‘The sublime objects of liminality’: the Byzantine insular-coastal koine and its administration in the passage from Late Antiquity to the early Middle Ages (ca. 600–ca. 850)

This paper focuses on the historical development and dynamics of political and administrative structures in regions of a fragmented empire that cannot be simply described as marginal ‘mouseholes’. Rather, it should be acknowledged that these spaces were part and parcel of a wider area (the Byzantine insular and coastal koine), which encompassed coastal areas as well as insular communities promoting socio-economic contact and cultural interchange. More importantly, they also boasted a peculiar set of material indicators suggesting a certain common cultural unity and identity. The koine coincided with liminal territories and the seas on which the Byzantine Empire retained political and naval rulership. Such liminal territories showed varied – yet coherent– administrative infrastructures and political practices on the part of local elites.

post-Justinianic reconquista to the invasions of Sicily and Crete by the Aghlabids and the Andalusian pirates. 1 This was a time marked by existential threats to the survival of the empire.Some were of human agency: the Last Great War of antiquity and the arrival of the army of the Caliphs, together with the so-called Iconoclast crisis as recently reassessed by Leslie Brubaker and John Haldon. 2 Meanwhile, natural events like the waves of the so-called Justinianic plague and the climatic crisis ushered by the end of the so-called Beyşehir Occupation Phase wrought havoc on social, economic, and governmental infrastructures. 3n this sense, during this crucial period, the economic, socio-political, cultural, and ecological outlook of the Great Sea was profoundly transformed: it turned into a more fragmented playing field (although not necessarily a level one) between competing powers such as the Caliphate, the Lombards, and later the Carolingians and the Byzantines; and this in comparison to its exceptionally unified political and economic Roman Imperial configuration. 4 However, the historiography on Byzantium has seldom given due weight to the role of islands and coastal hubs in this fragmentation: the exception is the case of Sicily (to which I will return later) because of its crucial role as a grain supplier for Constantinople from the early seventh century. 5In other words, coastal spaces and large islands (as well as archipelagos like Malta and the Balearics) have tended to be seen as figments of a long-gone dream of the Justinianic renovatio imperii. 6Marginal to the metropolis (for they were located far from the imperial centre) 7 , they were, it has been thought, relevant to the political and military fate of the empire only as bulwarks along the southern Mediterranean frontier vis-à-vis the Arabs (Cyprus and Crete being seen in this light). 8rguing against this prevailing view, I will pair the existence of a Byzantine insular and coastal koine with what Chris Wickham has famously described as 'the uneasy coupling of two wildly different geographical zones: the Anatolian and the Aegean'. 9ndeed, it was partly because of the differences and the crises of these liminal zones have made them more difficult to analyse that the Byzantine heartland became the gravitational centre of the empire's historiography in the passage from Late Antiquity to the early Middle Ages. 10 One should exercise caution, though, especially when considering constitutive elements of such a koine.Large islands and coastal communities served as connective hubs along shipping routes, with the stability of their territorial organization rooted in the intricate network of connections linking production centres and maritime emporia. 11With this in mind, and considering that the Byzantine empire never functioned as a single economic unit (while acknowledging the significant fiscal pull of the capital and buying power of the imperial elites, landowning and salaried, centred around the Constantinopolitan court) 12 , analysing the economic trajectories of the Byzantine koine may help us bridge the gap between what Wickham famously described as the two economic cycles. 13Indeed, Cosentino clearly highlights the economic strength, particularly of large Byzantine islands, well into the ninth century.He concludes: If the empire survived if it was not conquered, this also depended upon the fact that it remained in possession of a strip of islands that connected it from the Cypriot Levant to the Balearics [allowing] the Byzantine Empire to continue to operate large-scale movements of men, foodstuffs, and artifacts between.Constantinople and its insular ports until the reign of Michael II. 14 In turn, this should help us grasp the idea that a multilayered crisis (with clear consequences in territorial and demographic terms as well as economies of scale) did not turn into a jolt pushing the empire over its critical threshold but rather bears witness to its adaptive resilience.
Cécile Morrisson has recently sought to explain the resilience of 'an empire that would not die' by stressing the continuity of fiscal practices and a monetary economy based on the adaptation of coin supply as well as the survival of the urban and trade network and large investment in the army and military installation. 15Johannes Preiser-Kapeler and Alkiviadis Ginalis have recently stressed the robust endurance of maritime connections at the local level within the countless harbours, anchorages, and coastal landing spots of the Byzantine Mediterranean, as persisting even amidst political or economic turmoil that sometimes affected regional or trans-regional maritime trade. 16As will be seen, material culture provides clear witness to the diminished but still coherent density of exchange across the insular-coastal koine as not exclusively depending on the fiscal distributive system. 17inally, Morrisson identifies the emergence of a restructured and effective navy as essential to control the strategic Mediterranean sea routes.Here, she echoes the conclusion of scholars like Hélène Ahrweiler and more recently Elizabeth Jeffreys and John Prior, Telemachos Lounghis, and Salvatore Cosentino, who have pointed to the Byzantine fleets as playing a major role in confronting Arab raids and patrolling shipping routes (especially the so-called trunk route linking the Tyrrhenian and the Aegean) often using insular and coastal urban sites, gateway communities, and 'bunkers' as important bases. 18Indeed, one may conclude that the control of islands (and strategic coastal choke-points) was essential to Byzantium in order to retain control of the 'confetti' of a de facto sea empire.This dispersed but resilient political, cultural, economic, and military 'thalassocracy' has often been neglected by a historiography which has given giving the pride of place to the so-called Byzantine heartland. 19his paper will focus on the historical development and dynamics of political and administrative structures in regions of a fragmented empire that cannot be simply described as marginal 'mouseholes' (as pointedly defined by Richard Hodges). 20ather, it should be acknowledged that these spaces were part and parcel of a wider area (the Byzantine koine), which encompassed coastal areas as well as insular communities promoting socio-economic contact and cultural interchange. 21More importantly, they boasted a peculiar set of material indicators (such as lead seals, coins, and globular amphorae) suggesting a certain common cultural unity and identity. 22The koine coincided with liminal territories and the seas on which the Byzantine Empire retained political and naval rulership, for they showed variedthough coherentadministrative infrastructures and political practices on the part of local elites. 23

Liminality
My use of the adjective 'liminal' to connotate the koine in conceptual and analytical terms is deliberate.Liminality has been described as an umbrella term whose ubiquity and vagueness has made it at once popular and problematic. 24Its use has spanned across disciplines and methodologies, although its origins hark back to the work of anthropologists like Arnold van Gennep and Victor Turner. 25Van Gennep rooted liminality in rites of passage: liminality is therefore identified with an intermediate stage between a separation/detachment from and the final aggregation and reincorporation in a certain social structure and order. 26Turner focused on 'understanding the human reactions to liminal experiences as they shape personality, suddenly foreground agency, and (sometimes dramatically) bind thoughts to experience (like in the case of Christian pilgrimage)'. 27In particular, Turner's conceptualization of liminality has encouraged us to objectify moments and spaces when the distinction between structure and agency ceases to be resolved and understood in classical terms; in this light, spatial and temporary qualities have been highlighted as shared aspects positing the very ambivalence and ambiguity inherent to the idea of liminal. 28Certain landscapes that are intrinsically liminal, islands (and coastal areas) as places where sea and land meet indubitably among them. 29his should help us avoid regarding insular and coastal outposts as liminal spaces for their remoteness and peripherality in relation to centres: liminality does not imply marginality. 30In this sense, my definition of liminal concerning the insular and coastal koine hinges on three foundational characteristics.The first is the notion of the transitional (between an inside and an outside) as intrinsic to the creation of spaces of hybridity in the Mediterranean compounded by frequent peddler/cabotage movements. 31The second relates to the quintessential 'instability' (or, more accurately, temporariness) which dictates the tempo and the unique conceptualizations of cultural and social (and, as will be seen, political and administrative) life in liminal spaces. 32his is clearly connected to the role of the sea, which acts both as a natural barrier and as a looming threat, serving as a perpetual reminder that (maritime) landscapes, whether symbolic or tangible, are subject to continual transformation. 33As Matthew Harpster concludes: 'Considerations of the sea's relations with human activities are a reminder that the koine was not a human-centered world.' 34 My third and last characteristic relates to the concept of 'islandness' as opposed to that of insularity.Islandness is a linguistically neutral term that has socio-identitarian connotations (defining all that concerns islands) as well as spatial associations (for it may be regarded as a constitutive part of the triad of heterotopic spaces together with the sea and the ship). 35he islands of the Byzantine koine are to be considered geophysical, spatial, and human objects of the liminal and sometimes even pointing to the sublime.The sublime has its roots in the prefix sub-(up to) paired with -limen 36 , and, therefore, it hints at moving up and above a threshold.In particular, islands can sublimate (or better elevate) the liminal.In fact, some islands have an 'increased liminality' as embodied by their geophysical characteristics, which often determine their sacred and otherworldly character. 37oth the constitutive spaces of the Byzantine koine, thenthe insular and the coastallay at the interface of the two most sophisticated and coherent socio-economic systems (Caliphate and Empire).I will try to show that travellers, and indeed locals, were cultural actors who connotated the koine as organizational spaces of liminality as mirrored by the fluidity of administrative practices and political structures of governance. 38 shall focus mainly on archaeology and material culture (along with the scanty literary sources available for the period), for these will allow me to track the diverse trajectories of the local bureaucratic machinery.Byzantium did not apply a one-size-fits-all model of administration but rather showed the ability to adopt 'fluid' modes of government defined by its ability to withstand the diverse challenges or adversities of the hour in different geographical areas of the central and eastern Mediterranean. 39I will examine how Byzantium 'shaped its waters'; or, better, how it moulded the structures of government on islands as various as Sicily, Sardinia, Malta, Crete, Cyprus, and the Balearics, using a coastal enclave like Butrint on the southern Ionian coast as a control test case.In this respect, the above-mentioned islands present us with different (and sometimes even contradictory) incarnations of a 'borderland' liminality in the equivocal political allegiances of local elites vis-à-vis temporary or permanent insular residents. 40On the one hand, thematic strategoi based in Sicily played a central role in asserting the military might and political clout of Constantinople in the Tyrrhenian and southern Ionian Sea, as well as projecting the Byzantine soft power along central Mediterranean shipping routes. 41On the other hand, we should weigh in the political ability of prominent figures (though less prominent than the strategoi) like the dukes and archontes in charge of both large insular spaces like Cyprus, Sardinia, or the Balearics and liminal gateway communities like Butrint. 42Indeed, and as liminality pertains to individuals as well as spaces, these 'liminal' figures also contributed to producing creative and expedient socio-political arrangements at the edges of different political systems, economic networks, and religious worlds (the Carolingian, the Byzantine, and the Caliphal).

Sicily: at the heart of the Tyrrhenian sea
The centrality of Sicily to the Constantinopolitan administrative machinery is well illustrated by the story of the Byzantine official Herakleides, in the ninth-century life of Leo of Catania. 43Herakleides traveled to Constantinople from Sicily by entering one bath in Catania and exiting another in the capital. 44The mental proximity between the island and the capital showed by this account pairs with the numerous travels between Sicily and Constantinople documented by the literary and material sources, for the island lay astride the trunk route linking the Tyrrhenian with the Aegean and southern Anatolia. 45But through Sicily, it was possible to reach the Byzantine possessions in the Adriatic and evenvia Malta -Aghlabid in North Africa. 46This is shown by pilgrims' travel logs (like that of Willibald, who in the early eighth century travelled to Jerusalem via Syracuse and Catania, or of Gregory Dekapolite in the ninth century); by the travels of diplomats (like Daniel reaching Sicily from Constantinople in 799); and by a string of Sicilian hagiographies. 47The latter (in this period beyond) witnesses to the social, administrative, cultural, and religious links between the Byzantine Tyrrhenian and the eastern Mediterranean. 48hink, for instance, of the naukleroi who ferried St Pankration from Antioch to Taormina, as reported in the eponymous Life dated to the mid-eighth century: 'they took the journey to the island after they loaded their boat with Cretan oil and different sorts of [luxuries like] incense and sacred vestments from the weavers of Armenia.' 49 In a similar vein, we can mention the Life of St Elias the Younger (born in the Sicilian city of Enna), penned in the mid-tenth century but incorporating written and oral tradition harking back to the ninth (the period of the Aghlabid invasion of Sicily).
Elias' travels (forced or voluntary) took him to North Africa (as a slave captured during a Saracen raid). 50Upon working miracles to the benefit of local Christian and Muslim communities, he was set free and travelled back to Sicily via Alexandria, Antioch, and Jerusalem, sailing 'like a ship filled with all sorts of goods'.(A biblical metaphor for sure, but echoing another mercantile ship: the one the rebellious Byzantine commander Euphemius embarked on to flee to Aghlabid Ifriqya in 827.) 51 Oddly enough, when Elias' fame as a miracle worker had already been established, he was arrested by a local Byzantine stratelates in Butrint (on his way to Constantinople) with the accusation of being a spy for the Muslims. 52s I will return to Butrint shortly, it is important to stress that Sicily's centrality to the routes crisscrossing the Byzantine Mediterranean bolstered its economic resilience and its political importance for the Constantinopolitan governmental structure. 53The analysis of ceramics and globular amphorae, one of the most important guide-fossils for the period under scrutiny (to which I will return), bespeaks a two-fold network of connectivity: Sicily's northwestern coast was oriented towards the Tyrrhenian, while whereas the south-eastern one gravitated towards the southern Adriatic, the Aegean, and Constantinople. 54Moreover, the analysis of coinage's supply and circulation confirms the economic vitality of Sicily: it points to the high level of use of bronze and gold coins (though with a devaluation and ponderal reduction in the first half of the eighth century). 55hen it comes to the political and administrative dimension, one should start from the famous journey of Constans II to Italy in the 660s, when he chose Syracuse as its provisional capital vis-à-vis the mounting Arab threat to the Byzantine possessions in the western Mediterranean. 56After his tragic death, Byzantium started re-shaping its naval commands as a response to the Caliphate naval forces storming the Mediterranean and besieging Constantinople. 57Since the commander of the fleet of the Karabisianoi mutinied at least twice and could not prevent the fall of Carthage to the Arabs in 698, regionalization of the navy commands ensued. 58Sicily, sui generis within the Byzantine insular-costal koine, was elevated to the rank of a theme at the very end of the seventh century, when also the operation of a Sicilian apotheke is documented by sigillographic evidence. 59The establishment of the Sicilian theme not only paired with the creation of the strategia of Hellas and the Kybirrhaiotai theme as based in Attaleia but also entailed 'the acquisition of a new military organization on the part of Sicily between the 730s and the 780s, whose essential characteristic was a capacity for offensive warfare'. 60This coincided with the seizure of the papal patrimony in Sicily (and Byzantine southern Italy) by Leo III and was enhanced by the establishment of a detachment of the Byzantine navy (dromorum stolus Siciliae) under the command of the Sicilian strategoi who led it into battle against the Arabs in 720 and 763. 61igillographic evidence shows that Sicilian strategoi were indeed often members of the Cubiculum, one of the most important offices of the central administration almost exclusively composed of eunuchs (an occurrence seldom documented elsewhere in the Byzantine empire) chosen for their peculiar diplomatic and financial abilities, for they could also broker peace treaties between Byzantium and the local armies of the Caliph (as at the beginning of the ninth century). 62Although some of the Sicilian strategoi famously revolted against Constantinople (like the above-mentioned Euphemius) in the course of the eighth and ninth century, these mutinies were never meant to promote secession from the empire but rather called for more (political) attention on the part of Constantinople. 63The strategoi were often rewarded with a promotion to the rank of Exarch and ended up in Ravenna as the capital of Byzantine Italy till its final fall to the Lombards in 751. 64Moreover, and notwithstanding the confiscation of the ecclesiastical estates in southern Italy and Illyricum, Sicilian strategoi kept diplomatic back-channels with the Papacy open.They were also actively involved in the Tyrrhenian politics with active links to Amalfi and Naples as both still nominally dependent on Byzantium. 65n the one hand, then, such examples of 'physical mobility' as sublimated by 'mental proximity' help us to characterize Sicily in terms of liminal relational space as 'made up of the entanglements and comings together of material, cultural, political [and] boundary-defying forces of through wide ranges of networks of relations; for islands are indeed laboratories for thinking through relationalities'. 66On the other hand, one should stress how contemporary material evidence and archaeology help us to read Sicilian insular spaces as liminal; for liminality often embodies the possibility of cultural hybridity without an implicit or forced hierarchy. 67Indeed, once relocated to Sicily, some Armenians brought with them both a new ceramic type (later exported across Tyrrhenian shipping routes) as well as new architectural traditions (circular or elliptical rural huts) not previously found on the island. 68Sicily is paradigmatic of islands as landscapes of liminality, both for the sheer availability of material and literary evidence and for its centrality in the human geography of the Byzantine Mediterranean.We should, however, not lose sight of the importance of links rather than nodes (in relational terms) as characterizing the 'liminal centrality' of other insular spaces. 69The boundary between Muslim and Christian spaces was in flux, and the Byzantine koine, as a borderland region, saw activity-both contestation and communication-both along and across the frontier. 70

Malta: a liminal archipelago
One good example is Malta, at the crossroads linking the routes Sicily with Ifriqyia as well as those connecting Sardinia and the Balearics (to which I will return) to the eastern Mediterranean. 71The Maltese archipelago acted as a useful stop-over, rather than being simply a distant outpost on the imperial frontier.Instead, Malta and Gozo showed both their relational dependence on Constantinople (via Sicily) and a 'degree of autonomous action in communal terms, managing to economically bridge the political divide existing between the Byzantines and the Arabs'. 72This is shown both by the almost uninterrupted series of Byzantine coins yielded on the island (lasting well into the ninth century as minted in Sicily 73 ) and a series of seals of Byzantine officials in charge of administering the archipelago. 74Indeedunlike Sicily, but in tune with other large Byzantine islands like Sardiniathe Maltese archipelago was ruled by a dux and later an archon, and the islands showed a strong political, military, and religious dependence on Sicily. 75Lead seals of Maltese archontes have been found on the island, including an eighth-ninth century specimen issued by Niketas, archon kai droungarios of Malta; which, however, was not found on the archipelago or in a nearby region under Byzantine control but rather surprisingly in Tunisia. 76he evidence provided by globular amphorae, as well as other Byzantine productions, also points to the role of the Maltese archipelago in redistributing goods to the western Byzantine outposts (Sardinia, the Balearics, and fortress of Septem/ Ceuta (the latter falling to the Muslims on the verge of crossing to Iberia in 711). 77In this connection, recent extensive surveys in urban (Mdina) and rural (Hl Safi) areas of Malta have yielded evidence of imports and trading networks centred on the archipelago in the eighth and ninth century (and even after its conquest of by the Aghlabids in 869/70). 78As Matt King concludes: 'The presence of amphorae and fine ceramics indicates that areas of Malta, [also those] not located on the coast, were populated in a period of supposed desolation.' 79 Indeed, in the tenth century, archaeology confirms the arrival of Constantinopolitan Glazed White Wares II, matted wares (ceramica a stuoia as produced in Sicily in the area of Rocchicella of Mineo), as well as Otranto amphorae. 80These finds complement locally made amphorae and cooking pots of Islamic tradition, bearing witness to the archipelago's role as a point of contact between the Islamic and Byzantine shipping and economic networks in the early Middle Ages. 81t is important to note that the commercial (and fiscal) network, which hinged on Sicily with outreach in Malta, extended to the Byzantine Adriatic and the Ionian Sea, especially after the loss of Ravenna. 82The end of the Byzantine exarchate brought about a drastic reshuffling of Byzantine administrative structures in the Byzantine southern Ionian and upper Adriatic.Sicily seems to have held a leadership role in the Adriatic after the fall of Ravenna (at least till the foundation of the Theme of Kephallenia around 765/70). 83This is shown by coins minted in Syracuse under Constantine V's reign and increasingly appearing in Dalmatia and by the literary reference to the activity of the stolus Siciliae in the Adriatic in the second half of the eighth century. 84The Byzantine military presence and diplomatic/religious soft power in both Dalmatia (a region which showed some dependence on Ravenna before 751) and the southern Ionian sea was an emanation of the Sicilian theme but predicated on local gateway communities whose socio-economic life and political legitimization had a clear Byzantine imprint. 85trint: a gateway community on the southern Ionian sea A good example of such a community is represented by Butrint, one of the best excavated and published sites in the early Byzantine Mediterranean, located on the southern Ionian coast.86 In particular, ceramics found in two collapsed towers of the walled enceinte (all dated to the seventh and eighth century) presents us with a complex picture.87 They consisted mainly of coarse wares of local production and of imported globular amphorae.The latter have been cogently defined as a new family of transport vessels appearing in the seventh century and soon becoming the main form for trade all over the Empire and those regions maintaining contact with Byzantine heartlands.' 88 A marker of connectivity to which one should add GWW of Constantinopolitan provenance.GWW all point to a southern Adriatic commercially, socially, and culturally (as well as politically) connected with the Byzantine world with more interregional cabotage as well as tramping voyages in the eighth and even the ninth century (on small status ships similar to some of those found in Yenikapı).89 Like Zadar, Butrint was then 'upgraded,' although only to an archontate in the ninth century (as shown by archaeological, sigillographic, and literary evidence).90 As the local archon interceded for St Elias when he was imprisoned, he also presided over an active shipping network that reached out to Constantinople and the central Mediterranean via Otranto, Sicily, and Malta.91 Although not an urban community (with a settlement and population in comparison to Late Antiquity), Butrint is revealed as one of those (fortified or unfortified) liminal communities poised to benefit from connectivity.Such connectivity until the mid-ninth century was predicated upon the funnelling of prestige goods exchange as well as the importance of naval stations acting as a trading hub: this is exemplified by the presence of several types of seventh-to-ninth century imported amphorae.92 Butrint was a gateway community whose liminality is predicated on its abutment between different patterned connectivities: a special kind of settlement: an entrepot in a new sense, at the same time drawing its character from its position at the interface between different regional or sub-regional network systems (Tyrrhenian, south Ionian, Aegean).93

Sardinia and Crete
For a better understanding of the function and importance of the above-mentioned coastal archontates as part and parcel of the Byzantine koine, we should, however, return to the insular worlds of Byzantium.In Sardinia and Crete in particular we can document a good degree of loyalty and quest for Constantinopolitan legitimization on the part of local elites who were invested with Byzantine dignities and titles and regarded as local military and imperial political representatives.On both islands, we are again confronted with the presence of archontes, although these appeared earlier in Crete and only much later in Sardinia. 94Archon is a rather neutral title: it generally defines any officials possessing powers [and later] mainly governors.Scholars often conclude that archontes were mainly in charge of the most isolated enclaves of the empire, but this remains a rather general reading of the evidence, and it implies a lack of interest on the part of Constantinople towards its periphery. 95Indeed, archontes could not only promote local political initiatives as stemming from the local elites, but, as we have seen, they also based their legitimacy, status, and political pre-eminence on the recognition received from Constantinople. 96n Crete, sigillographic evidence points to the existence of some basiliko[i] spathari [oi] kai archon[tes] Kretes, which were active in the capital Gortyn between the mid-eighth and the early ninth century. 97Cretan archontes are also mentioned by the ninth-century Taktikon Uspensky, along with a strategos who remains elusive, as he is not elsewhere documented: the Kibyrrhaiotai theme did not include Crete. 98Cretan archontes seem to have presided over a network of local imperial representatives, again documented by lead seals struck by local military officials and fiscal officers. 99An elaborate urban defensive upgrade was implemented in the late seventh century across different Cretan sites, pointing to the direct involvement of imperial administration with an important economic and political role also played by local landowning elites and bishops. 100In Byzantine Crete, then, we may surmise the existence of a persisting link between local secular and ecclesiastical elites and imperial representatives since the island was also part and parcel of the so-called Aegean economic system (as shown by ceramic evidence). 101At the same time, material culture also points to the role of the island as a real threshold between different empires: Cretan harbours were central to naval traffic moving east-west and north-south. 102The real watershed in the history of the island was indeed represented by the invasion by Andalusian pirates in the 820s (roughly the time of the Aghlabid invasion of Sicily). 103 rather similar picture can be drawn for Sardinia, this time as part of the Tyrrhenian system of exchange.Here, and by contrast with Crete, we possess a rather larger variety of evidence, including inscriptions, architecture, coins, ceramics, and, of course, lead seals. 104In Sardinia, eighth-ninth century globular amphorae, from several sites across the island, reveal an island connected with the outside world through long-range commercial traffic involving the entire Italian peninsula and linked to both the Byzantine and Arab long-range spheres. 105This is confirmed by local coins minted in Cagliari after the fall of Carthage, which marry with Islamic ones, pointing to a transregional and transcultural acceptance of different monetary units. 106This notwithstanding the collapse of Byzantine administrative and military structures in Africa, which led to their transfer to the island, then at the forefront of the Byzantine-Arab military confrontation until a peace treaty was signed in 752. 107vidence of the increased military importance of the Sardinian duke is shown by the presence of a Byzantine naval squadron in the Tyrrhenian, and it is boasted of in a mid-eighth-century inscription celebrating a great victory by the consul and dux

Conclusion
I have sought to offer a brief overview of the structures of political governance and administration that characterized some of the most important spaces of the Byzantine Mediterranean koine between the late seventh and the early tenth century.
First, I have sketched the character of the local ruling authorities, with Sicily being the only insular Byzantine theme; indeed, it retained direct communications and a strong political link with both Constantinople (as highlighted by the eunuchs of the Cubiculum) and the Adriatic, as exemplified by the political trajectories of Butrint, Malta, Crete, and Sardinia.If Butrint, Malta and Crete were ruled by archontes, Sardinia was a duchy which only later turned into an archontate as well; it also showed strong links with their regional economic and shipping system of reference.Crete benefitted more from the proximity to Constantinople via the Aegean, whereas Malta abutted into the Tyrrhenian Sea, where the gravitational pull of Sicily (and of Rome) was stronger.In a similar vein, the Balearics and Cyprus played a peculiar role: though ruled by Byzantine archontes they both experienced peculiar government-sharing policies involving local elites.
Second, it is clear that Constantinople did not turn its back on the western Mediterranean but retained a continuous and active interest even in the more distant territories like Sardinia and the Balearics into the tenth century, particular through trade involving as enhanced in particular by material culture (globular amphorae and luxury goods).Finally, this interest was modulated by the pressure exerted by other military and political actors and by the ebbs and flows of Byzantine naval power in the period.In other words, we are dealing with an empire whose local administrative structure adapted to the changing velocity of regional politics although retaining control of areas that remained nodal hubs as part and parcel of different economic systems.Finally, I showed such islands (and coastal gateway communities like Butrint) were less marginal military outposts than they were relational and connecting spaces of the liminal.
As I have included a rather general definition of the liminal as a transitory time and/ or place of transition, I have also stressed the importance of approaching liminality in organizational terms: islands (and coastal enclaves) remain distinctive spaces 'in-between' both cognitively and geographically.As they remained connected and isolated, definite and self-existent worlds apart and part of a network, I have used the concept of threshold to sublimate the ambiguity of insular (and coastal) spaces, as this has allowed me to overcome differences in size, composition, and geo-morphology (archipelagos like Balearics and Malta vis-à-vis insular microcontinents like Cyprus or Sicily), and, finally, location (very distant or very close, mentally or physically, from what is often regarded as the only centre of the empire).I have also bound liminality to the concept of relational space.Indeed, although in different ways, Sicily, Cyprus, Malta, Sardinia, and the Balearics (as well as Butrint) were all central actors of a matrix of fiscal, commercial, administrative, and cultural interactions as interwoven