Sulpicius Alexander and the Soldier Historians of the Later Roman Empire

I Sulpicius Alexander’s Historia

The sole surviving evidence of Sulpicius Alexander’s oeuvre is the already mentioned Historiae by Gregory of Tours. The bishop of Tours cited long passages related to the Franks extracted from Sulpicius’ original work.^[5] Gregory’s aim in quoting Sulpicius Alexander and a later historian of the first half of the fifth century, Renatus Profuturus Frigeridus, is clear – the bishop of Tours attempted to collect the evidence for the origins and history of Frankish kingship up to the time of Clovis.^[6] He did not hide his confusion at the terminology used by both authors in describing Frankish leaders as duces, regali, subregoli, and reges.^[7] Quoting Sulpicius, Gregory passed on to us clauses connecting parts of the original narrative concerning the Franks to the rest of this historical account, revealing that these were in fact merely excursuses marginal to the main topic – the imperial politics, usurpations, and civil wars:

Because when he says that Maximus, having lost all hope of power, retreated to Aquileia like a fool, he adds: At that time, the Franks, who had Genobaudes, Marcomer, and Sunno as their commanders, invaded Germania, and, after having killed many people when they crossed the border, they laid waste to the most fertile districts and also menaced Cologne. When this news reached Trier, Nannienus and Quintinus, the magistri militum to whom Maximus had entrusted his son, still a child, and the defence of Gaul, raised an army and gathered in Cologne. [...] But in the fourth book, when he narrates the execution of Victor, the son of the usurper Maximus, he says: at that time Carietto and Sirus, replacing Nannienus, were stationed with their army in Germania to oppose the Franks. And after a while, when the Franks had brought plunder from Germania, he added: Arbogast, who did not want to delay anything, encouraged Caesar to enact the punishments that the Franks deserved, unless they immediately returned everything they had stolen from the slain legions the previous year and handed over those responsible for the war, who would be charged with the treasonous violation of the peace. He recounts that these things happened when there were commanders (duces); and then he says: after a few days, when a brief meeting was conducted with Marcomer and Sunno, the royal men of the Franks, and when, according to custom, hostages had been demanded, he retired into the winter quarters at Trier. [...] While in the East various things were going on in Thrace, in Gaul public life was disrupted.^[8]

The geographical scope of Sulpicius’ original work extended beyond Gaul’s borders and included events in the eastern part of the Roman Empire.^[9] Chronological limits are more difficult to define, given the nature of Gregory’s selection. The fact that the last quoted event is Eugenius and Arbogast’s expedition to the Rhine and renewal of foedus with Frankish and Alammanic kings simply means that this was the very final mention of Frankish reges by Sulpicius Alexander. His work could have extended well into the fifth century but taken no further note of the subject matter of Gregory’s interest. However, there are some indications of the chronological as well as thematic framework of Sulpicius’ oeuvre.

Events marked as quoted from books 3 and 4 give us the idea on how extensive and lengthy they were in their original form. Citations describe military developments in great detail. They indicate that naming individual commanders, as well as relating the course of campaigns, battles, and skirmishes, even the marginal ones unfolding at the borders that parallel the main narrative about the civil war, were all characteristics of Sulpicius’ Historia. The timespan of the fragmented books appears to be circa one to two years structured by seasons. Although Gregory’s Et post pauca (“And after a while”) indicates that Arbogast’s handling of the Franks and return to Trier under Valentinian II (F4) still belongs to book 4, making it cover events from late 388 to at least 390, the remaining material regarding Valentinian II’s forced seclusion (F5) and Arbogast’s military campaign against the Franks (F6) is distinctly set apart by Gregory’s ‘research’ commentary on the terminology of kingship: Cum autem eos regales vocet, nescimus, utrum reges fuerint, an in vices tenuerunt regnum. Idem tamen scriptor, cum necessitates Valentiniani agusti commemorat, haec adiungit (“But although he calls them royal, we do not know whether they were kings, or whether they held power alternately. But the same writer, when mentioning the misfortunes of Valentinianus Augustus, adds the following”) (F4-F5), implying it in fact belonged to a different book. So is the case for the passage describing Eugenius foedus with the Franks preceded by Iterum hic, relictis tam ducibus quam regalibus, aperte Francos regem habere designat, huius que nomen praetermissum, ait ... (“Then again, without mentioning either commanders (duces) or royal men, he openly states that the Franks had a king, but without giving his name, and he says ...”) (F7). Valentinian stayed in Milan until 391 and only ended up under Arbogast’s ‘protection’ when he was moved to Vienne in Gaul afterwards. Therefore, book 4 could hardly stretch beyond that year.

Although there was no strict rule about the number of books a history work would be divided into in Antiquity, there still existed conventions and tendencies that may indicate whether Sulpicius’ Historia actually concluded with book 4, as believed to be the case by Paschoud, according to whom four books of Historia would span from 376/378 until circa 395.^[10] Ancient narrative histories were often composed of a number of books that were a multiplication of five or six.^[11] The five or six books structure would be, therefore, more in line with the habits of ancient historians, unlike the four books postulated by Paschoud as part of his argument in favour of Sulpicius Alexander being Ammianus’ continuator (see below). Even the unusual 31 books of Ammianus Marcellinus that could serve as an exception to push for Paschoud’s hypothesis may differ from the historian’s originally intended composition.^[12] It is likely that Sulpicius Alexander’s work followed that convention and was divided into at least five books. In such a case, the events leading to the next civil war and the conflict itself – a material of a little less than 3 years – were probably covered in book 5. This in turn may allow us to reconsider the time span narrated in the earlier books, the first and the second one, and the timeframe of the entire work.

The invasion of Italy by Magnus Maximus’ in 387, Valentinian II’s flight to the East, and the resulting civil war meant that the period of 387–389 abounded with material for a grand narrative. Earlier years were less eventful and dense with a subject matter for political history so books 1 and 2 could cover a lengthier period. With a plausible average of two to three years addressed in a single book, the beginning of Sulpicius’ narrative could be placed in the early 380 s. The most obvious choice of an event opening this narrative about civil wars would be Magnus Maximus’ proclamation as the emperor and Gratian’s death in 383. Such a beginning corresponds well to some of the themes included in the surviving fragments and the likely conclusion of Sulpicius’ Historia with emperor Theodosius’ triumph over Eugenius. It would be a history of usurpations quelled by that emperor.

Woudhuysen has suggested to me that the timeframe of Sulpicius Alexander’s narrative proposed above would perhaps coincide with the ending of the period covered by the second edition or a continuation of Sextus Aurelius Victor’s work the existence of which was recently postulated by him and Stover.^[13] That Sulpicius Alexander would be continuing Victor is indeed an intriguing possibility and one that would fit within the landscape of Latin historiography in the late fourth and early fifth centuries that is emerging as a result of Stover and Woudhuysen’s study. If Victor’s lost grand narrative did indeed hold the cultural significance they argue for (and I am convinced they are right in that regard), then Sulpicius Alexander undertaking a continuation of Victor rather than Ammianus Marcellinus, whose reception at the time appears rather limited in comparison, seems the more plausible of the two options.^[14] However, higher probability does not equate certainty and further exploration of that idea would necessitate a separate study that would account for the scholarly discussion that is bound to be initiated by the recent publication of Stover and Woudhuysen’s monograph.

Gregory’s use of Orosius in narrating Stilicho’s handling of the Franks and the barbarian crossing of Rhine in 406 is a clue in defining the temporal confines of Sulpicius’ narrative. As has been already pointed out by Paschoud, however, the strength of that evidence is limited given that Gregory cited only the passages that mentioned specifically Frankish kings and Sulpicius Alexander may have very well continued his narrative further, simply not mentioning in it what was of interest to Gregory.^[15] An important yet indirect indication of the end of Sulpicius Alexander’s narrative is its intertextual inspiration, which may betray author’s intent in composing the work. As argued convincingly by Humphries, Sulpicius’ line about Maximus remaining without hope and mentally deteriorating in Aquileia bears a striking resemblance to a panegyric delivered by Latinus Pacatus Drepanius in 389 in honour of the emperor Theodosius.^[16] This inspiration together with the hostility towards Magnus Maximus and his criticism of Quintinus in Sulpicius Alexander’s work, may superficially appear to counter the identification of Sulpicius Alexander with Alexander, tribunus et notarius serving under Magnus Maximus, proposed by Seeck. However, hostility towards the usurper expressed by his former officer could be easily explained as a strategy for gaining favour with the Theodosian regime or/and “settling scores” with leadership that failed its men. Gregory did not note from which books the passages about Arbogast and Eugenius’ usurpation originated.

Nevertheless, it is easy to envision a history of five or six books written with the purpose of praising the emperor Theodosius and his regime. It would have described his victorious wars against the two usurpers who were to blame for the death of Gratian and Valentinian and composed by a former official or a soldier of the first usurper, who, not unlike his contemporary and possible acquaintance Symmachus, sought to win favour with the emperor known for his affinity for history.^[17] Theodosius’ affinity for history may be deduced from Nicomachus Flavianus’ alleged dedication of his Annales to Theodosius stressed by Nicomachus’ descendants (ILS 2984). Regardless of whether Flavianus did dedicate his work to the emperor or not in 431, Theodosius’ interest in history was considered a fact.^[18] Another source that corroborates it is the Epitome de Caesaribus, which though recently proven to post-date Jordanes must reflect in the relevant passage a more contemporary tradition.^[19] Sulpicius Alexander’s work may have been aiming at meeting assumed literary taste of the emperor. If he had been involved with Magnus Maximus regime,it would be one way to ingratiate himself with the house of Theodosius. A shift of attitude towards Magnus Maximus, similar to the one speculated by Zecchini and discussed here, is observable in works of another member of gens Sulpicii at the time – Sulpicius Severus. Could there be a connection between the two?

II Sulpicius Alexander and Sulpicius Severus – a connection beyond names?

Sulpicius Severus, a Christian author and ascetic active at the end of the fourth and the beginning of the fifth century, who originated from the elites of Aquitania and married into a consular family, pursued a career as an advocate before abandoning it for monastic life, is rarely mentioned when considering Sulpicius Alexander. Even though Sulpicius Severus shared with Sulpicius Alexander not only the nomen gentilicium, but also the interest in Magnus Maximus’ usurpation, and was active most likely in the same generation. Sulpicius Severus’ opinion towards Magnus Maximus and his usurpation is particularly interesting set next to that of Sulpicius Alexander, as it shifted in time and in this way was rather revealing of the views of the Gallic elites of the late fourth and early fifth centuries. In the earliest of Sulpicius Severus’ works, Vita Martini, Magnus Maximus is presented negatively as vir ferocis ingenii who inspired undignified servility amongst bishops attending him (V. Mart. 20.1–3 and 8–9). In the later Chronicle image of this emperor is rather neutral, moving agency from him to bishops (Chron. 2.49.2 and 2.50.2). In Dialogorum Libri Duo, the latest of Sulpicius Severus’ works, author adopts an apologetic tone, praising the emperor and his wife’s piety and virtues. He argued Magnus Maximus would be risking his life if he did not accept the imperial acclamation. Sulpicius Severus claimed that the emperor was led astray by Ithacius and other bishops in the case of Priscilian. Magnus Maximus’ greed was explained by the needs of the state bankrupted by his predecessors. Last, but not least, in Dialogorum Libri Duo Sulpicius Severus stressed the importance of Magnus Maximus’ relationship with Martin of Tours, whereas in Vita he played it down (Dial. 2.6.2–3; 3.11.2, 8–11; 3.12.2, 4; 3.13.1). It appears with the passing of time, Sulpicius Severus felt more comfortable with sharing his real view of Magnus Maximus, giving lip service to the Theodosian propaganda in Vita published in 397, but rehabilitating the usurper in Dialogorum libri duo. It gives an interesting counterpoint to Sulpicius Alexander’s depiction of Magnus Maximus in the context of his speculated alignment with the regime of that emperor prior to composing his historical work.

An argument in favour of a connection between Sulpicius Severus and Sulpicius Alexander, as well as the latter’s ties to Gaul and the possibility of his work reflecting the official stance of the provincial elites at the time, is the dissemination of Historia itself. There are of course multiple ways in which this history work could have ended up in Gregory of Tours’ book collection: he may have inherited it from his family in Clermont, he may have obtained it himself or it might have been a part of the episcopal library of Tours. In all these cases, the work – unknown anywhere else – would have circulated in Late Antique and Merovingian Gaul; in the last one, it would have been stored in a city Sulpicius Severus had strong connections to – the seat of his hero, Martin of Tours. If the two Sulpicii were somewhat related, circulation of their works amongst the same social circles of Late Antique Gaul would be a possibility and explain Sulpicius Alexander’s Historia in sixth-century Tours. Another piece of circumstantial evidence for that would be the Gallic perspective ascribed to Renatus Profuturus Frigeridus, whose work was clearly used by Gregory of Tours next to that of Sulpicius Alexander and likely had similar patterns of circulation in sixth-century Gaul.^[20]

None of the other Late Antique Sulpicii can be decisively tied to Sulpicius Alexander in a convincing manner. A survey of the prosopographical material does offer some indications, even if of an indirect character, regarding the social background of this historian (see the Appendix). Sulpicii attested in this period do share some characteristics. Most of them can be identified or surmised to belong to the curial class. Much like Sulpicius Severus, who even though he became a close friend of a member of a senatorial class, Paulinus of Nola, and married into a consular family, likely belonged to the upper strata of the Aquitanian curial elites. This is suggested by his earlier legal career and his coming into great wealth only through his marriage.^[21] A later Sulpicius, a contemporary of Gregory of Tours, the bishop of Bourges Sulpitius Severus could be vir valde nobilis et de primis senatoribus Galliarum, a man of great nobility and descending from the most prominent senators of Gaul, but it seems the Sulpicii of the Later Roman Empire, even the ones in Gaul, by and large remained below the senatorial class, numbering amongst the curiales, governors and imperial officials of lower ranks.^[22] This is the likely background of Sulpicius Alexander himself based on the occurrence of his nomen gentilicium. Gaul may be tentatively pointed to as the most likely region of Sulpicius Alexander’s origin and activity, given the transmission pattern of his work and potential connections to Sulpicius Severus. However, further evidence about his social background can be, in fact, inferred from Sulpicius Alexander’s historical writing.

III A soldier’s perspective

The thematic focus of Sulpicius Alexander’s work is clear even in its fragmentary preserved state – the political turmoil of the late fourth-century Roman empire. In his selection of passages discussing Frankish kings and chiefs, Gregory of Tours transmitted for us glimpses of the main narrative, in which the Franks were incidental and even marginal. Sulpicius covered the civil wars of the 480 s and 490 s in the West, but clearly made some remarks on the politics in the East at the time, “While in the East various things were going on in Thrace, in Gaul public life was disrupted”.^[23] This in itself is unremarkable and reflects some of the typical historiographic interests of Latin literature as well as its usual phrasing, but what is characteristic of Sulpicius’ narrative is the perspective and details he provides us with. The description of Quintinus’ campaign beyond the Rhine, in the wake of the victory at Silva Carbonaria, and the resulting battle at castellum Nivisium in 388 in Book 3 is full of topographic details suggesting intimate knowledge of the area, possibly gained first-hand (F1):^[24]

At the break of dawn, they entered the forest for the battle with Quintinus in command. Around noon, they were completely lost, confused by their ignorance of the roads. At last, finding everything solidly closed off with enormous fences, they tried to clear a path to the marshy fields that lie next to the woods. Here and there enemies appeared who, secured to tree trunks or standing on barriers, as if from the tops of towers, threw arrows covered with poison made of herbs as missiles, so that wounds inflicted on the surface of the skin and not in vital places would undoubtedly lead to death. Then the army, encircled by a greater multitude of enemies, rushed hurriedly to the open ground of the fields, which the Franks had left accessible. And the first horseman plunged into the abyss, where bodies of men and animals were mixed up, and they were, in turn, all pushed down by the fall of their own [horses]. The infantrymen, too, who were not pushed down by the weight of their horses, got stuck in the mud, struggling to free their feet, and, scared, hid again in the woods that they had barely left moments ago. With the ranks thus in disarray, the legions were slaughtered. Heraclius, tribune of the Joviniani, and almost everyone who commanded the soldiers, was killed, and the night and hiding places in the woods provided safe refuge for few. This he tells in the third book of his History.^[25]

The criticism of how the campaign was conducted, commentary on tactics and placing its failure with Quintinus and the soldiers’ foolishness may imply an emotional attitude stemming from either a personal involvement in this expedition or some kind of association through joint military service with officers engaged in it. Nevertheless, it is worth pointing out that if Sulpicius Alexander participated in said campaign, then he certainly did not employ the experience in the way Ammianus Marcellinus had instrumentalised autopsy as a narrative tool in his history.^[26]

However, there is a different, distinct trait that Sulpicius’ prose shares with that of Ammianus. One that perhaps indicates that, much like Ammianus, Sulpicius Alexander had a military background: an interest in the battle losses of tribunes. Sulpicius Alexander not only included the information regarding the death of a tribune next to otherwise unspecified military leaders, who also fell during said battle, but named him and the unit he belonged to. He offers no further explanation on why Heraclius’ death should be of particular interest to his audience, nor is there any indication that this tribune appeared in the remainder of the historical narrative that was preserved. It is unlikely that this oddity is a result of Gregory’s intervention in the cited text – quoted passages served him to consider the historiographic material on the Franks, not Late Roman officers. This is a rather uncommon characteristic throughout Latin historiography, but one that, as already stated, Sulpicius shares with Ammianus Marcellinus, who reported such losses several times throughout his narrative: after an Alammanic ambush on Arbitio’s troops; at the battle of Argentoratum; at the conclusion of the siege of Amida; over the course of Julian’s campaign against Persia; and in skirmishes against the Goths preceding the battle of Adrianople, as well as in the battle itself.^[27] In a number of cases Ammianus reported the names and the units of tribunes who fell in battle or were otherwise killed; this seems to have been done in a manner very similar to that of Sulpicius Alexander’s mention of tribune Heraclius’ demise. The question is whether we are dealing with an imitation or an independent development of a literary characteristic. A characteristic of their prose that would be conditioned by the military background of both authors.

Paschoud associated Ammianus Marcellinus with Sulpicius Alexander, speculating that the latter wrote a continuation of the historical narrative of the former.^[28] However, if that was the case then there were no traces of that in Sulpicius’ style, which is clear, concise and brief in contrast to Ammianus’ more flowery writing.^[29] It is also likely that the time that had passed between the composition of both works had been relatively brief. Ammianus finished his work before Valentinian II’s death in 392, while there are reasons to believe Sulpicius’ history did not deal with events beyond Theodosius’ reign and was likely produced before his death in 395 or not long after.^[30]

Even if Sulpicius was inspired by Ammianus and took on his interest in battle losses of tribunes for no other reason than imitation, which taking into account the above mentioned considerations is rather unlikely, he would still require sources that could provide him with such specific information. The name of the tribune, Heraclius, and that of the unit, legio Ioviniani, is a very specific piece of detail placed next to a generalisation about other officers, who died in this battle: paene omnibus qui militibus praeerant extinctis (F1). Woods plausibly identified this legion as legio Ioviani, which usually operated in unison with legio Herculiani.^[31] These elite troops were originally established by Diocletian and are attested in multiple sources well into the fifth century when both legions are listed under the command of comes Italiae.^[32] Woods assumed that the legio Ioviani in question would be the seniores branch rather than iuniores, though he did not quite formulate the reasoning for his supposition. Wood’s identification is followed by Van Hoof and Van Nuffelen in their commentary on the fragments of Sulpicius Alexander’s narrative, but frankly may not be correct.^[33]

Yet the element of unit identification that is truly relevant for analysing Sulpicius Alexander’s prose is not the ambiguous seniores-iuniores designation, but the very qualification of legio Ioviani as a palatine legion. The elite status of this unit explains why it and its tribune were singled out and mentioned by name in the historian’s report on Roman battle losses, while deaths of other commanders who perished in the same clash were reported only in general, collective terms. Distinguishing the death of an elite palatine legion tribune at the hands of Franks served to emphasise the scale of this defeat. It also showed the interest of the author and elements of his expected audience in specific commanders of the units that constituted the elite even amongst the Imperial field army – their likely social network.

The combination of precise and generalised data runs in a similar fashion to the way that Ammianus reports battle losses.^[34] There are several cases in which he notes deaths of tribunes in battle, sometimes providing his audience with a precise number (Amm. Marc. 15.4.8; 24.3.1; 31.13.18). There are a few mentions of lost tribunes by name, some of them commanded palatine units, much like Sulpicius Alexander’s Heraclius: Bainobaudes, tribune of Cornuti (auxilia palatina according to the Not. Dig.) at Argentoratum; Iulianus, Macrobius, and Maximus in a clash with Persians during the Roman return from Julian’s Persian expedition of 363; Potentius, tribune of Promoti (vexillatio palatina), who was the son of Ursicinus and therefore someone to whom Ammianus had a personal connection.^[35] The passage concerning Iulianus, Macrobius and Maximus merits a short additional commentary. Deaths of these three tribunes are also reported by Zosimos: Ἀπέθανον δὲ καὶ Ῥωμαίων τρεῖς λοχαγοὶ γενναίως ἀγωνισάμενοι, Ἰουλιανὸς καὶ Μαξιμιανὸς καὶ Μακρόβιος (3.30.4). This raises an obvious question on how truly characteristic of a soldier-historian’s perspective, which I am proposing, is the inclusion of the tribunes’ deaths? Zosimos does not belong to the discussed category of historians of a military background, yet he still reports the same piece of information in a similar manner to Ammianus. However, neither Zosimos nor what remains of his main source – Eunapios’ Universal History – account for that kind of details (tribunes’ deaths, describing involvement of specific units) anywhere else outside the part of the narrative on Julian’s expedition against Persia. Even in the case of Tribigild’s rebellion in Phrygia, of which Eunapios had personal knowledge and could possibly provide that sort of details, Zosimos’ narrative only mentions the death of Roman commander Leo (5.17.2; see below for further discussion).

In the case of Iulianus, Macrobius, and Maximus it is difficult to attribute specific units to them. They were mentioned in the context of involvement of legiones Ioviani and Herculiani in the battle, but also Iovii and Victores, called by Ammianus legiones, but in Notitia dignitatum listed as auxilia palatina (Not. Dig. or. 5.28, occ. 6.44). This gives the impression of the kind of knowledge obtained first-hand or through witnesses, rather than any official records. Sulpicius’ access to similar kinds of information and his use of it reveals the character of the network he may have possessed. The interest in the final fate of tribunes shared with Ammianus also speaks to a perspective focused on military matters, one stemming from an understanding of the function of the mid-rank officers but also from a personal attitude towards men in that role and a desire to recognise their deaths.^[36] Tribunes were an important element of the Roman imperial war machine in the middle of the command structure.^[37] Oftentimes those chosen for the more prominent commands originated from among the protectores and therefore belonged to the elite of the Roman military.^[38]Ammianus’ understanding of this reality originating from his military background is expressed through the attention to the group and individual losses of tribunes.^[39] Tribunes are the most often mentioned kind of Roman officers in Ammianus’ narrative and their prominence in the narrative corresponds to the passages describing the time of Ammianus’ active duty, diminishing in the later parts of his history.^[40]

Sulpicius Alexander seemingly shared such sensibilities. This is of course based merely on the material that was preserved via Gregory of Tours. F1 is the only fragment of Sulpicius Alexander’s narrative to include that level of detail and it would be easy to dismiss the mention of tribune Heraclius’ death as coincidental. Doing so, however, would require us to presuppose Sulpicius Alexander included that piece of information for no good reason. The passage offers no indication that Heraclius appeared in the earlier parts of the narrative nor does it immediately elaborate on his particular significance. One can only surmise that his interest is based on his command of an elite, palatine unit. If Sulpicius Alexander had been himself uninterested in such a detail or assumed it to be unwanted by his audience, but aimed to stress the magnitude of Roman losses, he could have used many other conventional literary devices to do so. The preserved material reveals Sulpicius Alexander as a fairly competent and intentional author. It is therefore more plausible to conceive of this element of the discussed passage as evidence for the type of information repeatedly included throughout the narrative. It would have a relevance within the context of the entire work, one that it somewhat lacks if only treated as a peculiar element of the passage pertaining to Quintinus’ defeat.

Including tribunes in historical narrative in such a way hardly has a precedent in extant Latin historiography. The closest analogy may be found in Caesar’s late Republican historical commentaries. Some late Republican centurions played a comparable tactical role on the battlefield to that of Late Imperial tribunes. The most senior centurion of the cohort commanded this tactical unit numbering 600 men roughly similar to the effective size of Late Antique auxilia (400–1200) or legiones (800–1200). Although Caesar mentions individual centurions by name in a number of instances, it is within the context of them conducting daring actions, not as Sulpicius Alexander or Ammianus Marcellinus do in reporting deaths of earlier unmentioned tribunes in their narratives (see, e. g., BGall. 3.5; 5.44; BCiv. 2.35). In fact, the only time Caesar accounts for the death of a centurion is in a case of a certain Caesarian centurion named Fabius attempting to kill Publius Attius Varus during the retreat of Pompeian forces during one of the skirmishes in Africa in 49 BC and dying in the process (BCiv. 2.35). In all the remaining cases of centurions’ deaths, only numbers are given, serving as a narrative device to illustrate the gravity of skirmishes and battles (BGall. 6.40; 7.51; BCiv. 3.67; 3.71; 3.99). Ammianus Marcellinus, on the other hand, mentions the names of lost tribunes and their units for no apparent narrative reason other than relating the very fact. Sulpicius Alexander’s mention of tribune Heraclius’ death seems to be of the same kind, with no reference to earlier appearance of said Roman officer in Sulpicius’ narrative. The similarity between Caesar’s narrative and that of Ammianus and Sulpicius in this regard is therefore only superficial. This peculiar trait appears to be a new development in the Latin historiographic literature.

One could point to many other ancient authors who had military experience that they drew upon in their histories: Xenophon, Polybios, Velleius Paterculus and Arrian to name just a few. None of them, however, belonged to a fully professionalised cadre of soldiers. Military service was simply a stage of their public careers. Therefore, it is unsurprising that such authors and their audiences demonstrated little concern in reporting the kind of information about individual officers that could be possibly of interest to other members and former members of an officer cadre of the fully professional military of the Later Roman Empire. There are instances in the writing of Xenophon of reports of injuries to mid-ranking officers without such details having an obvious role in his narrative that could be compared to Sulpicius Alexander and Ammianus’ works.^[41] Indeed Xenophon’s experience as a “mercenary” could be perhaps likened in certain ways to that of a professional officer in the Later Roman military. However, there is little indication that Xenophon served as a direct model for Ammianus and other writers of his era.^[42]

The role of reporting tribunes’ battle deaths for the portion of Ammianus’ audience, significance for the author himself as well as the very access to that sort of information is consistent with his military background. It cannot be overstated how brimming with details on military matters and personalia Ammianus’ narrative is. As noted by Woods, Ammianus did not aim to offer a thorough examination of the organisation and practice of the Roman army as an institution, nor did he intend to narrate histories of specific units.^[43] Nevertheless, his work facilitates many modern studies of particular aspects of the Late Roman army, for example allowing us to put together lists of commanders of elite units of scholae palatinae or examining the roles of officers of different ranks not only just on the field of battle, but also within the broader structure of the imperial government.^[44] It is difficult to find a comparable perspective and selection of information in any other earlier extant Roman history work. Yet was Ammianus Marcellinus and his history truly unique or perhaps a case of a broader phenomenon? Is Ammianus’ example sufficient to draw conclusions in regard to Sulpicius Alexander’s professional experiences and social networks? Pieces of the narrative of yet another “soldier historian” – Magnus of Carrhae – serve to strengthen these suppositions.

IV Magnus of Carrhae

Magnus of Carrhae is perhaps best known from John Malalas’ Chronographia, which contains quotations of his work in part of the narrative devoted to emperor Julian’s Persian campaign (13.21–23). The fourth-century author described said military expedition in detail drawing on his own experiences as a participant (13.21). His identification with the tribune Magnus, who during the course of the campaign heroically stormed Maiozamalcha through an underground tunnel, is not an uncontested one, but appears to be the still dominant view in the field.^[45]Even if Magnus was not the tribune in question, a military background of this participant of Julian’s campaign is readily apparent.^[46]

The characteristics of Magnus’ work that can be inferred on the basis of surviving citations are terminological archaization, suggestive of a Classical education and inspiration from Classical models, and a very detailed narrative including toponyms, emperor’s speeches, names of individual commanders, credible numbers of soldiers involved in the campaign and even engagements of specific units.^[47]These are reminiscent of Ammianus Marcellinus’ selection of details and, to an extent, the ones provided by Sulpicus Alexander. However, none of the fragments of the work of Magnus of Carrhae transmitted by John Malalas contain the distinct characteristic of both Sulpicius and Ammianus: the reporting of individual tribunes’ deaths. The presence of such a piece of data in Magnus’ work would further corroborate that such a selection of information is indeed conditioned by the military experience and service of the author.

Nonetheless, there is a piece of evidence that may suggest Magnus’ description of Julian’s expedition against Persia did in fact include such details. The already quoted passage of Ammianus Marcellinus concerning deaths of Julian, Macrobius and Maximus in a battle with Persians bears striking resemblance to that found in Zosimos’ Historia Nova, differing only slightly in the order of listed tribunes and form of one of their names (Maximianus instead of Maximus).^[48] The common source of Zosimos (through Eunapios) and Ammianus is nearly a matter of an academic consensus.^[49] What is, however, a contested issue is the character of that source. Engaging thoroughly in this discussion exceeds the subject of this article, but a broad description of existing positions is not out of order.^[50] The two main proposed identifications name either Oreibasios of Pergamon or Magnus of Carrhae as the common sources of Eunapios/Zosimos and Ammianus Marcellinus.^[51] A more recent hypothesis proposed by Bleckmann is based on a model of complex literary relationships in which authors derive their material from many sources engaging in a wider historiographic discussion between each other.^[52] According to his hypothetical stemma of textual relationships, that he describes as more of an illustration of his model rather than a reconstruction, the common “branch” of an influence of Ammianus Marcellinus and Eunapios/Zosimos would be Nicomachus Flavianus, Oreibasios, and, finally, the diary of Philagrios. Philagrios was a participant of the campaign and Libanios’ addressee, who would also be the source for Magnus of Carrhae. However, no direct relation between Eunapios/Zosimos’ histories and that of Magnus of Carrhae would exist within this model.

A definite solution to this problem is certainly out of reach with the current evidence. However, Oreibasios as a source of the piece of information on the tribunes’ death, which is of interest to our topic, is the less likely candidate. Eunapios’ testimony on his composition presents it as a ὑπόμνημα written for the sole purpose of assisting him in working on his history (F15). There is little to suggest his work was known to contemporary authors beyond Eunapios himself and any later evidence for it appears to be dependent on the transmission of Eunapios.^[53] That attestation also describes the work as biographical in nature with focus on the deeds of Julian, rather than details of his campaign against Persia and military matters. Finally, the narrative of Zosimos based on Eunapios’ shows little interest in details such as battle losses of tribunes, the only such mention belonging to the part of his narrative on the Persian campaign. This raises the question – would Oreibasios include such information in a collection of notes intended for his friend, who demonstrated indifference towards such details?

The exclusion of Oreibasios does not immediately mean the piece of information noted both by Zosimos and Ammianus originated from the history of Magnus of Carrhae. Given the characteristics of his work known through John Malalas’ citation, however, he remains the most likely source for this information. He was, after all, the author interested in individual commanders and specific units operating during Julian’s campaign. If such is the case, then his work shared with Ammianus Marcellinus and Sulpicius Alexander the very same characteristic discussed above.

V Literati in arms

Ammianus Marcellinus and Magnus of Carrhae were by no means the only cases of (ex)soldiers displaying both literary ambitions and skills necessary for composing such works in this period. It seems the final separation of military and civil elites ushered in by Constantine’s reforms created a societal landscape from which a new group of Late Antique literary authors and audiences came to be.^[54] In the fourth and fifth centuries there are a few cases either of literati, who used to be soldiers, or of men in active military service, who exercised skill and literary talent or were assumed to possess such. Magister equitum Promotus had been requested by the famous fourth-century rhetorician Libanios to write an account of the emperor Theodosius’ war against Magnus Maximus himself or have someone in his retinue to do it in his stead (Ep. 867). Van Hoof and Van Nuffelen ascribe to Libanios an assumption of the existence of a ‘report’ of that war in line with their own category of a ‘report’ as a kind of official dispatch.^[55] However, Libanios’ statement, ἡ φωνὴ δὲ ἔστω σὴ τοῦ μηδὲν τῶν πεπραγμένων ἀγνοοῦντος μήτε μεῖζον μήτε ἔλαττον, “but let your voice be not one ignorant of any action, neither greater nor lesser,” should be rather interpreted as one indicating that he asks for composition of such a work rather than presupposes its existence. Though the letter expressing this request is not specific on the character of the work, the description of it as a favour (χάρις) not just to Libanios, but to the emperor Theodosius himself, implies something more than just an epistolary report – a literary piece that could be circulated further. That Libanios expected Promotus to be capable of composing this potential Ὑπόμνημα or have someone in his retinue who could do so demonstrate that military men were seen as potential authors, who possessed rudimentary skills for the task.

Attesting the existence of capable writers amongst the Later Roman military is another case – Victricius of Rouen. The bishop of Rouen’s reason for fame among modern scholars is his only surviving work, a sermon De laude sanctorum. Composed for the occasion of introducing to Rouen relics given to the bishop by Ambrose of Milan, this text displayed familiarity with such works as Cicero’s Pro Caelio and Porphyry’s Εἰσαγωγή through its allusions and references, as well as standard philosophical arguments betraying the rhetorical education of its author.^[56] As reported by Paulinus of Nola, prior to his clerical career Victricius served as a soldier, likely during the reign of the emperor Julian (Ep. 18.7). If Paulinus’ account is to be trusted, Victricius was a military man of low enough rank to answer to a tribune. However, given that bishop of Nola aimed to pen a laudatory confessor-story about his addressee that, according to his own admission, he barely knew, the accuracy of his description is somewhat questionable. Regardless of that, Victricius is evidence of an (ex)military man of a rhetorical education and literary skill. Such individuals were certainly not unheard of in the following decades.

Flavius Merobaudes, originating from a family with military traditions, combined his own military career at the side of Late Roman ‘generalissimo’ Aetius with composing encomiac poetry for the benefit of his patron and commander (PLRE 2:756–758). The inscription accompanying a statue erected in his honour by the emperor and senate in the Forum of Trajan praised him both for his strength and learning, focusing on this combination of literary and military skills (CIL VI 1724). He was not an isolated case even among Aetius’ retinue. Quintianus, a soldier and a poet of a Ligurian origin, accompanied Aetius in his Gallic campaigns and wrote poetry praising him as well. An anonymous author from a similar milieu wrote compositions of a seemingly comparable type for the rival of Aetius – Bonifatius.^[57] All three were celebrated and praised by Sidonius Apollinaris in one of his poems (Carm. XI 277–301). Seemingly, they were perceived as part of the same phenomenon already in the subsequent generation. Although their literary production had little in common with classically understood historiography beyond the subject of military conflicts and actors involved in those, yet it again illustrates the presence of capable literati among the Late Roman military men and their inclination to repurpose traditional literary genres for their group’s needs and sensibilities.

This short survey demonstrates that the Roman military of the late fourth and early fifth centuries did not lack for men capable of producing literary compositions, which would imply that they could find at least a portion of their audience among the active and former military men as well. Sulpicius Alexander’s prose strongly indicates that he was such an author, demonstrating knowledge and sensibility that would resonate with other soldiers. Does this evidence support or rule out Seeck and Zecchini’s hypothesis on identifying Sulpicius Alexander with the addressee of Symmachus – tribunus et notarius Alexander?

VI Sulpicius Alexander – tribunus et notarius of Magnus Maximus?

The point of concern Zecchini focused on with his identification is the critical view of Magnus Maximus and some of his commanders. The hostility towards the usurper echoing the Theodosian propaganda, as pointed out by Humphries, is at first glimpse somewhat problematic given tribunus et notarius Alexander’s involvement with Magnus Maximus’ regime. However, Zecchini correctly argues that disavowal of the usurper would be entirely fitting for a former official who tried to overcompensate by exhibiting loyalty to the winning side of the civil war.^[58] The already discussed case of the shifting attitude in Sulpicius Severus’ writings is perhaps a better illustration of concealed views of Maximian’s supporters in the aftermath of Theodosius’ victory. In his Vita Martini composed in 397, Sulpicius Severus presented Magnus Maximus as a usurper, who had a corrupting influence on bishops, only to adopt a more neutral stance in his Chronicle and eventually a eulogistic and apologetic one in his Dialogues.^[59] This likely reflected the real widespread attitudes of Gallic elites and officials towards the emperor, whose base of support was in Gaul. The hostile tone of Sulpicius Alexander’s Historia and references to the imperial propaganda would be entirely suitable if he had been part of Magnus Maximus’ regime and sought to distance himself from that usurpation in the eyes of Theodosius’ establishment. The vivid criticism of Quintinus’ arrogance and incompetence, while presenting Nannienus as the more prudent of the magistri militum, implies some familiarity with military officials in Magnus Maximus’ service and his own personal views of them, which would fit well with former involvement with that regime (Sulpicius Alexander, Historia F1).

What is, however, problematic about this identification in light of the military background argued here is the civil character of the office of tribunus et notarius.^[60] As high-ranking members of schola notariorum, tribuni et notarii had little to do with military matters, as they were usually sent on diplomatic missions next to their administrative duties. The only attested cases of involvement of tribuni et notarii with the military in the fourth century are instances of overseeing redeployment of troops, distribution of payment or inspecting Roman commanders’ conduct.^[61] Although such an official could potentially have an opinion on Nannienus and Quintinus’ handling of the Rhine campaign and access to the information on it either through official channels or through his own network, his interest in tribune Heraclius’ death in a history penned by him would be puzzling. Another issue is that the Alexander known from Symmachus’ letter is said to have been in service earlier to Valentinian II, which means that just like Symmachus, who probably knew him from his time as praefectus urbi, he only became part of Magnus Maximus’ usurpation after the latter’s invasion of Italy in 386/7 (Symm. Ep. 5.39). He would be part of the Italian administration and likely have little to do with the Rhine frontier that Sulpicius Alexander was so clearly familiar with. The identification suggested by Seeck and upheld by Zecchini solely based on the matching cognomen is therefore impossible set next to the argument presented here.

VII The phenomenon of soldier historians

Considering all the evidence, it would seem that Sulpicius Alexander was a former soldier, possibly aligned in some form with the regime of Magnus Maximus, likely originating from the provincial, curial elites of Gaul (see the Appendix). If this hypothesis about Sulpicius Alexander’s military background and the military portion of the audience his work was meant for is correct, then there are at least three Roman historians of military background active in the second half of the fourth century, two of them composing in Latin in the final decades of that century. Such a concentration of a literary activity, historical works sharing otherwise unique characteristics, conducted by authors of seemingly similar professional experiences and social backgrounds, should be considered a cultural phenomenon emerging from the social realities of the Later Roman Empire. These were elite professional military men, who were educated well enough to participate in and appreciate the literary culture of the traditional Roman elite yet constituted now a group clearly defined by their career limits. Their interests in other individuals of military profession, including military tribunes, specific units and so on were provided for and reflected in history works of authors sharing their sensibilities and social background.

At the same time, writing a historical work that had any chance of becoming appreciated by Late Antique audiences enough to go beyond the initial circulation and survive until today, at least in a fragmentary form or as a mention, required a level of Classical education not necessarily possessed by most Roman military men. In the three discussed cases, there are indications of such an intersection of a military career and an earlier social background that would usually go hand in hand with an education needed for composing works of history. The fact that Magnus had been known by his city of origin rather than a rank or any other moniker may perhaps imply a civic background and self-identification with its characteristic of belonging to the curial class, which would correspond well with the literary skill and familiarity with Classical historiographic model displayed by him. The archaising style found in the fragments of his work suggests a relatively high level of education as well.^[62] The Antiochean origin of Ammianus, alleged on the belief he was the addressee of Libanios’ letter, is currently in doubt. However, his curial background remains the most plausible of his proposed origins.^[63] The seeming prevalence of the curial class amongst the bearers of the nomen gentilicium Sulpicius in the third and fourth centuries may in fact indicate a similar civic background of Sulpicius Alexander. These authors and their works were very much the product of the socio-cultural conditions unique to the fourth-century Roman Empire.

However, acknowledging the existence of this phenomenon leads us to another question – why there is no trace of later authors and works of that kind? The disappearance of this burgeoning strain of narrative political histories created by soldiers and, at least partially, for other soldiers in the Latin West, would be dependent on the same factors that made its development possible in the first place. The regionalisation of the Imperial professionalised military class, a social group that provided this subgenre of historiography with both authors and an important portion of its audience, decreased incentives for composing histories of that type such as advancing one’s career and social standing within the hierarchy of the Empire. The same process lowered the chances of circulation of such works beyond the network of immediate recipients. The subsequent demise of the imperial military in the West in the late fifth century meant the death of this subgenre of political history before it fully developed and left a lasting mark on the Latin historiography. It appears Sulpicius Alexander was at the same time one of the first authors of his kind and one of the last.

Acknowledgement: I owe a debt of gratitude for all the helpful suggestions and advice I received regarding this article to Prof. Paweł Janiszewski, Dr. George Woudhuysen, Dr. Marzia Fiorentini, and last but not least, Dr. Karolina Frank. I would also like to thank the audiences of Ewa Wipszycka’s Late Antique Seminar, The Lost Latin Historiography of Late Antiquity Session of IMC Leeds 2021 and Nottingham Classics and Archaeology Research Workshop for all the feedback they gave me on the occasion of delivering earlier versions of the argument presented in this article. The research for this article was conducted as part of the project, The Missing Link. The Lost Latin Historiography of the Later Roman Empire (3rd–5th century) 2018/29/B/HS3/00796 funded by the National Science Centre, Poland, through the grant scheme OPUS 15.