Map I: Thrace, Macedonia and Bulgaria ca. 775
Excerpted from: Byzantium and the Bulgaria 775 to 831
As has been noted already, one of the strategic priorities for the imperial government at Constantinople was to protect the rich lands of hrace and Lower Macedonia, which provisioned the principal Byzantine cities, including the capital itself, from Bulgar attacks. From the late seventh century the theme system formed the basis for the defence of these regions. By that time the word theme designated a territorial unit and a military division headed by a strategos who combined both military and civil power.Local defence was organized around the tourma, likewise both a military and administrative division, which in turn was subdivided into a number of banda (or topoteresiai), each under a komes.Like the strategos, the tourmarchai had headquarters of their own, usually a fortress town; they had formal jurisdiction over all those directly under their military command, but were also responsible for the safety of the local population, and for dealing with local raids.the soldiers of the thematic army were almost entirely recruited from local sources. Although there was a regular core of salaried and full-time troops, the local military command and the state were able to rely increasingly on individuals who served, usually on a seasonal basis, at their own or their family’s expense (providing the soldier’s gear, provisions and—in some cases—mount); in exchange the soldier or his immediate dependents received certain iscal privileges (their property, for instance, was exempted from all but the basic public taxes).In this respect, military service entailed privileged status for the soldier’s household. he only formal obligation was with regard to the hereditary nature of military service. hus, the position of every dead or retired soldier was to be taken by a member of his family. his guaranteed, in theory at least, that the nominal number of troops in the thematic army would not be reduced.the first Balkan theme was that of thrace, established by Constantine IV soon ater the arrival of Asparuch’s Bulgars in the peninsula. Its precise extent is unknown. It appears that until the late eighth century its capital was Adrianople, the largest city in the region (although Heraclea, the residence of the metropolitan of Europe, has also been suggested).When Adrianople was detached from the theme of Thrace, the seat of the strategos probably moved to Arkadioupolis. here is evidence to suggest that there were at least two tourmai, one based at Bizye (some 20 km west of the Black Sea coast) and possibly another at Sozopolis.A seal dating from the second half or the third quarter of the eighth century may refer to a third tourma (Τζάτζιος β σπαθ. κ. τουρμαρχ. τῶν Μακεδό[νων]).
This system may have been complemented by the so-called kleisourai, that is, districts including frontier passes, and the territory most immediately threatened by hostile activity, whose commanders usually enjoyed greater autonomy than the other thematic subdivisions. Indeed, in his De hematibus, Constantine VII indicates that before its elevation into a theme in the second half of the ninth century, Strymon (i.e. the area between the Strymon and Nestos Rivers up to the southern slopes of the Haimos) had the status of a kleisoura. Although this is not conirmed by other sources, the establishment of such a command near the Bulgar border makes a great deal of sense in the light of the military situation ater ca. 681. It may be conjectured that the base of the kleisourarches (or archon) was the Aegean port of Christoupolis (mod. Kavala).
In the late eighth century, the territory between the Hebros and Strymon Rivers was detached from hrace and made into a new theme, that of Macedonia. he irst reference to it comes from heophanes who mentions a monostrategos in thrace and Macedonia active in 801/2.the violent Bulgar attack on the army of hrace at Strymon in the late autumn or early winter of 788 provides a irm terminus post quem for the creation of the new command. Because of the pressing need to prevent further Bulgar encroachment into Byzantine territory, Irene, who took a strong interest in Balkan afairs, is very likely to have established the Macedonian theme soon thereater. Her (temporary) removal from power in December 790 may therefore serve as a terminus ante quem. here is no evidence that the empress brought in soldiers from Asia Minor or Anatolia to serve in the army of the new theme. Instead, she seems to have divided the army of Thrace into two smaller units, thus increasing the lexibility of the empire’s defence along the Bulgar border. The total number of troops present in thrace and Macedonia is hard to estimate. The Arab geographer Ibn al-Fakīh, who wrote in 902/3 but apparently drew on material collected by a certain Muhammad b. Abī Muslim al-Jarmi, a Byzantine prisoner until 845/6, gives a total of 5,000 men for each army.It is possible that this igure actually represents the nominal roll of the army rather than a total of active troops.
The active soldiers on the military registers must have numbered considerably less than this, perhaps 2,000–3,000 men in each theme.
he fundamental principles of Byzantine strategy in hrace and Macedonia were, as far as we can tell, twofold: where possible, raiding forces should be held and turned back at the border before they could do any damage. To this end, the Byzantines built or restored and garrisoned a number of fortresses along the major invasion routes (particularly near the Haimos passes) which provided safe bases for the local thematic forces.Where this policy of meeting and repulsing hostile attacks at the frontier did not work (which seems generally to have been the case judging from the fact that Bulgar raids frequently reached the heart of the hracian Plain), then larger armies, oten made up of both the local themes and the tagmata, were to meet the invading raiders and either attempt a direct confrontation or force a retreat. In any event, it was essential to minimize the scale of destruction inlicted on the land and prevent Bulgar penetration into the Constantinopolitan hinterland. As noted already, a series of fortiied settlements and kastra, some of which had been erected during the military resettlements of Constantine V, constituted a strong rampart around southeastern hrace (the most fertile part of the plain), but could also act as refuges for the local population in times of need.These kastra by themselves could hardly prevent the passage of raiders, but at least they limited the availability of easy plunder and forced the invaders, unless they undertook laborious sieges, to extend their ravaging to less fertile areas.
However important these raids may have been for the khan (given that they provided a positive outlet for the energies of the restless warrior aristocracy), his military priority was to defend the Bulgar pasturelands which supported the nomadic culture, economy and institutions of the ruling stratum. Although less fertile than the Wallachian and Bessarabian pasturelands (which the Bulgars also controlled during the period in question), the plains of northeastern Bulgaria were far easier to defend, hence the selection of Pliska as the permanent residence of the khan in the second half of the eighth century. The principle advantage of the so-called “inner lands” was their positioning between two natural protective barriers—the Haimos on the south and the Danube on the north. As already mentioned, the Haimos is far more diicult to cross from the south than the north, above all because its southern slopes drop steeply into the hracian Plain and are composed throughout of limestone.In addition, the deiles are narrow and densely forested, and are therefore easily blocked by human agency or weather. To prevent the passage of raiders, the Bulgars constructed above or behind these passes, but also along the eastern slopes of the Haimos which are comparatively low, an intricate system of earthworks and palisades, the remains of which are still visible today at Riš, Kotel, Dragoevo and Tsonevo, among other places.At the same time, the Black Sea shore was defended by a series of fortiications and ditches with embankments, intended to prevent landing and bar access to the naturally unprotected coastal road which aforded easy access to the Bulgar heartlands.Along the southern banks of the Danube an elaborate system of earthen ramparts, ditches with embankments, and stone-built defences (the latter pre-dating the arrival of the Bulgars in the Balkans) served as a deterrent to nomad or sedentary attacks from the north.Furthermore, a series of fortiied camps along the river functioned as bases for the army from which raids against an invading leet or enemy installations might be mounted, but could also aford protection to the local population and its livestock.Particularly important was the so-called Small Earthen Dyke, built apparently in the late seventh century, and running across the Dobrudja, from the Lower Danube to the Black Sea—a total of 59 km. his is superposed by the Large Earthen Dyke, which is traditionally dated to the reign of Symeon, and by the even later Stone Dyke, one of the most impressive structures of its kind in the Balkans.
Further north, the Bulgars had constructed, probably in the early eighth century, a series of monumental barriers of embankments and ditches, which seem to have functioned both as defensive installations and symbolic lines of demarcation.the most important among them are the so-called North and South Bessarabian Ramparts, the former running from the River Prut to the Dniester (106 km in all), the latter covering an area of 126 km, from Prut to Lake Sasik (Sasicul Mare) by the Black Sea coast. Another embankment dated to this period runs for approximately 22 km by Galaţi, located in the angle formed by the Prut and Siret Rivers. here are two dykes in the Wallachian Plain: the Brazda lui Novac de Nord (some 305 km long) and the Brazda lui Novac de Sud (about half that length), but their attribution to the Bulgars is far from certain.
A key aspect of Bulgar defensive strategy was the movement of population groups to the regions behind the frontier. Theophanes reports that Asparuch transferred the Slavic tribe of the Severoi from its home by the pass of Veregava (Riš) to the eastern borders of the khanate, near the Black Sea coast; another group of Moesian Slavs, the so-called “Seven Tribes”, was established on two diferent parts of the Haimos Mountains, guarding the hracian frontier and acting as check upon the eastern adventurism of the Avars.Similar arrangements were undoubtedly made in Wallachia and Moldavia (near the mouth of the Siret River), where in the early ninth century Krum resettled a large number of Byzantine prisoners from Thrace, who formed a self-governing Bulgar borderland under the supervision of a komes.In all these areas the Bulgars had established a chain of look-out posts covering the various points of ingress into the khanate.
The task of the frontier populations, particularly the Moesian Slavs on the Haimos Mountains, was not only to warn of invasion but also to meet and repel their opponents before they gained access into the Bulgar hinterland.If that failed, they were to follow, harass and dog the invading forces, thus making their expeditions riskier than before.Beyond the frontier regions, we might expect to ind, by analogy with other Eurasian states, sparsely populated areas designed to ofer additional obstacles to the invaders—for instance, making it extremely diicult for them to secure adequate supplies, and limiting the availability of easy plunder.However, the creation of such “wastelands” in Bulgaria is highly improbably due to restricted space. hus, if the distant defence was neutralized, the “inner land” could only be protected by the actions of the mobile Bulgar cavalry, which supplemented the passive protection aforded by the circumferential fortiications of Pliska. On the whole, the Bulgar strategy of establishing successive lines of defence acted as an efective deterrent to most attacks from the north or the south. Nevertheless, an important factor afecting its application was that of human resources. As already mentioned, it is very possible that at times—in special circumstances—the Bulgars might have been able to raise armies of more than 10,000 men, but even then the Byzantine forces are very likely to have been greatly superior. Thus, as successive Byzantine emperors, most notably Constantine V and Nikephoros I, were clearly aware, a coordinated assault on the khanate from several fronts could easily overstretch the Bulgar defence. In the light of this strategic situation, maintaining, if not expanding, the size of the population in Bulgaria was one of the primary concerns of the khans. The Slavs of thrace and Macedonia represented an obvious solution, but any attempt to incorporate them into the khanate was always bound to be met with stif Byzantine resistance.
Although the Bulgar state had only limited resources with which to oppose the Byzantines, it proved far more resilient that one might have expected; for not only was it defended by a ruling elite of nomadic warriors who enjoyed an advantage over their sedentary enemies in virtue of their customs and way of life, but, more importantly, it possessed the institutions and ideology of a centralized state that could include the khan’s non-Bulgar subjects. As will be seen, the co-operation between the latter—whether Slavs or Christians—and the Turkic-speaking aristocracy was vital for the survival of the khanate in the late eighth and early ninth centuries.
