The fall of Merovingian Italy, 561–5

After the end of the Gothic War in the mid‐sixth century, northern Italy remained divided between the Merovingian Franks and the eastern Roman Empire. In the 560s the Frankish territories were finally taken by imperial armies, but the end of Merovingian Italy is variably dated between 561 and 565. Drawing on the eastern evidence provided by the panegyrist Corippus, this article argues that there is a hitherto overlooked conflict between the Franks and the empire around the year 565, which finally brought an end to decades of Frankish rule in Italy. As this victory occurred under Justin II, an emperor with a poor military reputation today, this reconstruction of western events further bolsters the case that the successes trumpeted in his early propaganda were grounded in reality.

In either case, Merovingian Italy is interpreted as an entity in terminal decline after Amingus' defeat.
With Merovingian history now revitalized by a wave of studies emphasizing Frankish links to the Mediterranean, it is appropriate to re-examine how the last Frankish-Roman war ended, to reconsider the authoritative syntheses of Stein and Ewig with recent historiographical developments in mind. 8I argue here that by considering Constantinopolitan sources alongside a systematic analysis of western texts, the final end of Frankish Italy can be more definitively placed at the end of the timeframe.Moreover, the often-cited turning point in the 560s, the fall of Verona in 561, ought not to be seen as an episode that prompted the decline and fall of Austrasian Italy, as the available sources force us to be more ambiguous.Instead, a comprehensive end to Frankish ambitions is more likely to have been linked to the revolts that consumed northern Italy around 565, especially once we turn to a so-far neglected text, Corippus' panegyric for Emperor Justin II (r.565-78). 9ustin II's foreign policy is rarely lauded in the scholarly literature, as not only were his wars unimpressive compared to his uncle Justinian's conquests, the beginnings of new conflicts against the Persians and the Avars also further weakened the empire.Justin's more aggressive treatment of his neighbours is then seen as only another damning indictment of the emperor's lack of ability.10 All this is a marked contrast with the court propaganda emanating from Constantinople, as exemplified by the laudatory poem composed by Corippus, which loudly 0 Historiographie (Berlin, 2018), pp.204-7; Heather, Rome Resurgent, p. 292; Meier, Geschichte der Völkerwanderung, p. 823; Arnold, 'The Merovingians and Italy ', pp.453-4.8  trumpeted the emperor's military successes at the beginning of his reign.If Justin had indeed succeeded in securing the last vestiges of non-imperial Italy and defended the province against another Frankish incursion, however, then the optimistic and hawkish foreign policy he subsequently espoused may have been rather more justified.A new narrative of the end of Merovingian Italy will therefore not only enhance our understanding of the Frankish kingdoms, but of the politics of Constantinople as well.

The fall of 'Gothic' Verona, 561-2
As the siege of Verona is often seen as a turning point for Frankish-Roman relations, it is worth beginning with a fresh look at the narrative sources for the city's fall to imperial forces.Verona had last appeared in the historical record amidst the dying embers of the Gothic War.According to Procopius of Caesarea, following the defeat of Baduila (more commonly, but less accurately, known as Totila), the penultimate ruler of the Ostrogoths, in 552, an imperial force sent to accept the surrender of the Gothic city was prevented from doing so by the Franks. 11As a significant portion of northern Italy, including parts of Liguria, the Cottian Alps, and Venetia had previously been handed over to the Franks, it is likely that the Merovingians intervened in Verona to defend their interests amidst the collapse of the Ostrogothic regime. 12his account could imply that the Franks took control of the city instead, or merely that Verona remained under the control of its Gothic garrison.The uncertainty about the city's fate mirrors much of what we know, or more accurately do not know, about the state of northern Italy, as we cannot determine the exact boundaries of Roman, Gothic, and Frankish territories, let alone how they changed in the chaotic 550s. 13erona next appears in the Greek Chronicle of John Malalas, who wrote under Justin II and offers a detailed narrative of Constantinopolitan events up to 565.According to Malalas, in November 562 Narses, the supreme commander in Italy, sent news that two 'fortified cities of the Goths', Verona and Brescia, were taken. 14This matches well the later report by the ninth-century Ravennate author Agnellus, who reported that Narses besieged and took the city in July 561, with his foes allegedly being the 'citizens of Verona'. 15The two dates offered do not conflict with each other, as the larger campaign Narses was engaged in undoubtedly would have involved more than the capture of Verona, so it is quite understandable that the keys to the two seized cities only arrived in the capital at the end of 562, as reported by Malalas. 16n undated fragment from Menander the Guardsman's Greek History, although it does not mention Verona itself, is likewise frequently associated with the fall of the city.He narrates that one Frankish general, Amingus, had refused to let imperial forces cross the river Adige, thereby allegedly breaking the terms of a truce with the empire. 17Despite this obstacle, it would appear from another fragment by Menander that Narses nonetheless crossed the river through his tactical acumen, but no other details are provided. 18As the first excerpt was preserved in between other episodes from 559 and 561/2 that were collected in the tenth-century Excerpta de legationibus Romanorum, it seems reasonable to place this particular encounter between these years. 19Since the Adige flows through Verona and given the chronological concordance with Malalas and Agnellus, the evidence from Menander thus suggests that Narses' attack on the city had involved dealing with the Franks as well.
It is surprising then that there is no definitive evidence for the Frankish Amingus from sixth-century Latin sources.The only possible evidence is the mention of a Chamingus in the Austrasian Letters, but there is no consensus on his dating, as some historians instead see him as someone active during the reign of King Childebert II (r.575-96). 20ven if we prefer an earlier dating, this letter only reveals that Amingus was located some distance away from Gogo, a prominent Austrasian noble, and that the latter had viewed the general as his patron. 21Since Amingus is firmly associated with Austrasia in other sources, as discussed below, the identification of the two figures as one man nonetheless does not strengthen our understanding of his actions in Italy.
The next source to name Amingus, the seventh-century biography of Pope John III (561-74) collected within the Liber pontificalis, is only a little more helpful; it describes Amingus as someone who had 'oppressed' Italy alongside Butilinus, and whose defeat by Narses was a cause for celebration. 22Interestingly, Amingus was listed first instead of Butilinus, who had campaigned in Italy as early as 539 and died during the Austrasian invasion of northern Italy in 553. 23Chronologically it is of course impossible for Amingus to have fallen before Butilinus, given the former's presence near Verona in 561/2, so the phrasing of the papal scribe here may merely be due to artistic licence or, perhaps more interestingly, the result of Amingus' lasting legacy in historical memory.
Based on these early sources, it is already clear that Amingus was an important Frankish figure, but the circumstances of his defeat at the hands of Narses are only described by the eighth-century historian Paul the Deacon, who situated it amidst the revolt of an Ostrogothic comes named Widin.Nothing is known of why this comes, presumably in service to the imperial army, rebelled, but Paul did add the further detail that Widin was apparently supported by Amingus.Both men were defeated by the Roman general Narses, with Widin being taken to Constantinople, while Amingus was slain. 24Combining these fragments together, most of the scholarship on this topic, bar a small minority noting the ambiguity in the evidence, view Amingus' intervention in the battle for Verona as linked to Paul's account of his defeat by Narses, thus making Verona the death blow to Frankish fortunes in Italy. 25t is worth noting then that Paul did not delineate the where and when of this conflict, for this tale was situated within Paul's story of three Austrasian generals' actions in Italy.Butilinus is named first, with Paul following Gregory of Tours in noting his involvement in King Theudebert of Austrasia's Italian expedition in 539.Butilinus and Amingus were then left behind to conquer Italy: the former was later defeated by Narses while wintering in Campania, with Paul then narrating Widin's revolt and Amingus' intervention.This summary concludes with the sorry end of Leutharis, Butilinus' brother, who died between Verona and Trent when he sought to return home with his treasures.Since this narrative offers an impossible chronologyfor Leutharis had died in 554 and so well before Amingus -Louis Duchesne suggested that Paul's source has some relation to the biography of Pope John III outlined above that likewise featured a misleading order of events, though it remains plausible that Paul had a more mundane literary reason to narrate the three commanders' deaths out of chronological order.26 Most importantly for our purposes, however, Paul's authorial choice means that we do not have a location or a time for Amingus' downfall, which opens the question of whether we could definitively place his death near Verona in 561/2.
The sole Greek source to mention the fate of Verona, Malalas, similarly suggests that we should disentangle Widin's revolt from the siege.The Chronicle did not mention the prisoner Widin's arrival in Constantinople alongside the keys to these cities, while its description of Verona and Brescia as 'Gothic' suggests that they were not imperial cities in rebellion at all, but instead holdouts left unconquered by Narses in the preceding decade. 27It would then be tricky to fit Widin's war with the fall of these cities, since he was explicitly described by Paul the Deacon as a rebel. 28Ingeniously, but perhaps unnecessarily, Stein had suggested in 1949 that Narses had first taken the city in 561, which was followed by Widin leading Verona into revolt, forcing Narses to retake the city in 562. 29The more economical solution, especially when we do not have any evidence of the location and time of Widin's revolt, is surely to instead separate the two events.
Moreover, Paul's account of Widin being captured and then sent to Constantinople, itself a very sixth-century imperial practice, would have surely been noted in Malalas Gothorum comiti contra Narsetem rebellanti.'I am grateful to the anonymous reviewer for pointing out that imperial sources would have nonetheless described independent Gothic warlords as 'rebels', regardless of whether they had previously served the empire.As Paul the Deacon was a post-Roman author, however, it is uncertain whether he would have adopted such a Constantinopolitan perspective, though it remains possible that this section was drawn directly from an earlier pro-imperial source.Regardless, even if we treat Widin as an independent warlord, it does not affect the following interpretation that his fall was linked to the end of Frankish Italy c.565. 29 Stein, Histoire du bas-empire, p. 611 n. 1.
the fall of Verona and Brescia.As the chronicler noted the arrival of the captured kings of Vandals and Goths in the capital, after their respective captivity in 533 and 540, as well another 'barbarian' leader's capture in 529, it would be logical for such a minded author to also describe a similar event in 562, instead of ignoring Widin's arrival in favour of only noting the keys to the cities sent by Narses. 30The chronicler's account of the 560s focuses almost exclusively on events in Constantinople, so it is tempting to wonder why further details of the final destruction of Gothic and Frankish Italy, which would have surely been wildly celebrated by Justinian in the capital, were not included.This then leaves open the possibility that, assuming Paul's narrative of Widin's capture is correct, the prisoner had actually arrived in the capital after the accession of Justin II in 565, the year the surviving recension of Malalas' Chronicle ended. 31he pivotal figure of Amingus, who appears in Menander the Guardsman, the Liber pontificalis, and Paul the Deacon, has perhaps understandably made historians keen to connect together his death in battle and his bold stand against Narses along the Adige.Menander's account itself, however, suggests that Amingus was not coming to the aid of an imperial rebel.After all, Amingus was supposedly warned that there was a truce in place between the Romans and the Franks, so he should desist from any violence against the Romans.32Assuming that there is a kernel of historicity to this no-doubt embellished literary account, it would be curious for a Frankish army to be reminded of a truce if it was already invading imperial territories.As the last appearance of Verona in the historical record witnessed Frankish forces preventing a Gothic garrison in the city from surrendering to an imperial army, it is quite plausible that Verona had remained independent in the intervening years, which would also fit with Malalas' description of it being a 'Gothic' city.Amingus may then well have been intervening in favour of the locals, just as the Franks did in 552, which led the imperial army to remind the Franks of their existing treaty.
Agnellus from ninth-century Ravenna offers similarly ambiguous evidence.He only describes how Narses had fought against the 'citizens of Verona' in 561, which potentially suggests that this campaign did not feature much foreign involvement. 33Perhaps more importantly, Agnellus' account completely separates the fall of Verona from the expulsion of the Franks from Italy.The latter event was instead placed eleven chapters later, just prior to his account of Justinian's death in 565 and a comet sighting that can also be dated to the same year. 34This narrative, like all the sources discussed already, needs to be treated with care, for Agnellus seems to have relied upon a local annalistic source that he integrated (not always organically) into his celebration of the archbishops of Ravenna. 35The notices on the fall of Verona and the expulsion of the Franks are likely both derived from this now-lost text, but it would be impossible to discern what Agnellus chose not to include in his Ravennate Liber pontificalis.Nonetheless, we can suppose that at least in ninth-century Ravenna, the end of Frankish Italy was remembered as taking place later than the 561 siege of Verona, which was seen as a separate conflict and not the direct cause of the Frankish defeat.
Indeed, we may even have evidence that the Frankish truce with Narses, alluded to by Menander's account, was quickly restored after Amingus' encounter with Roman forces, a swift timeline that prohibits the total conquest of Frankish territories.In Paul the Silentiary's celebratory poem commemorating the rededication of the Hagia Sophia, delivered either on 31 December 562 or 6 January 563, the panegyrist noted that both the 'Median lord' and the 'Celtic war cry' are silent. 36The former has long been understood as a (very generous) reference to the fifty-year peace recently agreed with Persia, with the passage as a whole praising Justinian's dealings with foreign peoples, even gaining the friendship of 'Indians' (Ethiopia). 37J.B. Bury suggested that the 'Celtic war cry' may have been an allusion to Justinian's intervention in Visigothic Spain, but most scholars, including Paul Friedlander and Berthold Rubin, have interpreted it as a reference to the Franks. 38There is admittedly some ambiguity here, as the same term can easily have referred to the Goths and their recent defeats at Verona and Brescia.But regardless of its precise target, and it need not be entirely accurate either given the nature of the genre, the text could still be a clue that some degree of peace was restored in the Italian peninsula.It would then have been difficult for Narses' sieges to have led to the collapse of Frankish Italy, as the time-frame between the fall of Verona in July 561 and Paul the Silentiary's panegyric could not have been enough to account for such a total imperial victory, especially given the glacial pace of reconquest in the previous decade.Perhaps the Franks were defeated, forcing them to seek terms with the triumphant Narses, or the two forces never came to blows at all, as surviving fragments of Menander do not describe a battle between the Romans and the Franks, only that Narses had circumvented the Frankish forces blocking a river crossing.This interpretation also provides a plausible answer for a question Ewig raised in 1983, of why the Franks did not respond to the mopping up of their Italian territories in 562-5. 39If Merovingian Italy was instead relatively secure thanks to a truce with the empire, then their inaction becomes rather more logical.
A slight corroboration for this interpretation can also be found in Gregory of Tours, the Gallo-Roman bishop key to understanding the sixth-century Merovingian kingdoms.Gregory's brief notices on Childebert II aiming to recover the Italian territories of his father, King Sigibert I (r.561-75), while lacking in geographical detail, do suggest that Sigibert had nonetheless presided over parts of Italy during his reign. 40It then strengthens the case for a quiet aftermath following the fall of Verona, as Sigibert's father, Chlothar I (r.511-61), had died in November or December 561 and what we now call Austrasia, presumably including its Italian possessions, was inherited by Sigibert. 41Unless Gregory only meant to suggest that Childebert was seeking to reclaim Italian territories that his father had only held for a matter of months before they were overrun by Roman forces, we must suppose that the Frankish hold on parts of Venetia held firm at least in 562 and likely beyond as well. 42aken together, we have some grounds to disentangle the fall of Verona and the defeat of Widin and Amingus, the latter still often seen as marking the end of Frankish Italy.Widin's revolt cannot be localized to Verona, while Amingus' face-off against Narses by the river Adige does not require his death at this point, since Menander's surviving fragments do not offer any clues for Amingus' ultimate fate.Even if we continue to 39 Ewig, Die Merowinger, p. 27. 40 firmly assign Amingus' death to 561, then we would still need to decouple the conquest of Frankish Italy from this crisis, as the evidence raised here suggests that some Merovingian territories in Italy were nonetheless preserved, perhaps via the new peace hinted at by Paul the Silentiary.
But regardless of the interpretation favoured, historians should still be more cautious in linking together undated events reported independently in Greek and Latin sources.There remain significant gaps in our knowledge of Italian events in this period and it is important to acknowledge the discrepancies in the available sources.At the very least, the fragmentary nature of the evidence should prevent us from decisively holding on to 561 as a transformative moment, for Amingus' intervention in favour of Widin can be placed at any time between 561 and c.565.Indeed, while in the following section I will propose that a fresh conflict with the Franks began around 565, which could plausibly have included Amingus' campaign to support Widin, it nonetheless does not demand Amingus to have survived until then.

Reconstructing the last Frankish war
The final war between the Roman Empire and the Merovingian Franks is not recorded in western Latin sources.While some historians have highlighted the drawn-out process of the fall of Merovingian Italy, regardless of how they view Amingus' fate in 561, discussions have gone little further than what is provided by the few sources surveyed above. 43ut a unique Latin text written in Constantinople provides a hitherto unnoticed glimpse into how the imperial elite viewed the Frankish threat.Following the accession of Emperor Justin II in 565, the North African poet Corippus composed a Latin panegyric commemorating the new ruler's elevation in 566/7. 44This panegyric naturally glossed over inconvenient facts and painted Justin in the best possible light, meaning that its historical value is often limited, perhaps explaining why it has not yet featured in studies of Italy in the 560s.Its allusions to Justin's military successes, however, deserve greater attention.In the preface of the poem, Corippus waxed poetically on how various foreign peoples, whether Avars, Franks, Goths, Lombards, Gepids, or Persians, had submitted to Rome and its new ruler, which on the surface is very much a topos in panegyrics celebrating an emperor's might and so is not necessarily reflective of the emperor's actual achievements. 45But beneath the rhetoric, this poet's praise is also somewhat grounded in reality. 46orippus' description of the Avar envoys' submission to the emperor, for example, mirrors other accounts of their 565 embassy to Justin. 47enander the Guardsman had described how the Avars, despite this nomadic people's fierce reputation, were nonetheless awed by their initial meeting with Justin and so sought an easier target among the Franks. 48John of Ephesus, a Syriac court insider, said much the same, with the Avar envoys being exceptionally submissive to Justin's harangue. 49As for the Lombards and Gepids, whom Corippus noted as fighting each other of their own accord while Justin's soldiers remained safe, they had likewise recently drawn the attention of the emperor. 50aving repeatedly fought against each other in the previous decades for the dominance of Pannonia, the Lombards and Gepids were once again at war in 566; perhaps to cement his influence among the latter, Justin sent his son-in-law, Baduarius, to intervene in favour of the Gepids. 51his imperial force had emerged victorious, so here too Corippus had some justification to praise Justin's skilful handling of these 'barbarians'.
What then of the Franks and Goths that appear immediately in between these historical allusions?It is worth quoting Corippus' description here: Who could count the Franks, so often defeated in battle, and the conquered Goths, the tyrants captured and cast down? 52s the last well-documented conflicts between the empire and the Franks or the Goths, particularly those recounted in Greek sources, had taken place in the 550s, Byzantinists have not seen this passage as representing the military situation under Justin II. 53In this interpretation, Corippus was not only recycling classical tropes regarding an emperor's military prowess, but also reusing the successes of Emperor Justinian to bolster the new regime, which is hardly auspicious for Justin's own abilities.On the other hand, if the references to recent imperial success against the Avars, Lombards, and Gepids can be supported by other near-contemporary sources, it is surely plausible that Justin's alleged victories over the Franks and the Goths had also occurred at the beginning of his reign. 54his is where I would suggest we place the Gothic comes Widin's revolt and the Frankish dux Amingus' intervention, as in this unique story, preserved only in a later Latin history, both Franks and Goths were involved, matching Corippus' pairing of the two peoples in his preface and mirroring his discussion of the Gepids and Lombards immediately afterwards.Moreover, their fates also coincide with what was briefly outlined by Corippus, for the rebel Widin was captured and sent to Constantinople, while Amingus was killed in battle.The use of the word 'tyrant' is similarly suitable, for it both referred to usurpers and 'barbarian' invaders. 55As any and all rebels against the emperor were termed usurpers by sixth-century imperial authors, even the treason of a relatively low-ranking comes would be suitable for this unfortunate sobriquet.The same was evidently the case in contemporary Burgundy, for the late sixth-century chronicler Marius of Avenches grouped together Gothic kings, the Frankish dux Butilinus, and the Herul rebel Sinduald in his list of 'tyrants' defeated by Narses, so Corippus' inclusion of Frankish and Gothic leaders as 'tyrants' vanquished by Justin is similarly fitting. 56If Amingus was indeed defeated at this point, then a celebration of his defeat in an imperial panegyric would also have been very appropriate, for he had long been a thorn in the Romans' side, having originally invaded Italy alongside Butilinus in the 530s according to Paul the Deacon. 57Given Justin's repudiation of the Justinianic legacy, it may well have pleased him to know that he had bettered his uncle by defeating such a long-lived foe of the empire.
Finally, the concluding statement to this section, noting that these 'barbarians' were punished by their own treachery, likewise would fit well with the reported Italian events, if Widin had recently rebelled and there was (at least recently) a truce between the Franks and the empire. 58These unfortunate descriptions of perfidious 'barbarians' are, of course, also common rhetorical topoi used to denigrate enemies of the empire, but it is again worth highlighting how Corippus, a contemporary witness from the late 560s, nonetheless correctly narrates Justin's successes against Avars, Gepids, and Lombards.The references to Franks and Goths sit between these 'historical' allusions, meaning that we do have to entertain the alternative, that Corippus' words are not just wishful thinking or callbacks to past glories, but rather reminders of recent dangers and, more importantly, recent victories that can be used to celebrate the new emperor.
To this we can add a number of clues from eastern sources that suggest Justin did face off against unspecified military threats early in his reign.In a remarkable poem composed by Julian, the first city prefect of Constantinople to serve Justin, the emperor was described as a 'tyrant slayer'. 59While this has previously been interpreted as a reference to Aetherius and Addaeus, two court intriguers who allegedly sought to poison Justin, Alan Cameron nonetheless concedes that the word 'tyrant' is generally used to describe rebels whom the emperor could claim to have been usurpers. 60The same can be said for another poem boasting of Justin's military success, this time written by Agathias in 566/7, which not only alludes to defeated tyrants (note the plural) in a section celebrating Justin's military prowess in western Europe, but also praises the emperor for bringing peace both at home and abroad. 61ropaganda, as always, needs to be read carefully, but there is no reason to limit possible 'tyrants' to only dissenters in Constantinople or those defeated by Justinian, when the fluid situation in Italy also offers plausible candidates.In any case, even without my reconstruction of Widin and Amingus' war, in 566/7 there was another revolt in imperial Italy, this time led by the magister militum Sinduald, a rebel who was described by the near-contemporary Marius of Avenches as a 'tyrant' and so an exact fit for the rhetorical language highlighted here. 62e can therefore be certain that there were tangible imperial victories in Italy for Justin to celebrate, successes that at least equalled in plausibility the discovery of Aetherius and Addaeus' court conspiracy.More obliquely, we can perhaps also consider a law issued in 566, in which Justin noted 'barbarian incursions' as a cause for the remittance of unpaid taxes. 63Given the lack of military threats to the empire itself in the Balkan and Persian frontiers at that time, continuous conflict in the Italian theatre may have been one motivation behind this novel's composition.
Returning to Corippus, some ambiguity does remain in this interpretation of his praise, as no possible military or diplomatic success against the Persians, the last people to be listed in the preface, can be plausibly spun from what we can gather from the other sources, which could suggest that Corippus' descriptions here remain more fantastical than real. 64However, as Corippus' account of the Persians submitting to Justin is separated from successes against the western 'barbarians' by several lines praising the emperor's family, we have some reason to suppose that even if the 'success' over Persia was more rhetorical than factual, the preceding list could nonetheless have more historical weight. 65Moreover, the tone of the Persian example mirrors the triumphal language used for the Persians by Paul the Silentiary's panegyric in 562, when there were similarly no great victories over Persia, only an uneasy peace treaty. 66Indeed, Corippus may simply have been celebrating the continuation of the peace, as just before the Persian section he boasted of how foreign peoples were eager to secure treaties with the empire. 67Since Menander, Agathias, and an anonymous court poet of this period all describe Justin's increasingly muscular stance towards Persia in 566/7, the language used here actually seems to have accurately captured the anti-Persian atmosphere in Constantinople and so does not imply that the entire preface was a fictional creation. 68ore significantly, in a later passage where Corippus listed the empire's foes at the time of Justin's accession, the poet no longer included the Persians, which suggests that his earlier numeration in the preface did have some grounding in tangible contemporary successes against 'barbarians'.In this part of the panegyric, Justin described to the recently dead Justinian the dangerous peoples surrounding the empire, namely the Avars, Franks, Gepids, and the Goths, who were all preparing for war, which in the (no doubt constructed) narrative is situated immediately after Justinian's death in November 565. 69As the Avars and Gepids were very real problems that Justin had to deal with early in his reign, the appearance of the Franks and the Goths again points to their military threat in late 565.Crucially, this can be contrasted with yet another grouping of hostile peoples in the third book of the panegyric, which explicitly listed Justinian's, not Justin's, victories over the Vandals, Goths, Alamanni, and Franks, a chronological sequence that accurately summarized the chief western foes of the empire, even referencing the shadowy 'Frankish war' of the 550s. 70The two previously highlighted sections, however, clearly referred to recent events that took place after Justinian's death, meaning that Corippus did not need to namecheck the decades-old conflicts with the Vandals and Alamanni, only the far more topical Gepids and Lombards.The repeated appearance of the Franks and Goths in the same extracts is thus evidence of their continuing relevance at the court of Justin II, not a reminder of Justinianic triumphs.
There is a certain logic to a renewed conflict involving the Franks and the Goths at this point as well, for the final years of Justinian's life were not a time of peace.Greek sources report unrest and conspiracies in Constantinople, so dissatisfaction in Italy is likewise possible, as clearly seen in the later revolt of the magister militum Sinduald in 566/7. 71This febrile atmosphere would have thus offered opportunities both for a disaffected Gothic comes and for the Austrasian Franks to intervene from their surviving territories in Venetia.Given the western provinces' importance to imperial propaganda, such conflicts also had a natural place in a contemporary panegyric, not least because Corippus was himself from North Africa.
This reconstruction would fit well with what we can recover from western sources.First is Agnellus' claim that the Franks were expelled from Italy sometime after the fall of Verona in 561.The Ravennate author placed the expulsion just prior to the notice of Justinian's death, which could hint that the last conflict took place around 565, though admittedly it is far from a conclusive interpretation.Secondly, although Paul the Deacon's account of Widin's revolt is undated, two clues later in his narrative are nonetheless suggestive of a struggle between the Franks and the Romans around 565.After the accession of Justin, Paul reports that an Italian bishop named Vitalis, who had previously fled to the Franks in Aguntum in modern Austria, was captured by the Roman general Narses. 72The demise of Frankish Italy then provides a plausible explanation for how this could have occurred, as the eventual Roman victory could have allowed imperial forces to seize a bishop who had resided among the Franks for 'many years'. 73nother more oblique clue lies in the appearance of a Roman general named Francio, who was described by Paul as having defended the island of Comacina in Lake Como for twenty years before he finally surrendered to the Lombards. 74This can only be dated approximately to 588, and 'twenty years' itself is a suspiciously round number, so at best Francio can only be said to have been based on the island in the late 560s; Heinrich Büttner opts for c.565 and Neil Christie c.568. 75Yet the last time we hear of this island is in a letter from around 550 asking Merovingian king Theudebald (r.547/8-555) to look after the population there. 76Given Frankish holdings in northern Italy, it is likely that the island was then ruled by Theudebald, making this request a very understandable one.The fact that an imperial commander was based there from c.568 or earlier may then again suggest the island had fallen under imperial control before that point.If Francio was first placed there after Narses' reconquest, a possible reconstruction given the uncertain dating, then Paul may have inadvertently offered another clue to when this part of Frankish Italy fell into imperial hands.
Taking a broader perspective, it is striking that Austrasia, the only Frankish kingdom with holdings in Italy, was attacked from another side in 566.The aforementioned Avars, having been terrified by Justin II's show of strength in 565, had then decided to find easier pickings among the Franks instead. 77We are fortunate to have a Frankish source to carry on the story, for according to Gregory of Tours, King Sigibert of Austrasia was defeated by the Avars in 566 and had to pay tribute to the nomads. 78s Amingus was affiliated with Austrasia as well, it is then possible that Sigibert faced two recent defeats within a year or so, against the Avars and against the Romans in Italy.Pushing this argument further, there is room to wonder whether the Avar offensive can be tied to the Frankish intervention in imperial Italy, with the nomad attack being prompted by Roman diplomacy.It is after all very convenient for the Avars to have attacked Sigibert just as his Italian holdings were lost and immediately following their embassy to Justin.
Even following the traditional narrative of the 560s, there have already been suggestions of the Avar attack on the Franks being ultimately the result of Roman diplomacy, to draw them away from Roman interests in the Balkans.We can also turn again to the events of 561/2, since in 562 the Avars had already attempted to attack the Franks, but were repulsed by Sigibert. 79Following the traditional interpretation of Amingus dying in 561, this Avar thrust westwards has been interpreted by some as being linked to a Roman diplomatic initiative to harass their foes on another front.While the 566 attack is not explicitly connected to any war in Italy, the assumption nonetheless is that it was intended to harm the Austrasian Franks, who had long been opponents of the empire. 80If we situate the downfall of Merovingian Italy in c.565 instead, as I have argued, then the latter incursion can be understood in the same way, for it aligned with imperial interests at that moment precisely.
Walter Pohl has pushed back against this view and argues that the Avar wars abroad should not be always interpreted as aligned with imperial interests, but even if we give the Avars more agency regarding their choice of targets, then it was still a happy coincidence that benefited the interests of the empire. 81In these circumstances, it should not be surprising to learn from Gregory of Tours that Sigibert then sent an embassy to Justin II 'seeking peace', with a resulting treaty concluded the following year. 82Exactly when this occurred has been debated, but using other evidence Stefan Esders has convincingly dated this treaty to 568 or shortly beforehand. 83The purpose of this mission is similarly unclear, with many suggestions so far arguing either that it was an anti-Lombard effort, for in 568 they had secured a large portion of northern Italy from imperial forces, or that it was to facilitate a united front against the Avars. 84However, as the Austrasian Franks did not attack the Lombards until 574/5 at the end of Sigibert's reign, nor were there further wars against the Avars in the same period, it may well be worth taking Gregory at his word instead, that the Frankish emissaries had only sought peace from Constantinople. 85Faced with major defeats on his frontiers and the loss of his Italian territories, it is quite logical for Sigibert to send ambassadors to make peace with the emperor, to finally put a formal end to decades of conflict between Austrasia and the Roman Empire.
This treaty did not, as it turned out, lead to peace in Italy, for the Lombards swiftly replaced the Franks as the empire's enemy from 568 onwards.They have been variously interpreted as invaders, federate forces invited into the province by Narses, refugees, or, most recently, as rebels against Justin II, but regardless of the interpretation we adopt, their success in securing northern Italy surely owed much to the preceding Roman-Frankish war. 86Despite Narses' strategic acumen, it is nonetheless clear that Merovingian Italy was not an easy entity to destroy, as evident in its survival until c.565.By ending the last vestiges of Frankish rule, however, a new power vacuum was opened, one that would be filled by the Lombards, whose longevity in Italy would far exceed the achievement of the Merovingians.

Conclusion
The end of Merovingian Italy remains a poorly chronicled one, but from the scattered references surveyed, particularly those written in the Roman Empire, it seems likely that there was a renewed conflict around 565. Austrasia fared poorly in this encounter, but its motive is easy enough to understand, thanks to the weakened empire at the end of Justinian's reign.As a result, 561 should not be seen as a decisive turning point, for Frankish Italy was certainly not destroyed and its fate may have had little to do with the fall of Verona that year.Indeed, I have also argued here that Amingus' invasion to support Widin, so essential to many narratives of Italian history in the 560s, can be placed later around 565, as their war fits well with eastern reports of slain and captured 'tyrants'.At the very least, Amingus' death should not be firmly associated with 561.Likewise, Corippus' panegyric, which is still frequently seen as only court propaganda, should be considered anew as a valuable text for historians of Roman foreign policy.
While it is easy to critique Justin's reign with hindsight, particularly due to the eventual quagmire on the Persian front and the loss of much of Italy to the Lombards in 568, his early reign nonetheless deserves a fairer hearing.His boasts may not have been so empty after all, if it was his forces that completed Justinian's conquest of Italy.Moreover, the decisive end of Frankish Italy also opened a new avenue for diplomacy between the empire and the Merovingian kingdoms.Without competing territorial interests in play, it became possible for Justin to pursue a diplomatic reset with the Austrasian Franks, the perennial foes of the empire in the preceding decades.Whether through religious relics, gold subsidies, or the odd hostage, Justin and his successors instead aimed to secure the Franks as their allies, not their foes, marking the beginning of decades of intensifying diplomatic contact between the Frankish kingdoms and Constantinople.Paradoxically then, the end of Merovingian Italy had also brought the Franks closer to the imperial Mediterranean.