remind his audience of what it was.

Against this background, one cannot help but note that the combination of details — the killing of a man and the sparing of the death penalty — provide an excellent match to the escalation and resolution of the crisis of early 342. Placed in this connection, the Adrastus story perhaps becomes more than a small rhetorical flourish: Themistius is perhaps trying, very allusively, to imply that the death of Hermogenes, in the thick of mob violence, should in fact be regarded as a tragic accident rather than a wilful murder — an interpretation of events that would, of course, help to justify the clemency shown toward the rioters. That there was a robust difference of opinion on the matter between the eastern senate and Constantius — such as that alluded to by Libanius, who had been present in Constantinople at the time — need hardly surprise us; and it provides a perfect backdrop for Themistius' speech.Footnote ¹⁴ On the internal evidence, it is certainly possible to regard the speech as having been consciously tailored to a vindication of Constantius' clemency in the circumstances of 342. By itself, this does not rule out other possible contexts for the speech. But the identification of the most plausible context is a problem that brings us squarely to the question of how we date the speech.

III PROBLEMS IN THE DATING OF THEMISTIUS, ORATION 1

The point of departure for all attempts to date the speech is the fact that the manuscript tradition prefaces the text with the following remark:Footnote ¹⁵

The Theodosian Code explicitly places Constantius II in Ancyra on 8 March 347, prompting Jean Hardouin to date the speech to that time in his annotations to the edition of 1684.Footnote ¹⁶

This remained the standard view until Otto Seeck argued for 350 in 1906. Seeck observed that Ancyra lay on the main route between Constantinople and Antioch so that, on any journey between the two, Constantius can be expected to have passed through it. Reliance on a single explicitly attested stop at Ancyra is therefore unnecessary and — given that Contantius made several known trips between Antioch and Constantinople — is potentially misleading.Footnote ¹⁷ The observation is a good one and Seeck was surely right that we are not compelled to date the speech to 347.

Conversely, Seeck's positive case for 350 rests on the rather more fragile notion that, had Constans still been alive, the speech would have needed to mention him as a courtesy, making Constans' death the terminus post quem. In light of Constantius' long western sojourn in the 350s, Seeck took the view that 350 itself was to be preferred. Yet although a living co-emperor would normally be acknowledged, there were certainly circumstances in which the etiquette was put aside. Vanderspoel notes very aptly, for example, that Themistius was also silent on Valentinian II when the boy-emperor was definitely still alive but despised by Theodosius; and Peter Heather and David Moncur have noted a strange silence on Julian in 357.Footnote ¹⁸ Themistius' silence on Constans need only reflect a phase of heightened tension — of which there was more than one — in relations between the eastern and western governments.

The corollary of removing Seeck's terminus is that other arguments, by Heinrich Scholze and Werner Portmann for 350 or 351 respectively, are severely weakened, resting as they do on the a priori acceptance of Seeck's terminus for the interpretation of highly ambiguous passages.Footnote ¹⁹

Conversely, Karl Otto Gladis treated the speech independently of Seeck as a celebration of the Battle of Singara, for which he accepted a date of 348. Gladis took the view that Themistius' reference to victory over Shapur referred specifically to the battle and hence that the speech fêted Constantius on his return from the front that year.Footnote ²⁰ But Gladis's argument is unconvincing because the mention of victory over Shapur is patently not the main focus of Themistius' speech.Footnote ²¹

Despite useful critiques of other treatments (including Seeck's terminus) by Heather and Moncur, their own position on the chronology remains uncertain, acknowledging both 350 and 347 as possibilities.Footnote ²² More interesting is the case presented by Vanderspoel, who similarly rejects Seeck's terminus. Vanderspoel avoids misdiagnosing the speech as a victory celebration when he argues that the reference to victory over Shapur can be identified, on closer inspection, with a military operation in 346. On the basis that Singara, for which he also accepts a date of 348, was too significant for Themistius to have ignored if it had happened recently, Vanderspoel takes the view that the speech was delivered at some time between the two victories, with Hardouin's date of March 347 still most appealing.Footnote ²³

Vanderspoel's case is perhaps the most satisfactory to date. But it demands closer consideration. The identification of the military operation turns chiefly on Themistius' remark: Τοῦτ’ οὖν ἐστιν ὃ διόλλυσιν ἐκεῖνον, οὐχ ἡ μέση τῶν ποταμῶν, ἀλλ’ ἡ βασιλέως ἀρετὴ πλησίον λάμπουσα.Footnote ²⁴ Vanderspoel translates this as: ‘This, then, is what destroyed him [Shapur]: not Mesopotamia, but the virtue of the emperor shining nearby.’ Vanderspoel takes the remark to signal that Shapur had been defeated recently in Mesopotamia, in an engagement in which Constantius had been nearby but not personally present.Footnote ²⁵ With reference to nine major engagements with the Persians listed by Festus for Constantius' reign, Vanderspoel rules out four on the grounds that Constantius was either too personally involved or too far away to be described as ‘nearby’ and then settles on the siege of Nisibis in 346, finding support in Ephraim and Athanasius for Constantius having been ‘nearby’ on that occasion.Footnote ²⁶

Though they do not critique Vanderspoel directly on the point, Heather and Moncur offer a slightly different reading which has implications for chronology. As they translate: ‘This then is what brings about the ruin of this man: not Mesopotamia, but the virtue of the king shining out next to him.’Footnote ²⁷ On their reading, the proximity (πλησίον) in question is not literal in relation to a specific military encounter, with Constantius ‘nearby’, but metaphorical — with Constantius ‘next to’ Shapur because the Roman and Persian empires are themselves side by side.Footnote ²⁸

It is a strength of Heather and Moncur's reading that it keeps the destruction of Shapur (Τοῦτ’ οὖν ἐστιν ὃ διόλλυσιν ἐκεῖνον) in the present tense. Vanderspoel's rendering of the phrase in the past tense tends to reinforce the impression that Themistius is alluding to a single, specific defeat. By contrast, a literal translation in this instance has the merit of preserving Themistius' open-ended language. This, in turn, tends to underline the clearly abstract nature of the point that he goes on to develop. According to Themistius, what Shapur fails to understand is that the sole merit of his proximity (γειτονήσεως) to Constantius is that he, Shapur, a poor helmsman for an empire, has the possibility of tying his manoeuvrable but feeble ship of state to the great Roman battleship, rather than pursuing a struggle in which he is ultimately overmatched.Footnote ²⁹

Themistius' premise, therefore, is not that Shapur has had a particular defeat but rather that his defeat is an ongoing inevitability. Against this background, Themistius' point is that that unfolding defeat, however evasive of it Shapur may seem, is not a result of the Mesopotamian battleground but of the virtues of Constantius. In this light, too, the quality of Persian manoeuvrability that Themistius mentions hardly applies to the tactical behaviour of the Persians in a particular battle. Rather, Themistius' point is that it is a quality of Shapur's strategy and statecraft in temporarily eluding his own ineluctable destruction.Footnote ³⁰

Hence, to try to identify an allusion in the speech to a specific military operation — whether the siege of Nisibis in 346 or some other engagement — presses the evidence too far. To say this does not exclude the possibility that the speech followed the siege of 346. But it does serve to emphasize that that engagement has not been established as a terminus post quem any more than has Singara or the death of Constans.

IV AN ALTERNATIVE DATE FOR THEMISTIUS, ORATION 1

With no reliable terminus from which to work, the whole question of the date of the speech is wide open and invites a fresh perspective. Beginning with the fact that Constans' death in 350 is not (pace Seeck) a necessary point of departure, and taking on board existing critiques of the arguments of Scholze and Portmann for 350 and 351, we can supplement these doubts about a relatively late dating by pointing out that, so far as we know, Themistius was born in about 317.Footnote ³¹ Aged about thirty-three to thirty-four in 350–1, Themistius was neither young by ancient definition nor likely still to be producing work that left an impression of immaturity.Footnote ³² A date of 347, as proposed by Hardouin and Vanderspoel, is probably the latest option that we should willingly contemplate for a work associated with youth. Themistius was then about thirty. Libanius, who regarded himself as something of a late starter, had a teaching post in Athens by his mid-twenties.Footnote ³³ Augustine had a teaching post in Carthage by the age of just twenty-two. At thirty, he had taken up the imperial chair of rhetoric in Milan.Footnote ³⁴ Nor was Themistius himself likely to be found in the ‘slow lane’. He came from an already prestigious family: his father Eugenius was a prominent philosopher in Constantinople and his grandfather, also a philosopher, had had a similar public profile under Diocletian.Footnote ³⁵

Given the lack of precise knowledge about Themistius' early career, appropriate opportunities for him to have delivered the speech must first be identified by reference to Constantius' itinerary.Footnote ³⁶ The first opportunity after the rioting is precisely on the return journey by Constantius from Constantinople to the eastern front in 342, after the city was becalmed. This journey is beyond doubt — attested not only by Libanius' remark that Constantius left Constantinople to face the enemy again without even resting, but also by attestations, in the Theodosian Code, of Constantius' presence back in Antioch in spring 342.Footnote ³⁷ The second opportunity is conjectural. Richard Klein, tentatively followed by Timothy Barnes, has suggested that Constantius might have visited Constantinople in autumn 343 to celebrate his vicennalia.Footnote ³⁸ If so, Themistius could have delivered the speech either when Constantius was travelling towards Constantinople or when he was returning to Antioch, where he is attested again in spring 344.Footnote ³⁹ No other opportunity is known before Constantius' appearance in Ancyra in 347.

Of these possibilities, the earliest is the most convincing on present evidence. 342 itself is the only time when we know that Constantius travelled between Antioch and Constantinople, when we can also identify a major occurrence in public life of a kind that forms a discernible basis for the principal concerns of Themistius' speech, and when Themistius (then scarcely twenty-five) can be regarded really convincingly as a youthful speaker. No other date fits all three criteria. Not only this, however, but the speech — coming, as it then would have done, within probably less than a month of Constantius' debate with the eastern senate — would have had the most immediate propaganda relevance. By locating Themistius' speech in the immediate context of political differences over clemency, one gives back to it a freshness and significance that it simply would not have possessed five years later in 347, when there is no comparable background of a killing and clemency; and which it would have possessed to a much lesser extent in late 343 or early 344, when the crisis was more than a year and a half or two years behind. Delivered in 342, probably March, the speech becomes much more than an abstract exercise in praising an emperor. It becomes a practical element in the promotion of Constantius' decision-making at a time when he faced criticism from eastern magnates.

There is also good reason why a panegyric on Constantius delivered in early 342 might not have mentioned Constans. Expelled from Constantinople, Paul appears immediately to have headed west for the territory of Constans, whose court in Trier he reached later that year.Footnote ⁴⁰ It is not too much to conclude that Paul's westward direction of travel was known to Constantius' agents and that this was reflected in the briefing from court officials that Themistius is likely to have received.Footnote ⁴¹

The significance of this hypothesis emerges against the background of the deep rift that had opened by summer 341 between Julius, Bishop of Rome, and many of the eastern bishops.Footnote ⁴² Julius had written to the eastern bishops in 340 to complain that Athanasius and Marcellus had been unjustly deposed from the sees of Alexandria and Ancyra respectively; and he had alleged that the eastern bishops were destabilizing the Church by failing to respect the Council of Nicaea. He had proposed, therefore, a council of western and eastern bishops. In response, a council of eastern bishops — meeting at Antioch in January 341 to dedicate the great church, which had been started under Constantine and was now ready — rejected the proposal of an East-West council, accused Julius of trying to over-reach his authority, and declared that, if he remained in communion with Athanasius and Marcellus, he would himself be excommunicated. Julius then replied with a full rebuttal in summer 341, accusing the relevant eastern bishops of being schismatic and of persecuting other eastern bishops besides Athanasius and Marcellus, many of whom had also fled to Rome from Thrace, Syria, Phoenice and Palestine.

The situation was made graver by the fact that Constantius had attended part of the Council of Antioch, at the very least to dedicate the new church.Footnote ⁴³ Whether he participated directly in the rejection of Julius' complaints is unknown but it is difficult not to infer at least his tacit support for the position taken by the eastern bishops — and hence, also, that the bishop of Rome was refusing to bow to tacit pressure from the eastern emperor.

The potential for Constans to become involved was obvious. Western intervention in eastern ecclesiastical matters had occurred before. After Constantine had exiled him to Gaul, Athanasius had attended the court of Constantine Caesar (latterly Constantine II) at Trier, who wrote to the Christians of Alexandria in June 337 in Athanasius' support, securing his return to the Alexandrian see by November. After a council of bishops, at which Constantius was present, subsequently met at Antioch in the winter of 338–9 and formally deposed Athanasius, causing him to flee Alexandria to avoid arrest, the deposed bishop had crossed into the territory of Constans and gone to Rome. As Barnes has argued persuasively, Athanasius then wrote from Rome not only to Constans but also to his earlier patron, Constantine II. Had Constantine II not been more preoccupied with invading Constans' territory, and had the invasion not placed Athanasius under suspicion for a time at the court of Constans, precisely because he had corresponded with Constantine II, a fresh intervention by a western emperor in eastern ecclesiastical affairs might have occurred even sooner than it did.Footnote ⁴⁴

As it is, from summer 341 to spring 342, the likelihood that Constans would become involved was increasing sharply. It was only natural that Julius and the medley of eastern exiles besides Athanasius would seek his support. Marcellus, the deposed bishop of Ancyra who had himself arrived in Rome in spring 340, appears to have set off again in summer 341, perhaps for the court of Constans.Footnote ⁴⁵ Moreover, as Constans was operating on the north-western frontier, he was almost certainly exposed to the influence of Maximinus, Bishop of Trier, who was a leading supporter of Paul and the very man who summoned Paul in 341 from his Pontic exile to return to Constantinople.Footnote ⁴⁶ It is not unduly teleological to say that the arrival of Paul at Constans' court in spring 342 was a natural tipping-point.

As a result, when Constantius was hastily travelling back from Constantinople to Antioch in March 342, his advisers are unlikely to have doubted that some sort of intervention from Constans was looming. Given the seriousness of the tensions within the Church, the potential gravity of that intervention could not be under-estimated. (The theological and ecclesiastical problems at stake were profound enough that later, in 345, Constans might have issued Constantius with a threat of war if Athanasius were not restored to the Alexandrian see.Footnote ⁴⁷) The possibility of a significant diplomatic escalation by Constans was dangerous territory for the eastern administration in 342. The defeat and death of Constantine II in 340, following his invasion of Constans' domains, meant that Constans now had at his disposal the full resources of the western empire; and although Constans was engaged in a successful spring campaign against the Franci, the eastern government was heavily committed in the ongoing war with Persia.

In these circumstances, it need not surprise us if Themistius had been briefed to omit the conventional protestations of amity. Such protestations would have been complacent in March 342, not only risking subsequent embarrassment but also potentially being construed as weakness if the speech were heard at the time by Constans' spies or the text later read in the West. Instead, by procuring from the young philosopher a speech that confined itself to eastern matters and which focused on the immediate issue of vindicating Constantius' handling of the crisis at Constantinople, the eastern court can be regarded as having addressed itself above all to an eastern audience on what, for that audience, was the subject of immediate controversy. This was the fact that, in the face of the murders and destruction of property caused by urban rioters, Constantius had been lenient.

Such an interpretation cannot be proven on present evidence. All attempts to date the speech come up against the shortage of specific content. But there are various reasons why the present interpretation is a more probable option than the alternatives.

First, even if we accept that Constantius did celebrate his vicennalia at Constantinople in November 343, there are practical reasons why delivery of the speech in late 343 or early 344 would be less opportune than 342. Not only would the issue of clemency, so central to Themistius' purpose, have lost its immediacy, being twenty months or more behind events. A second problem is that the speech would then have followed the failure of the Council of Serdica.Footnote ⁴⁸ Although the synodical letters issued by eastern and western ecclesiastical participants at the Council, replete with bitter mutual recriminations, were still being conveyed or absorbed, the tension between the courts themselves had in fact temporarily eased. Constans had taken a measured diplomatic approach; the two emperors had managed to agree and bring about an East-West council; neither emperor had attended, so neither was directly involved in the Council's failure; and, notwithstanding clerical wrangling, its aftermath witnessed mutual diplomatic overtures by the courts.Footnote ⁴⁹ This was a phase when it would have been politic to mention Constans in a panegyric, however briefly. A third problem with dating the speech to late 343 or early 344 is that Constantius was, by that time, in considerably less clement mood in his dealings with troublesome clerics. He would issue a death warrant against designated priests in the form of a proclamation authorizing governors to behead them.Footnote ⁵⁰ Themistius as panegyrist would go on to show himself quite capable of misrepresenting difficult situations; but to rest his central theme of clemency on a bare-faced reversal of the facts — emphasizing the sparing of the death penalty at a time when it was being actively deployed as a matter of policy — was not Themistius' style.

Later alternatives set the whole purpose of Themistius' speech at an even greater remove from any meaningful context.Footnote ⁵¹ This is hardly the territory of a career-making debut. Portmann recognizes this when he tries (albeit unsustainably) to push the occasion back to 351, when an act of clemency — in the shape of Constantius' first offer of an amnesty to supporters of the usurper Magnentius — again becomes immediately significant.Footnote ⁵²

Other more peripheral considerations also make it satisfying to date Themistius' Oration 1 to the early 340s. A major focus of Libanius' Oration 59 is, of course, the Battle of Singara, and the celebration of its inconclusive outcome as a victory. This being so, it has always been regarded, unexceptionably enough, as following sooner rather than later after the battle. But the date of the battle itself is uncertain. Often dated to 348, a persuasive case also exists to date the battle — that is, the nocturnal battle of Singara, since in fact there were two battles — to 344.Footnote ⁵³ On this basis, Libanius' speech could reasonably date from 344 or 345. That problem requires no detailed review here. What matters for present purposes is that a dating in the early 340s for Themistius' speech is consistent with either dating of Singara while still leaving room for the view that Libanius knew something about Themistius' text.Footnote ⁵⁴ Equally, even the earlier date for Singara would come after Themistius' speech, which is consistent with his lack of emphasis on a specific victory against Persia.

V THEMISTIUS' EARLY CAREER

Situating Themistius' Oration 1 in early 342 has implications for the chronology of his early career and for the development of his relationship with the Constantian regime. It is a long-standing puzzle that Libanius and Themistius apparently did not become acquainted before 350.Footnote ⁵⁵ The notion that two such high-profile teachers could have spent significant time in the same city without becoming acquainted is inherently implausible, given the obvious likelihood that they would wish to ‘network’ or that ‘networking’ would be thrust upon them by others. But Libanius is firmly placed in Constantinople between 340 or 341 and 342. He left the city in the aftermath of the unrest and Hermogenes' death — probably some months later — and spent, most likely, roughly one school-year at Nicaea (342–3) before settling down for five years at Nicomedia (343–8), and then returning to Constantinople in 348.Footnote ⁵⁶ The most obvious explanation for Themistius' presence in Ancyra is that he held a teaching post there at the time. Had Themistius already been engaged at Ancyra as a teacher by 340, when he was about twenty-three, then his absence from Constantinople during Libanius' first sojourn in the city has a ready-made explanation.Footnote ⁵⁷

Given the prominence of his father and grandfather, it is also unsurprising that Themistius, when still in his mid-twenties, should have been chosen by the court as a panegyrist.Footnote ⁵⁸ Coming as he did from an already prestigious family with a record of public appearances at a high level, the court's Ancyran stop provided the perfect opportunity to test the measure of the young scion of a fine family by giving him a brief to speak at a highly sensitive moment. The resonance of the occasion was no doubt increased by the fact that the occasion took place in Ancyra, which had been the seat of Marcellus, one of the leading episcopal exiles in the West. This might well have affected not only the court and leading citizens in the audience, but also the Christian community in the city — particularly as it was the ‘Nicene’ faction to which Marcellus belonged, and which bore the blame for Hermogenes' murder.

This view dovetails well with the received dating of Themistius' Oration 24, addressed to the Nicomedians, to the earlier 340s.Footnote ⁵⁹ The immediate professional outcome of Themistius' panegyric in early 342 may well have been that he delivered a series of lectures at Nicomedia in the school year of 342–3, when Libanius was in Nicaea. But by the time Libanius was himself installed in Nicomedia for the new school year in 343, Themistius is highly likely to have left for Constantinople. Not only did the two men apparently not meet at Nicomedia; a return to Constantinople in 343 also suits the early Themistian chronology. As Vanderspoel notes, Themistius' claim in 357 to have spent twenty years in Constantinople need not (and surely does not) mean twenty consecutive years, but rather a cumulative total.Footnote ⁶⁰ What we have, in that case, are seven years from the meeting of Libanius and Themistius in 350 through to 357 and some five years from 343 to 348, while Libanius was in Nicomedia. The remaining eight years would then have fallen before Themistius departed for Ancyra. If Themistius had arrived in Ancyra by 340, then he would probably first have come to Constantinople in about 332, at the age of about fifteen. This concurs with Vanderspoel's suggestion that Themistius arrived in the city that year, to begin his philosophical training under his father.Footnote ⁶¹ One might even speculate that Eugenius arranged for his son to be in Constantinople in time for the dedication ceremonies of 330 — not implausibly as his own father had also been associated with Byzantium, so that a link with the city seems to have been a thread in the family's history.Footnote ⁶² This would scarcely affect the calculation of twenty years in Constantinople, even if that figure is exact. Themistius might have arrived in Ancyra by 339, aged twenty-two, as Augustine did in Carthage; and he might have overlapped with Libanius for a short while, somewhere between 348 and 350, before they became acquainted.

Where Themistius was between 348 and 350 as a whole, however, remains unclear. A short overlap with Libanius is one matter. A two-year overlap without making one another's acquaintance is rather stretching a point. Themistius was probably absent again for much of the time, perhaps serving a stint in a publicly-funded chair in another city before returning to take up a position at Constantinople. What we know, however, is that he was back in Constantinople in about 350, by now probably established in a public chair, from which he would be adlected to the senate five years later.Footnote ⁶³

VI IMPERIAL IDEOLOGY

The present reconstruction of Themistius' early career suggests a very different trajectory in his relationship with the court from that which is normally posited. Received wisdom provides a rather telescoped account. Themistius makes a first impression on Constantius in 347 or 350, gains a distinguished teaching post soon afterwards and is adlected to the senate — with the benefit of a long and highly personalized letter of praise from the emperor — on the basis of five or eight years' acquaintanceship.Footnote ⁶⁴ What we are looking at instead might well be a more gradual process, taking thirteen years from an initial encounter in 342 to the adlection. On this view, Themistius' sojourn at Constantinople in the 340s (probably, as we saw, from 343 to 348) is likely to have been a crucial phase in the rise of his reputation at court, without which his prominence and honours in the 350s are much harder to understand.

How Themistius made his mark in the 340s, and how this coincided with prevailing views at court, might not be impossible to salvage. From the vantage-point of his later career, we know that part of Themistius' hallmark and talent was not only to avoid religious contention but to cultivate a positive language on the shared ground between Christianity and Hellenic culture. Already detectable in Oration 1, this approach was something that Themistius would go on to refine, develop and bring into the centre of Constantinopolitan political discourse.Footnote ⁶⁵

It is equally on these terms that we can perhaps best understand Themistius' emergence, in the 340s, as a person of interest to the Constantian court. According to Libanius, the official sophist of Constantinople at the time, Bemarchius, toured the eastern provinces shortly before the rioting at Constantinople broke out. He repeatedly delivered a keynote speech (which he would also give in Constantinople) about the construction of the Great Church at Antioch:Footnote ⁶⁶

Bemarchius, whose job and salary were normally based in Constantinople, is hardly likely to have made this considerable lecture tour on a pro bono, sightseeing basis. What we are looking at is more likely to be an overtly Christian vein of court propaganda being rolled out across the eastern provinces.

Coinciding with this, in 341, was a re-statement of the ban on ‘superstition’ and sacrifice inaugurated by Constantine.Footnote ⁶⁷ The copy of the law of 341 which was used in the fifth century by the editorial board of the Theodosian Code was of a version addressed to Madalianus, vicarius of Italy and Africa, under Constans' jurisdiction. Constans' name has dropped out in the transmission of the text of the law but it will clearly have been there originally.Footnote ⁶⁸ As we would expect from the collegiate ‘voice’ in which the laws of the time were written, this law also gave Constantius II's name, which does survive. Although this does not in itself mean that the law was issued in the East, there is a strong likelihood that generalitas was in this case recognized, at the time, in both halves of the Empire.Footnote ⁶⁹ Not the least reason for this, in the present case, is that Constans was re-stating to Madalianus a principle — the bar on blood sacrifice — which had already been introduced by Constantine in the East, in 324, and which, on the balance of evidence, appears likely to have remained in place (albeit poorly enforced) until the reign of Julian.Footnote ⁷⁰

The corollary is that Bemarchius' lecture tour might well have been paralleled by correspondence from the court of Constantius to the eastern administration, re-stating the ban on sacrifice in the same way as Constans was doing. The only notable difference in the East is that the legal re-statement was tailored to an eastern audience through its coincidence with, and promotion of, the completion of the Great Church at Antioch.Footnote ⁷¹

What is noticeable, however, is that any prospect in 341 of ideological momentum on the limitation of pagan practice yielded no material outcome. Jill Harries has recently observed that ‘Constantius' cautious, incremental approach extended to religious reform’, noting that it was only in the mid-350s that he legislated for the closure of temples.Footnote ⁷² Whether we should simply take this incrementalism for granted — as a reflection of personal hesitancy, perhaps, or a pragmatic desire not to alienate a largely pagan administration and society — is another matter. In the circumstances of 342 and 343, both Constans and Constantius had been given significant occasion to reflect on the challenges presented for imperial policy by a fractious Christian clergy. The experience might not have been an unmodified delight and its effect was not necessarily to advance the Church's wider cause.

In the West, we can perhaps already detect this. Another characteristically astute observation by Harries is that Codex Theodosianus 16.10.3 should be seen as a response, in 342, to Roman senatorial queries in the near aftermath of the law of 341. Sacrifices were banned (Constans had said), so could temples (the senators asked) remain open without sacrifice? To this, Constans said ‘Yes’, on the grounds that temples served a public function as facilities for holding festivals.Footnote ⁷³ Constans' law in favour of retaining temples is dated to November. Harries' convincing rationale gives us, therefore, a law of November 342. This places it after the year's campaigning season had passed, when the risk of military conflict between Constans and Constantius had given way to the diplomacy which paved the road to the Council of Serdica. It was a good context in which the two emperors might agree, at the request of the senatorial order, to constrain the claims of the Church on wider society.

Equally, in the East, the Church might be seen as having paid for its internecine violence and its receipt of an initially startling clemency. The price was a cooling of imperial ideology from overtly Christian themes. Shortly after the re-statement in 341 of the ban on sacrifice, Constantius was co-named in the decision that temples could remain open, with a strong likelihood of generalitas being understood at the time and being conveyed by imperial letters.Footnote ⁷⁴ In 343, the limits of clemency had also become clear. Troublesome clerics in the East began to face their doom. The proconsul of Constantinople, Donatus, would find himself in receipt of imperial orders to arrest (among others) Bishop Olympius of the see of Aenus, in the province of Rhodope in Thrace; while a number of other clerics were targeted for beheading.Footnote ⁷⁵

In the meantime, Constantius had campaigns against Persia to manage and a continuing medium-term risk from the West under Constans. It was no time to alienate his provincial aristocracies. In this context, Themistius offered a coherent language of power, free of contentious religious encumbrance but rooted in a deep tradition of Hellenistic philosophy — a language that resonated with eastern aristocracies. This was a language which could serve the eastern governing class of the 340s as a whole, both the court and the wider aristocracies. It is only realism to conclude that Themistius spent much of that decade honing this language, chiefly in Constantinople, and offering it to the imperial government when occasion provided. That he should have done so in a private capacity for some years, before he achieved the official recognition which came to him in the 350s, need not surprise us. What we might see, in March 342, is what the court probably saw: a young man worth keeping an eye on. As it turned out, he spoke a unifying language better than anyone else and, in doing so, proved in the coming years to be a man for the times.