At the end of the Persian wars (c. 628/629) Heraclius had reestablished the pre-existing arrangements familiar from the end of the sixth century. Two changes seem to have been the merging of the two praesental field armies into one, and the disappearance of the army of Illyricum as the area was overrun by Slav and other invaders or immigrants. There was also a partial re-establishment of Arab allies along the eastern frontier, along with the restoration of the system of at least some limitanei posts and garrisons. The regional command structure was restored to the situation before Heraclius. The system of defence in Italy and Africa had been unaffected by the Persian wars and remained unchanged.
The Arab Islamic conquests radically altered the strategic and political geography of the whole east Mediterranean region. Following the disastrous defeat of 636, the field armies were withdrawn first to north Syria and Mesopotamia, and shortly thereafter back to the line of the Taurus and Anti-Taurus ranges. The regions across which they were based were determined by the ability of these districts to provide for the soldiers in terms of supplies and other requirements. The imperial field army was pulled back to its original bases in north-west Asia Minor and Thrace, where it becomes known as the Opsikion division. That of the magister militum per Orientem (or ‘Master of Soldiers of the east’), occupied southern central Asia Minor, and became known as the Anatolikon army; and that of the Master of Soldiers of Armenia, now known as the Armeniakon, occupied the eastern and northern districts of Asia Minor. The army of the Master of Soldiers of Thrace, which had apparently been transferred to the eastern theatre in the mid-630s, and had been employed unsuccessfully to defend Egypt, was established in the rich provinces of central western Anatolia, and known thenceforth as the Thrakesion army. By the last decades of the seventh century, the districts across which these armies were garrisoned were known collectively by the name of the army based there. While the distribution of the various units of the field armies across the provinces in this way was certainly connected with logistical demands, it had obvious strategic implications, since it meant that Roman counter-attacks were relatively slowly to organise, and that defence was fragmented and organised locally on a somewhat piecemeal basis.
The provinces which had belonged to the quaestura exercitus established by Justinian did not survive the Slav and Avar invasions of the Balkan provinces (although the empire still controlled much of the Danube itself, through isolated fortresses on the Danube delta and along the coast of the Black Sea); but the Aegean regions continued to function as a source for men, ships and resources, and a maritime corps, known in the later seventh century as the ‘ship troops’, or karabisianoi, seems to have been based to begin with on Rhodes. In the light of the considerably increased threat posed to the empire’s exposed coastline and its hinterlands, brought about by the rapid development of Arab seapower from the 660s, these ‘ship troops’ were to develop into the core of middle Byzantine provincial naval power. In addition to these naval units, the imperial fleet at Constantinople (equipped from the 670s with ‘liquid fire’ projectors) was complemented by squadrons from the thema of Hellas. The armies of the magistri militum or exarchs of Italy, and Africa (which included Sardinia) continued to function, although the latter disappeared with the completion of the Arab conquest of North Africa in the 690s, the army of Italy surviving, on an ever more localised basis, until the demise of the Exarchate of Ravenna in the middle of the eighth century.
The themata or themes were at first merely groupings of provinces across which different armies were based. By 730 or thereabouts they had acquired a clear geographical identity; and by the later eighth century some elements of fiscal as well as military administration were set up on a thematic basis, although the late Roman provinces continued to subsist. The number of themata expanded as the empire’s economic and political situation improved, partly through the original large military divisions being split up into different ‘provincial’ armies (a process begun under Leo III and continued by his successors), and partly through the recovery in the last years of the eighth century and the re-imposition of imperial authority over lands once held in the southern Balkans (begun under Eirene and Constantine VI). The first large thema to be thus subdivided was the Opsikion, which was broken up into three corps, the names of which reveal their late Roman origins – the Boukellarioi, the Optimatoi and the Opsikion. The themata were complemented along the eastern frontier by a series of special militarised districts intended to control key passes and roads into and out of the empire, known as kleisourai. As the empire went back onto the offensive in the later ninth century and after, these were converted into themes in their own right.
The localisation of recruitment and military identities that resulted from these arrangements led to a distinction between the regular elements – full-time soldiers – and the less competent or well-supplied militia-like elements in each theme region. In the 760s a small élite force, known as the tagmata (‘the regiments’) was established under Constantine V (741–775), which quickly evolved into the élite field division for campaign purposes. It had better pay and discipline than both the regular and the part-time provincial units, and this was the first step in a tendency to recruit mercenary forces, both foreign and indigenous, to form special units and to serve for the duration of a particular campaign or group of campaigns. As the empire reasserted its military strength in the east in the ninth and tenth centuries, the role and the proportion of such full-time units became ever more important.
As the empire prioritised a more aggressive strategy, the relevance of the thematic armies, whose primary function had become defensive in nature, meant that new tactical and strategic command structures evolved. New military districts under independent commanders evolved, beginning with the conversion of former kleisourai – small frontier commands – to themata along with the incorporation of conquered regions as themata. Unlike the older themata these were usually quite small, based around a key strongpoint. As ever larger and militarily more effective detachments of the imperial tagmata and similarly-recruited professional units were established along the frontiers so this system grew in extent and significance. From the 970s, these divisions were grouped into larger commands, each under a doux or katepano, independent of the local thematic administration. They formed a screen of buffer provinces protecting the old themata, tactically independent of one another in terms of their available manpower. Similar arrangements were established in the Balkan and western provinces. Such forces, whether on the frontiers or within the provinces, consisted increasingly of mercenary, professional troops or of forces sent by the dependent rulers of the various smaller states bordering the empire.
Over the same period the empire’s naval arrangements expanded from the single provincial fleet and the Constantinopolitan imperial squadrons of the seventh century. By the 830s there were three main naval themata, of the Aegean, of Samos and the Kibyrrhaiotai, in addition to the imperial fleet, and the much smaller provincial fleets of Hellas and the Peloponnese. The maritime front was thus covered in the east, and while continued raiding and piracy was not stopped, it was at least checked and occasionally thrown back. In the west a different situation prevailed. The definitive loss of Carthage and the remaining North African provinces by the late 690s had deprived the empire of its naval bases there, although Sicily probably continued to support imperial flotillas, while there is some slight evidence for imperial naval activity in the Balearics. Sardinia remained an imperial possession. But by the early ninth century the empire seems to have lost interest in the western Mediterranean. Adequate naval support at the time when Sicily and then Crete were invaded in the 820s was not forthcoming, a costly strategic error, since the latter in particular became the source of disruptive raiding activity against the empire’s coastal lands. From the late 840s the Balearics too were providing shelter for Muslim pirates and raiders.