Abstract and Keywords
The Epilogue offers a summation of the main arguments of the book; namely that Ammianus Marcellinus casts the major episodes of Julian’s reign as topoi, which are familiar from classicizing historiography in Latin, and that in doing so he deliberately presents a more ‘Romanized’ account of Julian for a Western, Latin-speaking audience. The Epilogue also compares the various points in the narrative where the narrator appears as a participant, and suggests that these are all conditioned by Ammianus’ wish to use his ‘presence’ as a way both of critiquing Julian, and of creating authority for his version of Julian’s reign.
One of the narrator’s last direct comments on Julian relates to his place of burial in Tarsus (25.10.5):
But his remains and ashes, if anyone then showed sound judgement, ought not to be looked on by the Cydnus, although it is a beautiful and clear stream, but to perpetuate the glory of his noble deeds they should be laved by the Tiber, which cuts through the eternal city and flows by (praeterlambere) the memorials of the deified emperors of old.
Ammianus’ statement forcefully corrects Libanius’ similar claim that Julian should have been buried not in Tarsus but at Athens ‘in the Academy, next to Plato’s tomb’ (Or. 18.306).1 Nonetheless, Ammianus’ Latin phrase also alludes to Virgil’s evocation of the burial place of the first Roman emperors: ‘what funerals you will see, Tiberinus, as you flow past (praeterlabere)2 the new tomb!’ (Aen. 6.873–4).3 Ammianus’ dual allusive practice here neatly reflects his wider creation of a western, Latin Julian in opposition to the eastern Julian of the existing tradition.
By setting a portrayal of Julian within the revived genre of historiography in Latin, Ammianus was able to give attention to many aspects of Julian’s reign that had not received detailed elaboration in the existing discourse on Julian up to the 380s, particularly Julian’s elevation in 355 and the pre-emptive description of Silvanus. However, he also created a new narrative of several events that had already (p.204) received extensive discussion in the quarter century after Julian’s death (Strasbourg and Persia). In each case, Ammianus’ revival of historiography, particularly Latin historiography, granted the opportunity to create a definitive, conclusive Julian-narrative which stood noticeably detached from the majority of the earlier accounts that had been written in the East for easterners and mostly in Greek. The identity of the historian as a Greek writing in Latin offers considerable authority for this enterprise, but the definitiveness of Ammianus’ version gains additional authority by its shaping of contemporary history into large-scale, stock generic topoi that had previously been developed by the likes of Sallust and Tacitus (tyrannical representation, speech scenes, imperial adoptions, battle scenes, and the use of omens). The combination of composition in Latin and the promotion of Rome presupposes the ancient capital as the place of writing, and the city’s inhabitants if not the sole audience, an important and unavoidable one. Ammianus’ is a western Julian, recounted by an authoritative Greek no doubt, but presented in a very Roman context of Latin historiography.
Although the deployment of topoi is a recurring phenomenon throughout Ammianus’ narrative of Julian, it does allow for the development of Julian as a character. Julian progresses through the stages of the unprepared but divinely endorsed student, the increasingly competent military leader, to a high point as an ideal emperor at his acclamation who is finally aware of his divine endorsement, and ultimately to a more complex figure encountering failure in Persia. This development is also marked by the primary narrator’s control of his principal subject, especially the degree to which Julian’s interpretation of events is allowed to cohere with the (authoritative) presentation by the primary, external narrator. In terms of recent discussions of religion within the Res Gestae, I adhere to the view that Ammianus not only denigrates Christianity but also creates a metaphysical system that owes more to paganism and traditional philosophy. To what degree this framework, as revealed by Rike, Davies, and Brodka, is fully consistent is still a matter of debate,4 but I hope to have shown that forces such as fatum, fortuna, and the numen play an important narratological role, especially as regards which characters may perceive and interpret these forces correctly.
(p.205) More than just an external narrator, Ammianus is of course one of a select group of historians who is also a participant within the narrative world that he creates. As Kelly has shown, many of his exploits are described in such a way that the loneliness and independence of the participant is designed to reflect back onto the narratorial persona and thus act as a further corroboration of Ammianus’ authority to write his history. In Persia, the appearance of participant and of first-person narrative also acts as a point of aemulatio with existing, eastern accounts of the Persian campaign. Earlier, in the accounts of Gallus and Silvanus, the participant’s appearance fulfilled a structural function in precipitating a change in the narrator’s presentation of each of these imperial characters. Just as Gallus and Silvanus act as preparatory studies for Julian, so too can the role of the participant in each of the Gallus and Silvanus episodes be seen as preparation for the appearance of the first-person during the Persian campaign. Ammianus had been careful to remain distant from the figure of Julian. Julian is kept away from the Silvanus episode (Chapter 2) whereas Ammianus keeps himself away from Julian in Gaul. It cannot be coincidence that the most extensive and the most ‘personal’ of Ammianus’ appearances (at Amida) occurs on the other side of the empire from Julian, who was at that point in Gaul, and in a scene that is designed to provide criticism not of Julian but of Constantius.5 So, in much broader terms, Ammianus’ appearance at 23.5 at the crucial moment when Julian delivers a speech that reveals his flawed reading of the military situation and amidst a situation where the discourse on omens still remains unresolved, may suggest that this should be read as the ultimate moment of peripeteia in Julian’s career.
I am aware that my reading of Ammianus raises several wider possibilities for investigation. In this book I have been mostly concerned with the military episodes in Julian’s career (largely because they were the ones which had received most discussion by Latin authors prior to Julian, as outlined in the Introduction), but a comparison with Julian’s ‘civil’ roles in both Constantinople and Antioch would also be instructive. We know that Ammianus avoided discussion of the New Rome wherever possible,6 but the primary narrator is forced to use Constantinople as the setting for his narrative in the first (p.206) half of Book 22 during Julian’s stay in 361. It would seem the ideal opportunity to exhibit pro-Roman bias and rivalry with the ancient capital’s younger competitor. Looking beyond the narrative of Books 14–25, one could ask whether the interaction between Ammianus’ narrator and the character Julian is distinct from that with other emperors in the Res Gestae.
In the conclusion to his book, Kelly posed a question: ‘if [Ammianus’] autobiographical passages focus on the formation of a historiographical persona…is he comparable to those historians writing in Greek, such as Olympiodorus, Priscus and Procopius, who are equally prominent in their own works?’7 The same question could be posed for Ammianus’ structural use of the participant. The almost wholesale absence of narratological studies on works of Latin prose of all periods and of late antique texts both Latin and Greek makes it difficult to discern whether Ammianus’ narratorial technique is similar to that of other historians (or indeed to writers in other genres, particularly the novel8); and whether in this respect, Ammianus should be considered more Classical or Late Antique in his aesthetic. Nonetheless, I hope some of the features of Ammianus’ narrator that I have identified here will offer a comparison for future studies of Sallust as much as Procopius.
It is within the continuum of Roman historiography, as represented in its largely extant form by the termini of Sallust and Procopius, that we should be prepared to read Ammianus as a skilled literary practitioner, who is willing to create a thrilling and compelling narrative under the constant and pervasive control of a subtle narrator, and who engages with and adapts the models set by his predecessors to shape an authoritative view of his recent history.
(2) Praeterlambere in Ammianus is Valesius’ correction of V, but den Boeft et al. (2005, 320–1)’s preference for Gelenius’ praeterlaberetur strengthens further the identification of the allusion to Virgil.
(8) A genre that Ammianus seems to have been fully aware of (see Weisweiler 2015, 126–9, who points out that the story of Craugasius and his wife in Books 18 and 19 exhibits many of the topoi expected of the plot of an ancient novel. These observations may suggest that Ammianus also owes more directly to Heliodorus in the narrative of the siege of Amida, which is ‘framed’ by the Craugasius-narrative, than I allowed in Ross 2015c).