Futurum Praeteritum: Narrative Prolepsis in Ammianus (and Tacitus)
p. 421-433
Texte intégral
1Book 21 of Ammianus Marcellinus’ Res Gestae is principally concerned with Julian’s rise to supreme power (in the preceding book he was proclaimed Augustus by his soldiers): in chapter 1 he is in Vienne, celebrating the fifth anniversary of his elevation to the position of Caesar; in chapter 5, still in Vienne, he addresses his soldiers, revealing to them his decision to wage war on Constantius; in chapter 10 he is well advanced in his swift movement eastwards to confront the old emperor (he stops for a few days in Sirmium in western Illyricum); in chapter 15, when Constantius dies in Cilicia while marching to fight with his rival, Julian becomes the sole and uncontested ruler of the Roman empire. Taking into account the importance of Julian for Ammianus, both as his model emperor and as the major character of his narrative, the place of Book 21 within the framework of the Res Gestae is highly significant1. It is worth pointing out that Ammianus underscores the significance of the book by means of numerous references (more or less straightforward) to the future – the future, that is, from the point of view of historical figures within the narrative (but, from the point of view of the author and his audience, the events thus evoked belong, of course, to the past; hence futurum praeteritum in the title of my chapter). In what follows, I shall discuss Ammianus’ use of such references and the roles played by them in his work; also, I shall try to compare his techniques of indicating or suggesting future events with those used by Tacitus – an author often regarded as one of Ammianus’ most important historiographical models. Following G. Genette and other scholars of narratology, I call a reference to an event which, at the point of its being mentioned in the narrative, has not yet happened, (narrative) prolepsis, a term already well-domesticated, also in the field of classical studies2.
2It would be advisable to start our discussion by looking more closely at Book 21. Ammianus begins his narrative with Julian pondering over whether or not he should remain loyal to Constantius; he finally comes to the conclusion that to openly oppose the emperor would be the safest option. Although the section as a whole (21.1.1-6) is directed to the future, it is above all its last paragraph which is explicitly proleptical (21.1.6):
Acuebat autem incendebatque eius cupiditatem pacatis iam Galliis incessere ultro Constantium coniciens eum per uaticinandi praesagia multa, quae callebat, et somnia e uita protinus excessurum3.
3Thus in the first chapter Ammianus already looks forward to what he will narrate, suo tempore, in this book’s last chapter but one (the last chapter is mainly devoted to Constantius’ obituary, so the emperor’s death dominates the closing of Book 21). The time-reach of the prolepsis is rather short – some 12 months between Julian’s quinquiennalia in Vienne (early November 360) and Constantius’ death on the 3rd of November 361.
4One of the most obvious distinctions between various forms of prolepses is that between proleptical statements made by the narrator and those made by a character within the narrative, in other words, between narratorial and actorial prolepses4. But in this case the distinction is somewhat blurred since it is the narrator who makes the statement, but, at the same time, the reference is to his character’s (Julian’s) own conjectures, based on Julian’s interpretation of predictory signs and dreams5. Usually (but not unexceptionally, especially in modern fiction) the narrator’s prolepses are more authoritative: if the narrator anticipates future events, we may safely assume that such an announcement will come true. Sometimes actorial prolepses are coupled with narratorial ones – in this way the author either confirms a prolepsis made by his or her character (when they converge) or, on the contrary, undermines it (when they clash)6. To come back to our passage: it is obvious that Ammianus’ Roman audience did know, thanks to their earlier acquaintance with contemporary history, that the statement about Constantius’ impending death was true; Ammianus, nevertheless, has chosen to suggest, in two different ways, that Julian’s interpretation of predictions was correct. Firstly, he inserts a short subordinate clause quae callebat which defines uaticinandi praesagia, thus emphasising Julian’s expert knowledge of predictory signs. Secondly, he introduces a rather long digression discussing various methods of predicting the future (21.1.7-14); this section’s most important message is that learned people (including Julian) are in fact able to make reliable statements about events to come.
5The digression is followed by two episodes which are also of a proleptical character (21.2.1 sq.). The first may be classified as an omen7: during his military exercises, Julian’s shield suddenly falls apart and he is left with its handle only; while those standing nearby are terrified, Julian asserts that they should not be frightened: habeo firmiter, quod tenebam. The second episode is more explicit: imago quaedam splendidior appears to Julian in the middle of the night and recites, again and again, four hexameters in Greek which predict, in a very unambiguous manner, the death of Constantius8.
6When we come, towards the end of Book 21, to the last days of Constantius, proleptical statements also abound. Ammianus sets the tone at the beginning of chapter 14 (21.14.1):
In hoc rerum aduersarum tumultu haerens eius fortuna iam et subsistens aduentare casum uitae difficilem modo non loquentibus signis aperte monstrabat.
7There follow some examples of these signa; at the end of this section Ammianus mentions Constantius’ conviction that his tutelary genius has deserted him (21.14.2):
…et putabatur Genius esse quidam tutelae salutis appositus eum reliquisse mundo citius digressurum.
8The mention of the tutelary genius leads the historian to present, in the form of a digression, some views on this topic held by Greek poets and philosophers (21.14.3-5). The emperor’s death itself is narrated in the next chapter, but also here there is one more prolepsis looking forward to it: when the emperor is passing through a suburban settlement of Antiochia he notices a human corpse, headless, pointing towards the west: territus omine finem parantibus fatis destinatius ipse tendebat (21.15.2).
9We should note a certain structural symmetry which governs Book 21. Both at the beginning of the book and towards its end we have the following sequence: Proleptical Section (1) – Digression – Proleptical Section (2). In both cases the prolepses point to the death of Constantius, but in chapters 1 and 2 they are recounted from the point of view of Julian, whereas in chapters 14 and 15 Constantius dominates the narrative. Moreover, it seems likely that the phrase citius digressurum (placed immediately before the excursus, 21.14.2) is intended to recall protinus excessurum of 21.1.6, also the last words of the first proleptical section. Furthermore, we are probably invited by “the allusive historian”9 to connect territisque ut omine diro praesentibus cunctis (21.2.1, the first proleptical item after the digression) with territusque omine (21.15.2, also after the digression); the combination of omen and terrere is found in Ammianus, apart from in these two passages, only at 23.1.5 (terrebatur omine quodam). Julian, when faced with what may be deemed as an omen dirum, remains, in contrast to his companions, unterrified; for Constantius, on the other hand, the omen of the headless corpse is but the last straw in his growing terror as he becomes convinced that his death is very close indeed10.
10Two more observations may be made in connection with Book 21. Firstly, let us return to the passage quoted above about Constantius being deserted by his tutelary genius. Thus far, we have concentrated on differences between the presentation of Constantius and Julian in the proleptical sections; but here there is a curious connection between the two rulers – which, however, reveals itself only in Book 25. The Constantius passage (21.14.1) bears a marked similarity to the account of Julian’s nocturnal vision immediately before the battle with the Persians which will bring him death (25.2.3):
uidit squalidius, ut confessus est proximis, speciem illam Genii publici, quam, cum ad Augustum surgeret culmen, conspexit in Galliis, uelata cum capite cornucopia per aulaea tristius discedentem.
11Ammianus does not resort here to his favourite technique of cross-reference, but he undoubtedly invites his readers to look back for the account of an earlier apparition of this genius, on the night before Julian was proclaimed Augustus by his troops (20.5.10). Significantly, in the earlier passage the historian quotes the genius’ address to Julian, which ends thus: id tamen retineto imo corde, quod tecum non diutius habitabo. By means of this proleptical statement (habitabo is the last word spoken by the genius!), Ammianus connects the beginning and the end of Julianus Augustus11. To put it more clearly: the prolepsis at 20.5.10 looks forward to the analepsis (flash-back)12 at 25.2.3; the analepsis at 25.2.3 looks back to the prolepsis at 20.5.10. This is a technique much resembling that used by Vergil in the two arguably most significant passages of the Aeneid: Turnus’ slaying of Pallas and despoiling him of his belt (the prolepsis Turno tempus erit… at 10.503-505) and Aeneas’ giving up his intention of saving Turnus upon catching sight of Pallas’ belt on his shoulder (the analepsis at 12.943 sq.)13. On a smaller scale, but much closer in content to Ammianus than the Vergil passages, we may compare two chapters on Fortuna in Suetonius’ biography of Galba (Galb., 3 and 18): her first appearance predicts Galba’s rise to power and the second announces his fall14.
12My second observation concerns a prolepsis which refers to neither Constantius nor Julian – but which also serves to connect the events of 360 and 361 with the end of Julian’s rule. It comes at the very end of Book 21. The long obituary of Constantius is followed by a notice about the procession with the emperor’s embalmed body heading for Constantinople. Jovian, at that time an officer of the bodyguard (etiamtum protector domesticus), was in charge of this procession. Ammianus pays attention to some apparently quite unimportant things which happened at the beginning of the journey (21.16.21):
Eique [scil. Iouiano] uehiculo insidenti, quod portabat reliquias, ut principibus solet, annonae militaris offerebantur indicia […] et animalia publica monstrabantur et ex usu crebrescebat occursus. Quae et alia horum similia eidem Iouiano imperium quidem, sed et cassum et umbratile ut ministro rerum funebrium portendebant.
13That was in November 361; one and a half years later, on the 27th of June 363, Jovian became, quite unexpectedly, Julian’s successor as the Roman emperor (see 25.5 for the circumstances of his acclamation) – and just eight months after receiving the purple, he met his end (hence imperium […] cassum et umbratile). Thus the trivial matters recorded by the historian at 21.16.21 reveal their true significance – and Ammianus places them at the very end of the book, which was dominated by Jovian’s two predecessors, for dramatic effect. It is by no means accidental that portendebant is the book’s last word15.
14The historian uses the same technique in Book 17. In its last chapter he deals with the fiasco of peace talks between Rome and Persia; the Roman delegation, having realised that king Sapor has no intention of making any concessions, returns to Constantius. However, a new delegation is soon dispatched, consisting of two negotiators: Lucillianus missus est comes et Procopius tunc notarius, qui postea nodo quodam uiolentiae necessitatis astrictus ad res consurrexerit nouas (17.14.3, the last words of the book). The reference is to the usurpation of Procopius in 365/366, described in detail in Book 26; thus the time-reach of this prolepsis is six years and it spans almost one-third of the Res Gestae (ten books). As is the case with Jovian (see n. 15), Procopius has also not been mentioned before; in Book 25 there is another prolepsis looking forward to his usurpation16.
15Proleptical statements put at the very end (of a book, a major narrative segment or a narrative year) are common in Tacitus17 and it is tempting to suppose that Ammianus’ technique of anticipating later developments owes something to that historian. The question of Tacitus’ influence on Ammianus is much discussed, and no scholarly consensus has been reached so far18. On the one hand it is obviously relevant that Ammianus began his narrative at the point at which Tacitus finished his own history (Amm. 31.16.9: a principatu Caesaris Neruae exorsus); on a stylistic level, there are several phrases in the Res Gestae which are most probably modelled on Tacitus19; there is also some similarity of historical views, interests and attitudes20. But, on the other hand, nowhere in his work does Ammianus mention Tacitus by name (although he cites many Greek and Roman writers, among them some historians, e.g. Thucydides and Sallust)21; more importantly, many of the two authors’ literary and historical methods differ so markedly that it would be highly injudicious to posit that Ammianus’ intention was to imitate Tacitus on a large scale and in all respects. Nevertheless, it is at least possible that the later historian’s use of narrative prolepsis is somewhat related to his predecessor’s techniques. Thus, in what follows, I would like to make some observations on how Tacitus and Ammianus make use of this narrative device, on what is common to them and how they differ. These observations will be of a preliminary character only; a full comparison of the two authors’ treatment of prolepsis is beyond the scope of the present study22.
16One of the most remarkable instances of Tacitus’ use of prolepsis is found in Book 4 of the Annals, in the context of the trial and execution of Titius Sabinus (this is the first item of the annalistic year AD 28). Tacitus the narrator restrains himself, as it were, from at once revealing, in defiance of the annalistic format of his work, the fate of those who brought about Sabinus’ fall (Ann., 4.71.1):
Ni mihi destinatum foret suum quaeque in annum referre, auebat animus antire statimque memorare exitus, quos Latinius atque Opsius ceterique flagitii eius repertores habuere, non modo postquam C. Caesar rerum potitus est, sed incolumi Tiberio, qui scelerum ministros ut peruerti ab aliis nolebat, ita plerumque satiatus et oblatis in eandem operam recentibus ueteres et praegraues adflixit. Verum has atque alias sontium poenas in tempore trademus.
17Indeed, at 6.4.1 (AD 32) the historian informs us about the punishment of the first of Sabinus’ accusers, Latinius (or, correctly, Lucanius) Latiaris: Latiaris, ut rettuli23, praecipuus olim circumueniendi Titii Sabini et tunc luendae poenae primus fuit. The books covering the principate of Gaius are lost; we should assume that the punishments of other accusers were recounted there. Of course, it would be unwise to take Tacitus’ statement about his narratorial self-restraint at face value. From the rhetorical point of view, the passage from Book 4 is a fine example of praeteritio: while pretending that the annalistic framework of his history prevents him from mentioning sontium poenas here and now, in actual fact the historian does mention them, and even puts them in high relief. As J. Ginsburg has shown, there is no question of Tacitus being constrained by the annalistic format which he inherited from his republican predecessors; on the contrary, he is capable of using it to his own ends to press home his own views on the history of imperial Rome24.
18I have selected the Sabinus passage as the starting-point for my comparative analysis because Ammianus in fact provides an interesting parallel. In the first chapter of Book 28 he gives a graphic description of the oppression in the city of Rome under the uicarius Maximinus (370/371) and his successors Simplicius (374/375) and Doryphorianus (375/376). Many Roman aristocrats were tortured, executed or exiled. The long chapter is summed up by the following epilogue (28.1.57):
Sed uigilarunt ultimae dirae caesorum. Namque ut postea tempestiue dicetur, et idem Maximinus sub Gratiano intoleranter se efferens damnatorio iugulatus est ferro et Simplicius in Illyrico truncatus et Doryphorianum pronuntiatum capitis reum trusumque in carcerem Tullianum matris consilio princeps exinde rapuit reuersumque ad lares per cruciatus oppressit immensos.
19Ammianus’ tempestiue is the exact counterpart of Tacitus’ in tempore. But the later historian does not fulfil his promise; the punishment of the three men is never described in the subsequent narrative25. This passage is a good illustration of how markedly Ammianus’ literary technique differs from that of Tacitus. For one thing, he is much more explicit in emphasising the motif of “just retribution”. Tacitus would never have written such a sentence as sed uigilarunt ultimae dirae caesorum26. For another, and more relevantly to our subject, Ammianus is much more specific than his predecessor about the gloomy fates of the three uicarii; the reader, in fact, may come to the conclusion that there is no longer any need to describe them tempestiue, in the following narrative27.
20We have discussed above Ammianus’ subtle introduction of the future emperor Jovian at the end of Book 21. His introduction of another prospective emperor, Valentinian I, is more straightforward. At 16.11.6 he mentions two officers taking part in a military action in Gaul: Bainobaudes tribunus et Valentinianus postea imperator. The year is 357, so Valentinian’s elevation is seven years ahead; it will be narrated in Book 26 (the prolepsis spans eleven books of the Res Gestae). The technique of noticing a historical figure’s later career, especially upon his first mention in the narrative, is common in Ammianus28 – and is by no means limited to future emperors. To give only a few examples:
– 17.6.3: huic pugnae Neuitta, postea consul, equestris praepositus turmae, et affuisse et fortiter fecisse firmatur [358; he became consul in 362; here Ammianus draws attention to Nevitta by placing this item at the end of the narrative of the battle with the Juthungi, and by the final alliteration].
– 21.10.6: Victorem […] uirum sobrietatis gratia aemulandum, multo post urbi praefectum [361; this is Aurelius Victor the historian; he became prefect of Rome in 38929].
– 22.1.2: Arpunculus Gallus orator promotus rector postea Narbonensis.
– 31.3.5: Munderichum ducem postea limitis per Arabiam30.
21It would be difficult to find exact parallels of this kind of proleptical notice in Tacitus. When he refers to a historical figure’s later career, he does so less straightforwardly and usually to make a point. The most interesting example of this is Ann., 3.18.4 on Claudius, a passage to which we shall return. On a surface level, Tacitus’ flash-forward reference to the rule of Gaius Caligula closely resembles Ammianus’ later career notices. The reference comes in the context of Livia’s death and funeral in AD 29. The historian rounds off31 the section devoted to Livia with this sentence: laudata est pro rostris a C. Caesare pronepote, qui mox rerum potitus est (Ann., 5.1.4; here mox corresponds to Ammianus’ postea). It may be asked why Tacitus thought his readers had to be reminded about Caligula’s elevation to the throne (in 37, eight years later). This was common knowledge. But the wider context of this passage is crucial, namely Sejanus’ influence on Tiberius and his actions against Agrippina and her children. Sejanus (this is the message Tacitus delivers by means of the prolepsis) will not be wholly successful; one of Agrippina’s sons will survive – and will become Tiberius’ successor32. Here, at least, we may detect some resemblance to one of Ammianus’ notices. At 15.2.7 sq. he writes about accusations brought against Julian; the situation was dangerous but, happily, the empress Eusebia intervened in his favour, aspiratione superni numinis, as Ammianus believes. Significantly, at the beginning of this section Julian is described as memorabilem postea principem – most probably to show that the machinations of evil people at the court of Constantius will come to nothing.
22If people other than emperors are introduced by Tacitus, he occasionally hints at their later historical significance (but not at their formal career). Such is the case with Nymphidius Sabinus (Ann., 15.72.2); here, the historian combines analepsis (a short digression about Nymphidius’ earlier life: pauca repetam) with prolepsis, namely a rather unspecific statement nam et ipse pars Romanarum cladium erit. Another example is his first mention of L. Vitellius (the emperor’s father) at Ann., 6.32.4; Tacitus refers to his later infamous role as Claudius’ servile courtier, but his point is a paradoxical one: that this man proved himself, nevertheless, an excellent provincial commander (an aspect of his biography which faded into oblivion: bona iuuentae senectus flagitiosa obliterauit)33. There is also a particularly intriguing reference to Cassius Chaerea’s role in the assassination of Caligula: Tacitus registers his brave action in AD 14 during the mutiny of the Rhine legions (Cassius Chaerea, mox caede C. Caesaris memoriam apud posteros adeptus, tum adulescens et animi ferox, inter obstantes et armatos ferro uiam patefecit). Here, it is crucial to realise that Gaius (then a two-year-old) was also present in the legionary camp (cf. 1.40.4; 1.41.2). Chaerea’s and Gaius’ ways crossed for the first time in AD 14; tum in this passage looks ominously forward to mox (when they will cross for the last time). All these examples show that here we are far away from Ammianus’ matter-of-fact career notices.
23There is more similarity between Ammianus and Tacitus in their treatment of what may be called proleptic epilogues. These are notices about a character’s later life which are inserted at the point where this character is mentioned for the last time in the narrative; the implication is that he (or she) will not play an important historical role in later events, but the reader may be interested in his (or her) personal fate. A good example is Amm. 15.7.5, a notice about what happened with the troublemaker Petrus Valvomeres after he had been arrested by the urban prefect Leontius:
…uulgus omne […] ita euanuit, ut turbarum acerrimus concitor34 tamquam in iudiciali secreto exaratis lateribus ad Picenum eiceretur, ubi postea ausus eripere uirginis non obscurae pudorem Patruini consularis sententia supplicio est capitali addictus.
24In Book 16, Ammianus speaks about Chnodomarus, one of the kings of the Alamanni; as a result of Julian’s victory in the battle of Argentoratum, he was captured and sent to Julian. The historian appends a short epilogue (16.12.66):
Et diebus postea paucis ductus ad comitatum imperatoris missusque exinde Romam in castris peregrinis, quae in monte sunt Caelio, morbo ueterni consumptus est.
25In both passages the story is brought to its obvious conclusion, namely to the character’s death. In the case of Valvomeres, Ammianus’ implied message is that this turbarum concitor remained troublesome even in exile; his later life confirms the rightfulness of the action taken against him by Leontius. In the case of Chnodomarus, the situation is different: what is important here is the contrast between his previous haughtiness and excessive confidence (e.g. 16.12.4; 24; 61) and his present condition as a Roman captive35. Citing Ammianus as his source, Gibbon writes that “Chnodomar experienced an honourable treatment: but the impatient Barbarian could not long survive his defeat, his confinement, and his exile”36. How did Gibbon know that Chnodomarus had died soon after his capture? He seems to have taken a wrong clue from the proleptic epilogue in Ammianus, believing that it must refer to the near future. But the Roman historian does not specify the time-reach of this prolepsis, and – since we do not have any other evidence bearing on Chnodomarus’ life after 357 – we are not in a position to determine the date of his death in Rome. What is more, Ammianus’ morbo ueterni may suggest that he died many years later, as an old man. But, more cautiously, we should say that he died somewhere between 357 and the late 380s, when Ammianus was composing his work37.
26Tacitus provides a good parallel to this passage. Maroboduus, the former king of the Suebi, is now expelled from his land and writes to Tiberius, seeking refuge on Roman territory. Tiberius agrees and Maroboduus is interned in Ravenna. An epilogue follows (Ann., 2.63.4, AD 19):
Et Maroboduus quidem Rauennae habitus, si quando insolescerent Suebi, quasi redit<ur>us in regnum ostentabatur: sed non excessit Italia per duodeuiginti annos consenuitque multum imminuta claritate38 ob nimiam uiuendi cupidinem.
27Differently than Ammianus, Tacitus specifies that Maroboduus spent 18 years in internment (so he died in AD 36 or 37). Having been driven out of his kingdom, he ceased to be an important historical figure and, consequently, this is the last appearance of him as an active character (at 3.11.1 he is briefly mentioned in a retrospective notice)39.
28Apart from such epilogues, more extensive biographical sketches may also contain some proleptical elements. In Ammianus, particularly interesting is his digression on Eutherius, Julian’s praepositus cubiculi (this is followed by another digression, which is on honest eunuchs). The digression comes in the context of the narrative of 356 (Eutherius bravely defended his master at the court of Constantius). Ammianus presents the whole biography of Eutherius, and the section on his later life (16.7.6-8) is as detailed as that on his fate before 356 (16.7.5). Ammianus’ introductory words (16.7.4: res monuit super hoc eodem Eutherio pauca subserere) bear some similarity to Tacitus’ opening formula in his biographical sketch of Helvidius Priscus (Hist., 4.5.1): res poscere uidetur […] ut uitam studiaque eius, et quali fortuna sit usus, paucis repetam. Both sketches are encomiastic and draw attention to their respective subject’s education and moral standards. But the Tacitean passage is from beginning to end retrospective; apart from a brief and vague reference in the first sentence (quoniam iterum in mentionem incidimus uiri saepius memorandi – this causal clause comes where I put dots to mark the omission) there is no proleptical material. On the other hand, Tacitus does refer to the future in his sketch of Sejanus at the beginning of Book 4 of the Annals – but also in that passage he is far from being specific (as Ammianus is when he presents Eutherius)40. A character’s later life is also presented, and in less allusive language than in the case of Sejanus, in Tacitus’ digression on Curtius Rufus (Ann., 11.21) – but in that chapter the historian’s concern is to depict the fulfilment of a prophecy concerning Curtius’ African proconsulship41.
29One of more important literary functions of prolepsis is to emphasise the gap between what is known by characters in the text at a given point in the narrative (and this knowledge reveals itself also in their plans, expectations or fears) and the hind-sight knowledge of the narrator. The tension thus arising may have a powerful dramatic effect. Tacitus, for example, presents an emperor sacrificing to the gods who have already deserted him; and another one, instituting rites to the gods of the place of his victory – which will soon become the place of his decisive defeat42. It is also Tacitus who makes a bitter comment about what he perceives as Fortune’s sport with all things human. The context is the trial of Cn. Piso in AD 20, and the striking narratorial prolepsis refers to an event which is 21 years ahead (Ann., 3.19.3-4):
Addiderat Messalinus Tiberio et Augustae et Antoniae et Agrippinae Drusoque ob uindictam Germanici grates agendas omiseratque Claudii mentionem. Et Messalinum quidem L. Asprenas senatu coram percunctatus est an prudens praeterisset; ac tum demum nomen Claudii adscriptum est. Mihi, quanto plura recentium seu ueterum reuoluo, tanto magis ludibria rerum mortalium cunctis in negotiis obuersantur. Quippe fama spe ueneratione potius omnes destinabantur imperio quam quem futurum principem fortuna in occulto tenebat43.
30Finally, let us briefly consider a passage from Ammianus in which attention is also drawn to the gap between what was known at a given point in the narrative and what will become known in the future. The year is 376 and emperor Valens has just agreed to the Goths’ request to be settled on the other side of the Danube, on Roman territory. Ammianus stresses the role which the emperor’s eruditi adulatores played in arriving at this decision; they assured Valens of the advantages, in terms of military manpower and money, which this decision would bring to the Roman state. This actorial prolepsis is followed by a narratorial one (31.4.4-6):
Hacque spe mittuntur diuersi, qui cum uehiculis plebem tranferant truculentam. Et nauabatur opera diligens, ne qui Romanam rem euersurus relinqueretur uel quassatus morbo letali. […] Ita turbido instantium studio orbis Romani pernicies ducebatur.
31Elsewhere in Ammianus we find Roman officials who, having received orders from a higher authority, for some reason neglect to execute them – and bad consequences follow (e.g. 14.7.10; 15.5.8). Here, on the other hand, we have most dutiful officials who pay special attention to the very accurate implementation of their task. And the result of all this will be nothing less than orbis Romani pernicies44. The Romans, led by their emperor, rush headlong into disaster, so graphically described in the final chapters of the Res Gestae (the battle of Adrianopole, 378)45.
32To sum up: the result of the above analysis is mainly a negative one. To be sure, there are some similarities in Ammianus’ and Tacitus’ treatment of proleptical statements and it cannot be ruled out (in the same way as it cannot be proved) that Tacitus exerted some influence on Ammianus in this respect. But, since prolepsis is a popular narrative device in other ancient historians, both Greek and Roman46, it is quite possible that Ammianus’ sources of inspiration lay elsewhere. Significantly, even in similar contexts (Tac., Ann., 4.71.1 and Amm. 28.1.57), the two historians’ approach is different. Tacitus seems to be more nuanced and allusive; some of his most meaningful references to the future may be classified not as explicit prolepses in the strict sense of the word, but rather as what G. Gennette calls, following R. Debray, “amorces” (“baits”)47. Ammianus, on the other hand, is more straightforward – but it would be wrong to conclude that this was caused by some artistic deficiency. We simply have to admit that his historical and literary priorities were different, not inferior48.
Notes de bas de page
1 For the importance of Julian in the Res Gestae see Ross 2016. For Julian as Ammianus’ (almost) ideal ruler see e.g. Blockley 1975, 73-103; den Hengst 2010, 306-320.
2 See Genette 1972, 82; Rimmon-Kenan 2002, 48; Martinez & Scheffel 2007, 33. In the field of classical studies: de Jong 1987, 81-90; 2014, 78-87 (with useful further reading); Hornblower 1994b, 140; Rood 1998, 11; also chapters in the collective volumes by de Jong & Nünlist 2007 and Lianeri 2016 (the latter is concerned solely with historiography). For “pre-Genettean” approaches to prolepsis (or foreshadowing) in Greek and Latin literature see e.g. Duckworth 1933.
3 Ammianus is quoted throughout this chapter after the Teubner edition by W. Seyfarth (Leipzig, 1978).
4 See de Jong 2014, 82.
5 Thus it may be said that the prolepsis looking forward to Constantius’ death is focalised by Julian – although in this case we are dealing with so-called embedded focalisation. For focalisation see Genette 1972, 206-223; Bal 1997, 142-161; de Jong 2014, 47-72.
6 A famous example of the latter is Achilles’ proleptical prayer for Patroclus, immediately followed by the narrator’s statement that Zeus fulfilled only the first part of the prayer (Hom., Il., 16.246-252).
7 The word itself is used in reference to the onlookers’ reaction: territisque ut omine diro praesentibus cunctis. The whole episode bears some similarity to the portent of the innocuous flame over the head of Ascanius (Verg., Aen., 2.679-691): nos pauidi trepidare metu on the one hand and Anchises’ joyful reaction on the other.
8 This prediction is quoted also by Zosimus (3.9.6) and Zonaras (13.11 C).
9 For the allusive Ammianus see Kelly 2008. He deals primarily with Ammianus’ intertextual references to earlier literature, but his observations also apply for intratextual allusions.
10 Two other passages in which the word omen appears are relevant here. At 21.10.2 Julian is euentu laetus et omine and his hopes for the future are strengthened (firmata spe uenturorum – note the wordplay between euentu and uenturorum); this refers to the truly royal welcome he received at Sirmium. At 22.1.2 he is in Dacia, anxiously examining the entrails of sacrificial animals and the flight of birds to learn something about Constantius. Predictions are ambivalent and obscure, but he becomes more confident after interpreting a soldier’s collapse as a favourable sign: omen multo praesentius ipse conspexit, quod excessum Constantii clare monstrabat. For Ammianus’ treatment of omina in the Julianic narrative see Ross 2016, 180-189.
11 There is a further link between the genius passages of Books 20 and 25. On both occasions we are told that Julian informed his close associates about the apparition (20.5.10: iunctioribus proximis rettulerat imperator; 25.2.3: ut confessus est proximis). On 20.5.10 see Brodka 2009, 73 sq., who compares this passage to Ammianus’ source, Julian’s letter to the Athenians (Ep. ad Ath., 284c-285.a). It is Ammianus who speaks about the apparition of the genius; Julian, on the other hand, speaks about Zeus sending a sign to show him the divine will.
12 For analepsis, “the narration of an event that has already taken place by the point in the story where we find ourselves”, see de Jong 2014, 78. As was the case with prolepsis, the term was introduced by Genette.
13 For the connection between these two passages in the Aeneid see e.g. Wlosok 1990b. Of course I am not suggesting that Ammianus was here inspired by Vergil.
14 See Poulle 1999, 37. For the motif of a tutelary deity finally deserting the emperor whom (s)he protected see Hekster 2010. He does not discuss Ammianus’ Julian or Constantius.
15 Jovian’s presentation as etiamtum protector domesticus is also significant. This is the first time he is mentioned in the narrative; he reappears only as the newly proclaimed emperor.
16 Procopius was entrusted with organising Julian’s funeral; Ammianus notes that he disappeared soon after the celebrations and, although a thorough search was made, he was nowhere to be found – nisi quod multo postea apud Constantinopolim uisus est subito purpuratus (25.9.13). Interestingly, Procopius also disappears from the narrative, since Ammianus’ next mention of him is in the context of his usurpation (26.5.8 sq.). His actions between Julian’s funeral and the usurpation are recounted retrospectively at 26.6.4 sq.
17 The end of a book: Ann., 1.81.2 (tanto eruptura ad infensius seruitium); 13.58 (the prodigy of the ficus Ruminalis); 14.65.2 (a flash-forward to the Pisonian conspiracy, three years ahead). The end of a narrative segment: Hist., 2.86.4 (ceteris fortunam secuturis); Ann., 3.19 (the epilogue to the trial of Cn. Piso with three more or less explicit references to the future); 14.9.3 (an old prophecy concerning Nero and Agrippina, adduced retrospectively at the end of the account of Agrippina’s murder). The end of an annalistic year: Ann., 12.57.2 (AD 52: Agrippina’s nimiae spes); 15.22.2 (62: on which see Pigoń 1999); 15.47 (64: prodigia imminentium malorum nuntia: the reference is to the Pisonian conspiracy of the following year). For the last type see Ginsburg 1981, 46: “Tacitus also employs the end chapter to anticipate later episodes in the work”.
18 See e.g. Neumann 1987; Kelly 2009; Bargagna 2015.
19 E.g. 22.1.1: dum haec in diuersa parte terrarum fortunae struunt uolubiles casus, cf. Tac., Hist., 2.1.1: struebat iam fortuna in diuersa parte terrarum initia causasque imperio (both phrases at the beginning of a book – and in both cases “a different part of the earth” refers to the East); 27.9.2: transferendae in alios inuidiae artifex, cf. Tac., Hist., 2.86.2: serendae in alios inuidiae artifex; 28.1.57: is urbanarum rerum status, ut ita dixerim, fuit, cf. Tac., Ann., 1.16.1: hic rerum urbanarum status erat (with ut ita dixerim as a marker of the allusion: thus Kelly 2008, 73). More material in Neumann 1987.
20 For instance, they both pay much attention to political trials and the activity of informers. But also here there is a difference in focus; see Blockley 1975, 114.
21 Cf. Kelly 2008, 183: “In the surviving books Ammianus never mentions Livy, Tacitus or Gellius, authors of evident centrality for his project”.
22 I discussed more fully Tacitus’ proleptical statements in my Polish book (Pigoń 2004); more easily accessible are Pigoń 2003 and 2017 (in which some of the analyses of that book are developed). For Tacitus’ use of anticipation (a topic not much discussed in Tacitean studies) see Devillers 1994, 107-119; Suerbaum 2015, 375-391; Lindl 2019 (I am grateful to the author for sending me a copy of this paper).
23 Note the cross-reference to the narrative of Sabinus’ trial in Book 4. Thus we have the prolepsis at 4.71.1 and the analepsis at 6.4.1, both referring to each other (and thus bridging the time gap between AD 28 and 32). We may compare Amm. 20.5.10 and 25.2.3, discussed above.
24 See Ginsburg 1981. See also, in the context of Ann., 4.71.1, Graf 1931, 19; Martin & Woodman 1989, 253 sq.; Suerbaum 2012.
25 This is one of Ammianus’ two unfulfilled promises (the other is 22.8.35). See den Boeft et al. 2011, 113. Since the flash-forward at 28.1.57 does not have its counterpart in the subsequent narrative, we may regard it as completing prolepsis. On the other hand, 17.9.7 (quoted below, n. 27) is an example of repeating prolepsis. For this terminology see Genette 1972, 109.
26 The closest he comes to such language is perhaps at Ann., 4.35.5, where he appends a few words of comment to his narrative of the trial and death of the historian Cremutius Cordus (whose books were officially burnt as part of the punishment).
27 For other prolepses in Ammianus which refer to a villain’s just punishment, see 14.7.17 (a certain Luscus incites Antiochians against Montius and Domitianus: qui haud longe postea ideo uiuus exustus est); 17.9.7 (Gaudentius spies upon Julian: quem postea ipse interfici iusserat, ut loco monstrabitur competenti; a prospective cross-reference to 22.11.1). The motif is expressed perhaps most strongly at 14.11.24 (non diu postea: the deaths of Scudilo and Barbatio), a section introduced as follows: sed uigilauit utrubique superni numinis aequitas. As for 28.1, we should remember that in this chapter Ammianus departs a long way from annalistic chronology; see especially 28.11.43 (et quamlibet tempestiuum est ad ordinem redire coeptorum, tamen nihil impedituri temporum cursus immorabimur paucis […]: Maximinus’ successors in office). As a result, Ammianus’ proleptic epilogue is only yet another instance of chronological irregularity.
28 See den Boeft et al. 2018, 46. They speak more generally about references “to later events in the life of his [Ammianus’] characters”; here I treat “later career notices” as a separate category.
29 Thus this is an example of external prolepsis, looking forward to an event beyond the chronological framework of the text (e.g. references to Achilles’ death in the Iliad); see de Jong 2014, 80. Most prolepses in Ammianus (and Tacitus) are internal (but see Amm. 29.6.16 on the reign of Theodosius I; other examples are listed in Kelly 2008, 27 n. 46).
30 This obscure figure is not mentioned elsewhere in Ammianus. On the other hand, there is no proleptical notice at 31.7.4, where Richomeres’ later career (e.g. consul in 384) is only implicitly suggested (domesticorum tunc comes). As Kelly notes (2008, 27), “reference to the future is all but absent in Book 31”.
31 Once again, the prolepsis comes at the end (of a short section in this case). See above, n. 17.
32 Moreover, we should connect the beginning of Book 5 and the closing of Book 4. The last item of Ann., 4 is Agrippina the Younger’s marriage to Cn. Domitius. A son of this marriage will be born nine years later, and in 54 will become emperor. There is no prolepsis at 4.75; as R. Syme remarks (1958, 267), “the names were enough”.
33 And, we may add, it is Tacitus the historian who brings it back to his audience’s memory.
34 Probably inspired by Tacitus, cf. Hist., 3.2.1: acerrimus belli concitator (also concitor, a less frequent word, is attested in Tacitus).
35 This is explicitly stated at 16.12.61. Note especially immensum quantum ab eo differens, qui post feros lugubresque terrores cineribus Galliarum insultans multa minabatur et saeua. This may be an allusion to Vergil’s quantum mutatus ab illo/ Hectore (Aen., 2.274 sq.) and especially, in view of the idea of haughtiness, to Ovid’s heu quantum haec Niobe Niobe distabat ab illa (Met., 6.273); Vergil’s passage was also imitated by Statius (Theb., 7.706) and Valerius Flaccus (4.398). For Ammianus as a reader of Ovid see Kelly 2008, e.g. 209-211. For the importance of Vergil in the Res Gestae see (apart from Kelly 2008, passim), Hagendahl 1921; den Hengst 2010, 269-273. The connection between Aen., 2.274 sq. and Amm. 16.12.61 has been noticed in some nineteenth-century commentaries on the Aeneid.
36 Gibbon 1806, 224 sq.
37 For the date (and for the view that Ammianus did not publish his work in two instalments) see Kelly 2008, 8 (“the emerging consensus”).
38 This goes back to clarissimum quondam regem from Tacitus’ rendition of Maroboduus’ letter to Tiberius (2.63.1). The historian underlines that he wrote this letter non ut profugus aut supplex, sed ex memoria prioris fortunae. As in Ammianus, the proleptic epilogue makes evident the contrast between Maroboduus the king and Maroboduus the refugee, but Tacitus puts the accents somewhat differently.
39 For other examples of this kind of prolepsis in Tacitus see e.g. Ann., 12.52.2; 13.32.3; 13.43.5; 14.62.4; 15.64.2.
40 Tacitus says that Sejanus managed to take control over Tiberius non tam sollertia (quippe isdem artibus uictus est) quam deum ira in rem Romanam, cuius pari exitio uiguit ceciditque (Ann., 4.1.2; the whole sketch is placed between genitus at the beginning and ceciditque at the end). Wiesen (1970, 403 sq.) draws attention to ceciditque immediately following uiguit: “Tacitus […] has artfully constructed his period as to suggest […] how short and rapidly covered is the distance from the zenith of power and glory to the nadir of disgrace” (but Wiesen’s interpretation of isdem artibus uictus est is far from certain; see now Woodman 2018, 64 sq.).
41 For Tacitus’ (and Pliny the Younger’s) account of Curtius Rufus see Suerbaum 2015, 493-511. The date of his African proconsulship is not certain, perhaps 56/57 or slightly later (Vogel-Weidemann 1970, 80), so the time-reach of Tacitus’ prolepsis is some ten years.
42 Tac., Hist., 1.29.1: ignarus interim Galba et sacris intentus fatigabat alieni iam imperii deos; 2.70.4: laetus ultro et tam propinquae sortis ignarus instaurabat sacrum dis loci (see Morgan 1992a, 16 n. 10 for a possible additional irony of instaurabat; as for laetus, cf. Verg., Aen., 1.35 and 7.288 with Lyne 1989, 181-185: “disaster-prone happiness”). The so-called second battle of Bedriacum was not fought in exactly the same place as the first one, but Tacitus tends to identify both battle sites.
43 This is the last item of the senatorial narrative devoted to the trial. There follows an epilogue, also strongly proleptical (3.19; see n. 17 above). For a possible allusion in fortuna in occulto tenebat to the circumstances of Claudius’ rise to the throne (see Suet., Claud., 10.2 sq.; C.D. 60.1.2 sq.) see Woodman & Martin 1996, 192 sq.
44 The elevated language used in this passage strengthens the idea of impending disaster. For res Romana, see Constantius’ proud response to Sapor: in proeliis quibusdam raro rem titubasse Romanam in summa uero bellorum numquam ad deteriora prolapsam (17.5.14). For orbis Romanus see 19.2.4: cernentes populos tam indimensos ad orbis Romani incendium diu quaesitos in nostrum conuersos exitium (the siege of Amida).
45 Ammianus emphatically announces this disaster in the opening chapter of Book 31. See Blockley 1975, 53; Brodka 2009, 110 sq. The imperfect armabat in the first sentence is particularly foreboding.
46 See studies collected in Lianeri 2016.
47 Genette 1972, 112: “On ne confondra pas ces annonces, par définition explicites, avec ce que l’on doit plutôt appeler des amorces, simples pierres d’attente sans anticipation, même allusive, qui ne trouveront leur signification que plus tard et qui relèvent de l’art tout classique de la ‘préparation’”. Tac., Ann., 4.75 may be cited as an example (see n. 32 above).
48 Some research on this paper was conducted during my stay in Oxford in the summer of 2018. I am grateful to the De Brzezie Lanckoroński Foundation for granting me a scholarship which made my stay in Oxford possible.
Auteur
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Architecture romaine d’Asie Mineure
Les monuments de Xanthos et leur ornementation
Laurence Cavalier
2005
D’Homère à Plutarque. Itinéraires historiques
Recueil d'articles de Claude Mossé
Claude Mossé Patrice Brun (éd.)
2007
Pagus, castellum et civitas
Études d’épigraphie et d’histoire sur le village et la cité en Afrique romaine
Samir Aounallah
2010
Orner la cité
Enjeux culturels et politiques du paysage urbain dans l’Asie gréco-romaine
Anne-Valérie Pont
2010