In the Balkans, the subject of Albanian history tends to straddle the spheres of politics and academia. For example, it is not uncommon for Balkan academics outside Albania to favour hypotheses which see the Albanian people migrating as late as possible into the region, thus countering Albanian claims to autochthony.Footnote ¹ Albanian scholars can use the same evidence to support very different conclusions. One point of contention is whether Albanians inhabited territories near Dyrrachium (modern-day Durrës in Albania) at the beginning of the eleventh century. If so, the likelihood increases, for various reasons, that their ancestors were already established in these same territories in previous centuries. The challenge facing scholars of all disciplines grappling with such matters is an unfortunate dearth of contemporary textual sources related to the Balkans and its peoples before the eleventh century.

The controversy centres on a passage by Michael Attaleiates, a prominent and influential lawyer and judge close to the centre of imperial life in Constantinople. Attaleiates completed a history of the Byzantine Empire around 1080.Footnote ² Known in English as the History, this work describes events from the 1030s to Attaleiates’ own day. The narrative body of the text opens with a short moralistic anecdote involving one of Byzantium's greatest generals, George Maniakes.Footnote ³ In 1038, the Emperor Michael IV the Paphlagonian sent Maniakes to recover Sicily from Arab Muslim control.Footnote ⁴ After two years of hard fighting, and having gained most of the island's eastern seaboard, Maniakes led his elite army to decisive victory over the enemy's forces on the plains beneath the western slopes of Mount Etna. Yet with the ultimate prize at last within reach, Maniakes’ admiral of the fleet, Stephen patrikios, a man with close family ties to the emperor, accused the great general of sedition.Footnote ⁵ Maniakes was arrested, imprisoned and then replaced by his accuser. The army now fell under incompetent leadership, and the island was soon lost to Byzantium forever.

At first glance, Attaleiates appears to give this event only a cursory examination. Nevertheless, his brief account has generated debate in recent decades over questions concerning the ethnic identity of one of the peoples named in Maniakes’ army – the Albanoi, who appear together with another people he calls Latinoi. There was once consensus among most scholars that these Albanoi were Albanians. Then, in 1970, Era Vranoussi published an extensive, detailed and erudite study in Greek, contending that Attaleiates’ Albanoi were not a Balkan people, but rather mercenaries arriving from Normandy who fought to establish themselves in southern Italy in the 1020s and 1030s.Footnote ⁶ The eminent French Byzantinist, Alain Ducellier, responded in defence of the traditional interpretation.Footnote ⁷ Their dispute ran for decades.

In the first English translation of Attaleiates’ History, published in 2012, Anthony Kaldellis and Dimitris Krallis included a footnote in their text referencing Vranoussi's thesis, without qualification, in which they defined Albanoi as ‘an antiquarian term referring probably to the Normans (from ancient Alba, near Rome), not modern Albanians’.Footnote ⁸ Given that a future generation of scholars working with the History in English will confront Vranoussi's ideas more as fact than unproved theory, it is timely to review her core arguments concerning the identity of the Albanoi.

Vranoussi began her 1970 paper with an observation not previously addressed by scholars. She noted that Attaleiates’ History includes three passages which, according to the traditional interpretation, all refer to Albanians from the Balkans.Footnote ⁹ In a first step toward establishing her central hypothesis, Vranoussi drew attention to the fact that the three passages employ two different ethnonyms – the first two use Ἀλβανοί, the third Ἀρβανῖται.Footnote ¹⁰

Table 1. Three passages from Attaleiates’ History traditionally interpreted as referring to Balkan Albanians

Observing that Attaleiates was very careful in his use of language and terminology, Vranoussi argued he would never employ different terms to describe the same people group. Noting that the third passage, using Arbanitai, refers to peoples from regions around Dyrrachium, she concluded that Arbanitai was the term Attaleiates used for Albanians.Footnote ¹⁴ By this reasoning, Albanoi cannot also denote Albanians.

Ducellier challenged Vranoussi's assumptions concerning Albanian uniformity in this period. He suggested that Attaleiates’ use of these two ethnonyms could indicate that the Albanians were not a uniform people group. In his view, Albanoi might depict Albanians who were well integrated into Byzantine society and culture, to the point of being equal to the Byzantines in their legal status (isopoliteia), while Arbanitai (named as soldiers among an army hastily assembled by doux Basilakes from regions around Dyrrachium in 1078) could describe Albanians who were less acculturated to Byzantine ways and more pugnacious than their Albanoi counterparts.Footnote ¹⁵ Owing to a paucity of surviving textual evidence, the question remains unresolved.

Vranoussi did not consider the possibility that different Albanian groups might have been known by different names. Her conclusion that Attaleiates used Arbanitai alone to describe Albanians is fundamental to her reading of the first of Attaleiates’ two Albanoi passages. She also argued that a full understanding of this passage requires special attention and wider commentary, and suggested that certain facts only hinted at from one point of view can be clarified by comparing Attaleiates’ text with other contemporary sources, both Byzantine and western.Footnote ¹⁶ The sources she used to this end were, primarily, John Skylitzes’ Synopsis Historiarum and Skylitzes Continuatus.Footnote ¹⁷ This is significant because of the conclusions she drew. By interpolating external information into Attaleiates’ version of events and arguing from silence in the sources, she reinterpreted the first of the History's Albanoi passages so as to preclude the possibility that Attaleiates’ Albanoi were ever a Balkan people.

Before addressing the texts of Skylitzes, Vranoussi scrutinized a short and obscure but significant phrase in this passage which Attaleiates used to describe the geographical location of the Albanoi and Latinoi. It reads, ὅσοι μετὰ τὴν ἑσπερίαν Ῥώμην τοῖς Ἰταλικοῖς πλησιάζουσι μέρεσι.Footnote ¹⁸ Henri Grégoire suggested replacing μετὰ with κατὰ, and gave this phrase in his French translation as ‘qui, du côté de la Rome occidentale, sont voisins de l'Italie’.Footnote ¹⁹ Vranoussi dismissed this offering as incomprehensible, and chided Grégoire for failing to grasp the meaning of the source text.

While admitting the phrase is difficult to understand,Footnote ²⁰ Vranoussi contended that Attaleiates did not describe the Albanoi as ‘voisins de l'Italie’ (neighbours of Italy), but πλησιάζουσι τοῖς Ἰταλικοῖς μέρεσι, in which she read πλησιάζουσι as γειτνιάζουν (to abut upon), and τοῖς Ἰταλικοῖς μέρεσι as ‘Italian territories’ (in line with regionibus italicis from Nathan Rosenstein's nineteenth-century Latin translation).Footnote ²¹ She explained that after the establishment of the Catepanate of Italy (969–1071), Byzantine authors ceased to employ Italia for the entire Apennine Peninsula, but used it instead to describe those specific and confined administrative regions still under Byzantine rule.Footnote ²² Citing an eleventh-century text by Michael Psellos, in which Italia refers ‘not to the whole coast-line, but only to that part which lies opposite us and has appropriated the name of the whole peninsula’,Footnote ²³ Vranoussi concluded that Attaleiates’ phrase τοῖς Ἰταλικοῖς μέρεσι, in this instance, must denote these same regions.

By piecing the various elements together as she defined them, πλησιάζουσι (= to abut upon), τοῖς Ἰταλικοῖς μέρεσι (= Byzantine-occupied regions of southern Italy), and μετὰ τὴν Ῥώμην, which she gives as νοτίως τῆς Ρώμης (= south of Rome),Footnote ²⁴ she interpreted Attaleiates’ obscure phrase as indicating lands between Rome and Sicily or between Rome and Apulia adjacent to Italian territories occupied by the Byzantines.Footnote ²⁵ Thus, in her view, the meaning of Attaleiates’ text becomes clear. This assertion, that the Albanoi lived in southern Italy, lays the foundation for her subsequent arguments that the Albanoi were Normans. It remains to be seen if southern Italy is the only interpretation possible for this short phrase in the History concerning the geographical location of the Albanoi and Latinoi.

Drawing from a very wide body of textual sources, Vranoussi's lengthy paper incorporates an impressive array of citations. When analysing the phrase ὅσοι μετὰ τὴν ἑσπερίαν Ῥώμην τοῖς Ἰταλικοῖς πλησιάζουσι μέρεσι, however, she appears to overlook two passages by a famous Byzantine author that could enable a different interpretation to the one she presents. At issue are tenth- and eleventh-century Byzantine perceptions concerning the geography of ancient Italian territories. Should Attaleiates’ obscure phrase be seen as referring to the Balkans, it would undermine Vranoussi's attempt to re-identify Attaleiates’ Albanoi as Normans in territories ‘south of Rome’.

We begin with a review of the wider historical context. In the fourth century A.D., Constantine I established himself in the eastern half of the Roman Empire and built the city of Constantinople on the shores of the Bosphorus. His successors made the new metropolis their permanent residence. Relevant to this present study is the fact that Constantine also ‘split up the Empire into three large prefectures, each consisting of a number of dioceses’.Footnote ²⁶ These were: the Prefecture of the East, including Constantinople and Thrace; the Prefecture of Illyricum, with its capital at Thessalonike and lands extending to Dyrrachium and environs; and the Prefecture of Italy, which included Dalmatia. According to George Ostrogorsky, ‘it was not until the end of the fourth century that their boundaries took firm shape’.Footnote ²⁷

When the Emperor Theodosius I died in 395, his two sons inherited a divided empire which they ruled over separately, with one in the east and the other in the west. According to John Fine,

Written sources are sparse in their coverage of the Balkans for these centuries. The peoples inhabiting the territories along the eastern Adriatic immediately south of Fine's dividing line, including the mountainous lands of what is today northern Albania, likely engaged with the ancient port of Dyrrachium. When the Byzantines established Dyrrachium as a thema in the early ninth century much of the region fell under its jurisdiction, though determining precise borders across the different periods is difficult.Footnote ²⁹ This background may prove useful when it comes to interpreting Attaleiates’ phrase, τοῖς Ἰταλικοῖς μέρεσι.

In the tenth century, Constantine VII Porphyrogenitus wrote a work concerning the geography of the Byzantine Empire, best known by its conventional Latin title, De Thematibus. In a section dealing with the thema of Dyrrachium, the emperor stated that Dalmatia ‘is a region of Italy’ (τῆς Ἰταλίας ἐστὶ χώρα).Footnote ³⁰ In a later work, known as De Administrando Imperio, Constantine described Dalmatia's territories ‘in olden times’, as follows:

Constantine's assertion that Dalmatia was among τῶν ἑσπερίων θεμάτων (the western themes) with a southern border starting ‘at the confines of Dyrrachium’ is significant. In his obscure phrase regarding geographical location, Attaleiates also used hesperia, describing Rome as τὴν ἑσπερίαν Ῥώμην. Though hesperia means, literally, ‘western land’,Footnote ³² it is also found in ancient Greek texts and Latin epic poetry depicting Italy and Italian regions.Footnote ³³ Kaldellis and Krallis seem to recognize an archaic nuance in this instance, for they translate τὴν ἑσπερίαν Ῥώμην as ‘the Elder Rome’, instead of ‘the Western Rome’.Footnote ³⁴ Attaleiates may have been reflecting on the distant past when he wrote of ‘hesperian’ Rome and related ‘Italian regions’Footnote ³⁵ if, like Constantine Porphyrogenitus a century before him, he understood that territories once part of Italy had extended to the very borders of Dyrrachium. Judging from Attaleiates’ prestigious education, and the breadth of knowledge exhibited in his History, it is unlikely he was ignorant of this perception of ancient Italy's Balkan reaches. Given Constantine's claims that Dalmatia ‘is a region of Italy’, and that, ‘in olden times’, Dalmatia ‘used to start at the confines of Dyrrachium’, then if, in the eleventh century, the Albanoi inhabited areas south of these old borderlands,Footnote ³⁶ it is reasonable to read ὅσοι μετὰ τὴν ἑσπερίαν Ῥώμην τοῖς Ἰταλικοῖς πλησιάζουσι μέρεσι as Attaleiates describing Albanians whose lands abut upon regions to the north which, in antiquity, belonged to Italy and were administered from Rome.Footnote ³⁷ Perhaps when Attaleiates wrote his History in the late 1070s, the lands of the Albanoi were no longer under Dyrrachium's jurisdiction, and his description is thus a form of archaism. Attaleiates also explains elsewhere in his first Albanoi passage that the Albanoi (and Latinoi) remained close allies of Byzantium until c.1040, when they unexpectedly broke with the empire – which, if the Albanoi were Albanians, would have included Dyrrachium. Other evidence suggests that, in these same regions, around the first half of the eleventh century, ecclesiastical sites previously affiliated with the Eastern Church suddenly came under the auspices of Western Christianity and the pope in Rome.Footnote ³⁸

Without considering alternative interpretations for Attaleiates’ obscure phrase, Vranoussi set hers as the first entry in the following list of six elements depicting the Albanoi and Latinoi, extracted from the History's first Albanoi passage: (1) they were established in a specific region of southern Italy, (2) they were formerly σύμμαχοι (allies) of the Byzantines, (3) they shared in the Byzantine's ἰσοπολιτεία (equality of civil rights/commonwealth),Footnote ³⁹ (4) and in their θρησκεία (‘practicing the same religion’), (5) they had their own ἄρχων, and (6) they became enemies (πολέμιοι παραλογώτατοι) of the Byzantines.Footnote ⁴⁰

Vranoussi observed that these elements, taken together, do not depict a tagmatic military unit within the Byzantine army (as Grégoire assumed in his translation), but entire populations under the rule of an archon.Footnote ⁴¹ This observation has merit. Her subsequent conclusion that these Albanoi were somehow Normans in Italy in the mid-eleventh century is, as we shall see, far less robust. It seems remarkable that a scholar of Vranoussi's ability should suggest, in the absence of firm supporting evidence, that Attaleiates – whom she acknowledged was precise in his use of language – applied meaning-laden terms such as symmachoi and isopoliteia to the small and largely disparate bands of recalcitrant Norman mercenaries before 1040,Footnote ⁴² happily selling their swords to the highest bidder and wreaking havoc across southern Italy,Footnote ⁴³ or that Attaleiates would have expressed surprise over one or more of these same bands breaking an agreement with the Byzantines.Footnote ⁴⁴ Later in his History, Attaleiates described Normans (using Φράγγοι) as ‘a race treacherous by nature’.Footnote ⁴⁵ Alexander Kazhdan noted Anna Komnene's awareness of the Normans’ ‘habit of breaking oaths’.Footnote ⁴⁶ For Byzantines writing in the eleventh and twelfth centuries, both treachery and martial prowess are recurring themes in their developing concepts of ‘Norman-ness’. From the little that is known of Normans in southern Italy before 1040, it is difficult to see how they conform to Attaleiates’ depictions of the Albanoi and Latinoi.

In search of other support for her Albanoi = Normans hypothesis, Vranoussi turned to Skylitzes’ Synopsis.Footnote ⁴⁷ Skylitzes reported that the protospatharios Michael Dokeianos replaced Maniakes as head of the imperial army in southern Italy.Footnote ⁴⁸ One day, a Lombard named Arduin, who had led the Norman unit in Maniakes’ army in Sicily, complained that his men were not treated fairly, including in matters of pay. Dokeianos responded by flogging Arduin in public. As a result, Arduin and his Norman followers turned against the Byzantines, inflicting two major defeats on Dokeianos’ army in 1041.Footnote ⁴⁹ In the second battle, at a place called Horai, a large force of Lombards joined with the Franks.Footnote ⁵⁰ Shortly after, the emperor recalled Dokeianos from Italy. Vranoussi drew attention to apparent similarities between Skylitzes’ account of these events and the first of Attaleiates’ Albanoi passages. Both texts report Michael Dokeianos mistreating a foreign leader. According to the Synopsis, Dokeianos’ flogging of Arduin caused a Norman revolt that was later supported by Lombard forces. In the History, Dokeianos’ offence against the archon of the Albanoi and Latinoi turned them into enemies of the empire. On the assumption that the two authors were reporting the same events, Vranoussi equated Attaleiates’ anonymous archon with Skylitzes’ Arduin, the Albanoi with the Normans, and the Latinoi with the Lombards.Footnote ⁵¹ This reading of the History's first Albanoi passage offers support for her hypothesis that the Albanoi were Normans. Unfortunately for Vranoussi's argument, however, this approach to these texts is flawed by a substantial error in Skylitzes’ Synopsis.

In a landmark paper, Jonathan Shepard (without drawing attention to the point) challenged the central feature in Vranoussi's interpretation of the two accounts. He argued that Skylitzes’ report of Dokeianos’ flogging of Arduin was a ‘distortion’, and ‘almost certainly incorrect’.Footnote ⁵² A close review of the textual evidence in various languages led Shepard to deduce that it was Maniakes who punished Arduin, not Dokeianos. Wolfgang Felix arrived at the same conclusion.Footnote ⁵³ Felix also questioned Arduin's role in Maniakes’ army, and determined that the historian Geoffrey Malaterra had correctly described the Greek-speaking Lombard as a liaison and interpreter between the Normans and the Byzantines, whereas Skylitzes saw him erroneously as an independent prince.Footnote ⁵⁴ Vranoussi would have been aware of discrepancies between the sources over who flogged Arduin, for at least two of the authors she cited elsewhere in her paper depicted Maniakes as the perpetrator.Footnote ⁵⁵ The fact that she fails to engage with this issue should raise concerns over her handling of existing evidence. If, as now seems likely, it was Maniakes who flogged Arduin, then Vranoussi's contention that the Synopsis and the History report the same event unravels, for the only specific element the two accounts share – the name of Michael Dokeianos – is eliminated. Furthermore, the dates and locations do not tally, for Maniakes could only have beaten Arduin while still in command in Sicily, long before the Lombard and his followers took to fighting Dokeianos in southern Italy. Given that Skylitzes may have used Attaleiates’ History as a source for his Synopsis,Footnote ⁵⁶ it is also possible that, having read in Attaleiates about Dokeianos offending the Albanoi/Latinoi archon, Skylitzes interpreted Latinoi as ‘Normans’, assumed the passage concerned Arduin's flogging and aligned his account with the History's version. This would explain why Skylitzes ‘erroneously’ depicted Arduin as leader of the Normans (ἀρχηγὸν ἔχοντας Ἀρδουῖνον).Footnote ⁵⁷

In an attempt to further strengthen her case, Vranoussi proposed an etymology for her Albanoi = Normans hypothesis.Footnote ⁵⁸ Working back from the two occurrences of Albanoi in Attaleiates’ History, she suggested that when the Normans first arrived in southern Italy, the existing populations referred to them using a conjectural term albani/aubains, which she derived from the Latin alibi, meaning ‘aliens’. This supposed usage in southern Italy (for which there is no evidence) then passed to Constantinople (again, without evidence) where Attaleiates later learned of it and, alone among Byzantine or any other authors, applied it to Normans (instead of his usual Italoi or Frangoi) – but only twice and both times in events involving Maniakes in the early 1040s. It is worth noting that in the History's second Albanoi passage, Attaleiates named the Albanoi, this time together with Romaioi (regular Byzantine soldiers), as members of Maniakes’ rebel force in 1042–43. This is significant because William of Apulia, writing at the end of the eleventh century, claimed that the Normans of southern Italy all refused Maniakes’ offers of payment to join his insurrection.Footnote ⁵⁹ Jonathan Shepard observed, ‘there is no explicit evidence that they or other Normans took part in the rebellion which [Maniakes] mounted upon returning to Italy in 1042’.Footnote ⁶⁰ Ducellier's response to Vranoussi's proposed etymology was unambiguous: because no evidence exists to show that the term aubain was used in Italy, it cannot be transferred to the text of Attaleiates as denoting Normans. On the basis that the only Greek examples come from the two passages in Attaleiates’ History, Ducellier concluded that the Ἀλβανοί were neither aliens nor Normans.Footnote ⁶¹

The cornerstone of Vranoussi's hypothesis is her insistence that the identity of the Albanoi must be sought in southern Italy. Most of her other arguments stand or fall by this. Given the lack of evidence for her attempt to equate the Normans and Lombards of Skylitzes’ Synopsis with the Albanoi and Latinoi of the History, the highly tenuous nature of her Ἀλβανοί–Albani–Normans etymology, and the possibility that Attaleiates’ obscure phrase ὅσοι μετὰ τὴν ἑσπερίαν Ῥώμην τοῖς Ἰταλικοῖς πλησιάζουσι μέρεσι refers to lands around Dyrrachium, Vranoussi's attempt to make the Albanoi into Normans is, at best, unconvincing.

Vranoussi's approach to the Albanoi question focussed largely on language. She berated scholars for engaging with the History in translation without consulting editions in the original Greek, and she admonished those whose knowledge of Greek she deemed inadequate.Footnote ⁶² She worked closely with Attaleiates’ text, assigning definitions to specific terms and phrases and clarifying its meaning using information from other sources. In her analysis she appears to treat Attaleiates’ first Albanoi passage as a dispassionate historical narrative. Viewing beyond the minutiae of the text, however, reveals that Attaleiates never intended to give a full account of the Sicilian campaign. This point takes on considerable importance, for the passage's structure and limited content suggest a very different purpose, further weakening Vranoussi's hypothesis that the Albanoi were Normans.

In his preface, Attaleiates stated that the study of history ‘has proven to be exceedingly useful for life, as it reveals the lives of those who were virtuous and those who were not, describes illustrious deeds born of flawless planning and effort as well as inglorious actions caused by the faulty planning or negligence of those governing public affairs’.Footnote ⁶³ When past deeds are ‘stripped bare by history’, he continued, ‘they convey clear instruction and set patterns for the future. They simply lead us to imitate what was discerned well and to avoid ill-advised and shameful deeds’.Footnote ⁶⁴ In Attaleiates’ compact and carefully crafted account of the causes and consequences of Maniakes’ downfall – with which he opens the narrative body of his entire work – he clearly aimed to increase the didactic value of the passage by ‘stripping bare’ the ‘inglorious actions’ directed against the great general, presumably to encourage his audience to avoid such ‘ill-advised and shameful deeds’ in the future.

Attaleiates began by establishing the period and context of his anecdote, writing: ‘While the sceptres of the Romans were still held by the blessed emperor Michael, whose homeland was the province of the Paphlagonians, the race of the Hagarenes in the west, in Sicily, was pressed hard by Roman forces, both at sea and on land.’Footnote ⁶⁵ This terse introduction to the 1038–40 Sicilian campaign assumes considerable familiarity with the subject matter on the part of readers, if they are to comprehend the scant reference to ‘the race of Hagarenes’ (Arabs) in Sicily ‘pressed hard’ by Roman (Byzantine) forces ‘both at sea and on land’. Attaleiates adds no information. Having set the scene, he interpreted the events that played out in Sicily as follows:

From other sources we know that Stephen patrikios – admiral of the fleet for the Sicilian campaign and brother-in-law to Emperor Michael IV – falsely accused Maniakes of sedition. We also know that the emperor's brother, John the Orphanotrophos, who held great authority over the empire at that time, had the general arrested and transported in chains to Constantinople. Without naming the perpetrators, Attaleiates links Maniakes’ mistreatment (slandered, removed from his post) with the disasters that followed. This brings us to the crux of the story. Attaleiates now strips bare the ‘inglorious deeds’ behind Maniakes’ undoing, and reveals the tragic consequences for Byzantium: ‘As it was, however, envious resentment brought down the man, his accomplishments, and that great enterprise. For his successors in command made wretched and base decisions, causing the Romans to lose that island along with most of their army.’Footnote ⁶⁷

This first Albanoi passage exemplifies the type of didactic moralizing promised in the History's preface. Attaleiates, a well-informed and well-connected author with access to a very wide range of sources, gives an account of Maniakes and the Sicilian campaign that is devoid of detail. He includes just enough information to show that moral failure in leaders can bring disaster. The shameful debacle involving Maniakes resulted in three significant ‘losses’ for the Byzantines: (1) the island of Sicily and all its riches, (2) ‘most of their army’, and (3) a pair of close allies, deemed so significant that their description takes up a full third of the entire passage. Attaleiates concludes his opening anecdote as follows:

Vranoussi would have us accept that the revelation this passage builds to and culminates in – symmachoi peoples becoming staunch enemies of the empire – involved unruly bands of barbarian outsiders, largely ignorant of the imperial language, culture and religion, who, together with their Lombard insurgent overlords, had tried to drive the Byzantines from territories they still held in southern Italy. If Attaleiates hoped to dissuade his readers from engaging in ‘ill-advised and shameful deeds’, the example of disgruntled Norman mercenaries fighting Dokeianos in 1041 presents a woefully ineffective deterrent. Attaleiates’ moralistic purpose in this short passage is badly served by Vranoussi's reading of Albanoi as Normans in southern Italy.

If, instead, we read Attaleiates’ Albanoi as Balkan Albanians – concurring with Ducellier that (1) the weight of evidence favours this traditional interpretation over ‘Normans’, and (2) the Albanians were not a uniform people group known by one ethnonym only – then it remains to consider briefly the identity of their Latinoi counterparts, described together with the Albanians as former allies of the empire who had also participated with the Byzantines in their commonwealth (isopoliteia) and religion (thrēskeia). By allowing for options beyond just Normans and Lombards when attempting to define Attaleiates’ Latinoi, it may be possible to reconsider the meaning of the History's first Albanoi passage.

Byzantine authors began using the term Latinos as a ‘generic appellation for Western peoples’ in the eleventh century.Footnote ⁷⁰ Of the units serving in Maniakes’ Sicilian campaign, Attaleiates’ Latinoi could denote any of the following: (1) the 300 Normans sent from southern Italy by Prince Guaimar of Salerno,Footnote ⁷¹ (2) the Lombards who joined them or served with regular conscripts from the Byzantine-held regions in southern Italy, (3) the Varangian Guard, (4) Harald Hardrada, the future king of Norway, and his five hundred Scandinavian warriors,Footnote ⁷² or (5) some combination of the above.

Many scholars have assumed that the Latinoi of Attaleiates’ first Albanoi passage were the Normans serving in the Sicilian campaign who left for Southern Italy after Maniakes flogged Arduin.Footnote ⁷³ Furthermore, a reference in Skylitzes Continuatus asserts, among other things, that Maniakes (c.1042) took many Normans (Frangoi) with him from Italy to Byzantium to fight in his rebellion.Footnote ⁷⁴ This text also claims that, after their defeat at the battle of Ostrovo (now in north-western Greece) – which they reached from Dyrrachium on the Via Egnatia – some remained in Byzantium and took on the name Maniakatoi.Footnote ⁷⁵ Shepard observes that this passage in the Continuatus is ‘inaccurate in several respects’.Footnote ⁷⁶ Among the issues he raised is the fact that earlier sources do not name Normans as insurrectionists in the cause of Maniakes. As noted above, William of Apulia stated explicitly that no Normans agreed to join with the general. Outside of Skylitzes Continuatus, those named in Maniakes’ rebel force were Albanoi, Romaioi (regular Byzantine troops) and Varangians.Footnote ⁷⁷ The surmise that Attaleiates’ Latinoi were Normans invites the same difficulties as Vranoussi's Albanoi = Normans hypothesis, for Attaleiates described the Albanoi and these Latinoi in precisely the same terms. Furthermore, the History states it was doux Michael Dokeianos who had offended their archon, which could have occurred only after they left Sicily for southern Italy.

From this, it is possible to suggest two tentative conclusions: (1) that Attaleiates’ account of Dokeianos offending the Albanoi/Latinoi archon concerns a separate and unrelated incident to Maniakes’ flogging of Arduin, and (2) that the three hundred Normans in Maniakes’ Sicilian campaign were not the Latinoi Attaleiates mentioned in his first Albanoi passage. Otherwise, we would have to accept that Attaleiates mistakenly named Dokeianos as Arduin's tormentor, incorrectly described Arduin as an archon and then left him anonymous, despite the detailed description of the Albanoi and Latinoi at the end of this passage. On the basis that Attaleiates’ History was a major source in Skylitzes Continuatus (a text which Skylitzes may also have compiled),Footnote ⁷⁸ and there is evidence to suggest Skylitzes used the History in his Synopsis, it seems reasonable to conjecture that the author in both instances read Latinoi as ‘Normans’ in Attaleiates’ first Albanoi passage and viewed his other information through the prism of that assumption. Catherine Holmes has warned historians that ‘it would be dangerous to assume that Skylitzes was merely a passive copyist and abbreviator whose testimony can be accepted as an accurate transmission of the materials he collates’, and that his ‘active authorship can impose serious distortions on the contents and interpretations of the underlying materials he transmits’.Footnote ⁷⁹

The other candidates for Attaleiates’ Latinoi, apart from Normans and Lombards, are the Varangians and/or a contingent of Harald Hardrada's men.Footnote ⁸⁰ In this case, according to the History's account, these Varangians/Rus/Scandinavians would have – together with the Albanoi – rejected all ties with Byzantium, which, for the Varangians c.1040, were very close, and become staunch enemies of the empire. Such behaviour by Varangians would have shocked the Byzantines, for, since its establishment by Basil II in the 980s, the Varangian Guard had won high esteem and a place of honour in Byzantine society.Footnote ⁸¹ Anna Komnene observed of the Varangians that they ‘bear on their shoulders the heavy iron sword, they regard loyalty to the emperors and the protection of their persons as a family tradition, a kind of sacred trust and inheritance handed down from generation to generation; this allegiance they preserve inviolate and will never brook the slightest hint of betrayal.’Footnote ⁸² Varangians in the emperor's service enjoyed the status of a symmachoi people.Footnote ⁸³ They might also be described (in contrast to Normans) as having shared with the Byzantines in their isopoliteia. Equating the Latinoi of the History's first Albanoi passage with former Varangians presents a more plausible explanation for Attaleiates’ insinuating surprise at their having turned against the empire than the surmise that they were recent Norman arrivals in Italy.

As a working hypothesis, the interpretation advanced here could expand into new and unexpected areas of research. If we accept that Skylitzes may, in this instance, have misread Attaleiates’ Latinoi as Normans, then it opens the way to explore the possibility that the later Maniakatoi of Skylitzes Continuatus were not Normans, but, instead, a combination of Albanians and Varangians from Maniakes’ rebel force who survived defeat at Ostrovo in 1043 and together escaped back up the Via Egnatia to Dyrrachium and the Albanians’ homelands.Footnote ⁸⁴ Psellos wrote, ‘As for [Maniakes’] army, some got away to their native countries without attracting the enemy's attention, but the majority deserted’.Footnote ⁸⁵ That the Maniakatoi Latinoi reappear in the company of a doux of Dyrrachium in events dating to 1078 and the early 1090s, still bearing the name of the great general, suggests they lived to establish themselves as an organized military force and passed that identity on to their offspring.Footnote ⁸⁶ This fits with Attaleiates’ statement that the Albanoi and Latinoi dwelt together under a single archon in lands abutting upon ‘Italian regions’ related to ‘hesperian’ Rome. Viewed against a possible background of Albanians with former members of the Varangian Guard inhabiting the ancient borderlands near Dyrrachium, we can begin to revisit necessary questions about the wider history of this region in the second half of the eleventh century.