If the period 400-550 AD had seen Greece in a phase of revival and heightened activity in town and country, not least because it had become one of the near-hinterland regions of a Roman Empire focused on Constantinople at the head of the Aegean, the succeeding centuries saw rising crisis and a constant threat of the complete disappearance of Eastern Roman power. Following immense depopulation occasioned by the first outbreak of the (probably bubonic) Plague in the mid-sixth century (Durliat 1989), continual recurrences till the eighth century ensured that Greece was underpopulated and its economic productivity shrunken. The migration of Slav peoples from the Baltic regions of North Europe had been underway throughout Central-Eastern Europe for some centuries, so that their arrival in the South Balkans in the late sixth century found insufficient resistance from a weakened Byzantine army or local militias to prevent their rapid colonization of the countryside, from Macedonia in the north to the Peloponnese in the south (Malingoudis 1991). The Empire was preoccupied with prolonged wars in the East against the Persian Sassanian Empire and then, from the seventh century, against the seemingly irresistible Arab armies. The available forces to meet this Slav challenge were only capable (with the help of the imperial fleet), to retain Byzantine control over the capital, major towns, and many coastal strips and islands. Essentially the Mainland Greek countryside left imperial control for much of the period 600-800 AD and came under the dominance of Slav tribes. But evidence on the Aegean islands during the EB era for churches and walled military posts confirms a tenuous grip by the Byzantine fleet on many coastal localities.
The permanent loss of Byzantine control over its Levantine and North African provinces in the seventh century to a new Islamic empire arising in Arabia, although confining the Empire to the South Balkans, Anatolia, and restricted parts of Italy (Figure 17.2), at least allowed it to direct all its energies to reorganizing its rump state and recapturing its core hinterland in Greece. During this phase of crisis a new system of provincial management emerged, where the surviving Empire was divided into a series of small provinces or themes (Haldon 1990), each with an integrated government combining military and socio-economic control. Soldier-farmers ensured the regional productivity needed to make each theme both economically viable and defensible. Steadily between the seventh and early ninth centuries imperial forces defeated the
Slavs and reincorporated the Aegean countryside into the Empire, as well as stabilizing the frontiers to North and East, respectively against new Slav states and the dynasties of the Islamic Empire (the Umayyads and their Abbasid successors).
Settlement in Early Byzantine town and country
Late Roman ceramics up to the early seventh century can be understood reasonably well, evolving out of forms arising in the earlier Empire (Hayes 1972, 1997). Fine wares, especially from North Africa and in lesser quantities from the East Mediterranean, and transport amphorae from numerous centers in the East Mediterranean and the Aegean, allow collections to be dated with considerable accuracy (Pettegrew 2007). Increasingly, the domestic and coarse wares (generally locally made in each region) are also being classified for many provinces, including the Aegean variants. Field survey, our most significant tool for the history of rural sites between Late Antiquity and the Middle Byzantine period, demonstrates that despite the accumulating crises, people continued to dwell or exploit rural estates till the end of the LR era. On the Methana Survey for example (Mee and Forbes 1997), 11 LR sites survive into the seventh century AD. The large villa at the abandoned city of Halieis in the Argolid, partly excavated, was deserted after fire destruction in the seventh century (Jameson et al. 1994). Late Antique towns were hit by a succession of catastrophes, not only natural ones such as earthquakes and the Plague, but increasing Barbarian raids from the third through sixth centuries, culminating in the permanent settlement of the Slav tribes in the sixth and seventh century. Nonetheless many small and large Greek urban sites continued to be occupied throughout the sixth even into the early seventh century AD (the last imports of African red-slip tablewares being a primary dating tool), including excavated sites such as Corinth as well as town sites explored by intensive surface survey, such as the several Boeotian cities.
Archaeological research in Greek cities and rural sites focusing on the post-Roman periods remains very rudimentary, further compounded by a severe problem in recognizing assemblages typical for this period, whether in provincial towns or in the countryside. The archaeological evidence in the Aegean is, however, extremely scanty for this period. If we accept that population was brought to a very low level from the sixth century onwards as a result of warfare and recurrent Plague outbreaks, then the amount and dispersal of material culture should be limited for this phase of two hundred years, with the exception of the larger towns. Even there, however, our recovered evidence remains slight, despite historical sources confirming their widespread continued existence through Early Byzantine times.
When the Byzantine Empire recovers its demographic, military, and economic power from the ninth century onwards, there are also radical developments in material culture, one of which is a new ceramic assemblage of “medieval” character, paralleled around this period in both other Christian lands of Southern Europe and in the Islamic Near East and North Africa. The easily-recognizable feature is glazing on table and other wares, initially perhaps to make them watertight and cleanable, soon however for visual attractiveness (probably imitating expensive metal containers). Such wares spread slowly out ofprecocious seventh-century production foci such as Egypt, Syria, and the city of Constantinople. It is generally in the tenth to eleventh centuries that they become common in the Byzantine provinces, by which time they had already stimulated imitative local production. Outside major cities and maritime-oriented smaller settlements, a serious gap in recognized ceramics for the crisis phase, the later seventh to later ninth centuries, hinders understanding of everyday Aegean life.
As in Italy (Potter 1987), pushing back regional glazed ware production to meet the last early seventh-century Late Roman forms is unsupported, leaving rare occurrences of glazed white wares exported from Constantinople to the provinces as occasional dating evidence (Sanders 1996). What is beginning to be confirmed is that the “Dark Age” saw derivatives of Late Roman wares, chiefly local domestic and coarse ceramics, but also in lesser amounts red-slip and amphora types, survive until the appearance of the “medieval” assemblages (Armstrong 2009,Vionis et al. 2009). Since we still lack a clear assemblage for EB sites after the mid-seventh and until the tenth century AD in the Greek provinces, only future research will create regional settlement maps for Greece during these centuries, or a plan of occupation zones in a surviving town site. Instead we can at least present case studies from rural settlements and excavated urban sites.
The failure of a distinctive Slav culture in Greece to fill the gap surprises scholars. Slav-occupied areas in East-Central Europe do possess distinctive assemblages, although in the early period of migration it is usually in cemeteries that such wares are most clearly recognized. But in Greece comparable pottery is very rare and often their forms and decoration merely assign them to either an earlier or later time-period within the seventh to tenth century AD. Our historical sources (Malingoudis 1991) leave no doubt that the Slav colonization of Greece was a large-scale population movement. Thessaloniki, second city of the Byzantine Aegean, was besieged by a sea of Slav settlers from the surrounding countryside, and Slavs exercised genuine control over the Greek rural landscape rather than existing as harmless infillers of land left empty by earlier depopulation. Indeed for this reason we could not accept the apparent emptiness of the EB landscape, since such colonizers undoubtedly revived land use and settlement activity. We can also dismiss the theory that early Slavs lacked tangible houses, pottery or mixed farming: all these features are well-evidenced throughout the remaining Slav world (Malingoudis 1991).
The Greek peasant population which survived the raids and plagues of the LR era we might suspect largely stayed in their landscapes and intermarried with Slav incomers, hence the occurrence of subRoman ceramics in the EB period. Our working hypothesis to account for the rarity of“Slav” ceramics is that Roman-derivative wares were preferred by both population groups. This merger was aided by the creation of a minor but widespread local production of hand-made coarse wares in final LR times, perhaps to substitute for declining professional supplies, which look similar to genuine Slav ceramics (Gregory and Kardulias 1990, Rautman 1998, Oikonomou-Laniado 2003). On present knowledge, the colonizers arrived with their own Slav ceramic traditions but rapidly converted to utilizing regional pottery in a sub-Roman tradition, making their presence almost invisible in the archaeological record. Up till 2005, our Boeotia project in Central Greece, despite having surveyed many Late Roman villas, villages, and urban sites, and numerous Medieval villages and farms, producing a database of 100,000 sherds, included just one piece of “Slav ware,” from ancient Hyettos city. We favor the view that a mixed Slav-Hellenic peasant population occupied many of our sites during EB times but their ceramic culture lies largely unrecognized in the innumerable “LR” sherds which lack a distinctive fine-dating. The latest research (Vionis et al. 2009) shows that careful examination of such assemblages and those of “medieval” settlements can reveal new variants of older Roman wares, or distinctive new unglazed wares, as probable type-fossils (recognizable types) for the EB era.
Finding EB settlements may be hindered by a new preference for surviving local populations to occupy isolated hilltop refuges, little researched in Greece. Such locations, analogous to post-Bronze Age refuge sites, are widely documented historically and archaeologically throughout the rest of the LR Balkans (Dunn 1997,Poulter 2007),and we discovered one such (Aghios Constantinos) near Tanagra, Central Greece (see Chapter 15). In Italy similar resettlements occur (Francovich and Hodges 2003). But the incoming Slav colonizers were confident militarily, and skillful farmers seeking the best land (Malingoudis 1991); this would surely encourage them rather to reoccupy ancient village and city locations. Surface survey and place-name evidence support this second scenario for the EB era, and if some open locations retained existing peasant occupants there may well have developed a peaceful coexistence with dominant Slav settlers. We do however have one very clear Slav community discovered archaeologically. The ancient sanctuary at Olympia lost its ritual role probably in the fifth century, but was occupied by a domestic Late Roman settlement till perhaps the early seventh century, sheltering behind a fortification (recalling Aghios Constantinos village). It is argued that this community left and was soon replaced by a Slav village, only recorded from a cremation cemetery of at least 35 burials (Vida and Volling 2000). The typology of these burial urns, beginning with early hand-made, undecorated forms and continuing into increasingly more complex wavy-line ornaments and the use of the potter’s wheel, suggests a period of use
Figure 17.1 (Upper) Generalized distribution of major foci of settlement in Byzantine and Frankish Boeotia, from extensive and localized intensive survey, compared with (Lower) the distribution of Greco-Roman cities (triangles) and villages (circles) in Boeotia. A high proportion of settlements are in use in both eras, but their names changed in the intervening period.
J. Bintliff, “Reconstructing the Byzantine countryside: New approaches from landscape archaeology.” In K. Blelke et al. (eds.),
Byzanz als Raum. Wien 2000, 37—63, Figures 11 and 12.
Of some 200 years over the seventh and eighth centuries, paralleling sequences in the Lower Danube and Dniepr Valleys in the North Balkans. Other evidence suggests that we are on the right track.
Thus a map of known Byzantine-Frankish settlements in Boeotia (but excluding small farms and hamlets) is strikingly similar to that of Greco-Roman towns and villages (Figure 17.1). We might argue that in EB times some elements of a reduced Greco-Roman population remained on or near former LR foci of settlement from Late Roman times, merging through intermarrying with incoming communities of Slav farmers. A subsequent major rise ofpopulation and regional prosperity during mature to final Middle Byzantine (MB) times (tenth to twelfth centuries AD), predicted from traditional study of churchbuilding and the historical sources, is certainly matched by a dramatic increase in rural sites from our intensive survey in Boeotia, generally without LR or possible EB occupation. Yet I would suggest that this later phase of revival builds upon the “survival capsules” of “Sub-Roman” times, which in their turn are a simplified network of the Greco-Roman settlement system. We also suggest that these EB sites are in two forms: defensible locations used by indigenous communities on or near antique settlements, as widely evidenced in the north Balkans (Liebeschuetz 2007), at least in the most insecure early part of the EB period, and open sites also on or near ancient sites, where Slav farmers settled down with or without local surviving peasant families.
In the Siedlungskammer (settlement-chamber) model of the German Landeskunde (landscape-knowledge) school (see Chapter 8; Lehmann 1939, Bintliff 2000a— b) diversified landscapes create “constraints and possibilities” for settlement, often in or around certain recurrent places in the landscape (the product of physical topography, soil types, technological and land use regimes, and natural paths of communication). Apparent continuities in the placing of settlements over time can thus result from natural conditions, or recurrent ways of using the landscape, and need not imply genuine continuities of local peoples.
To clarify these contrasted processes we need more precise information. Intensive survey of small districts at the Siedlungskammer scale, combined with historical and archival evidence, can provide major insights in Boeotia (following Text Box).
The Transition from Antiquity to the Middle Ages in Boeotia
TheValley of the Muses is a small, largely enclosed upland valley landscape, which has in the long term normally only seen one nucleated settlement (see Figure 8.4). Intensive survey (Bintliff 1996) found some 50 archaeological sites dating from Neolithic to Turkish times. In Greco-Roman antiquity the central nucleation was a large village at Askra, which was particularly flourishing in Late Roman times. Byzantine ceramics from Askra securely dated to the ninth to twelfth centuries AD suggest a smaller village, but are associated with a large church in the south center of the site (Bintliff and Snodgrass 1988). Lock (1995) has argued that this community is “Zaratova,” which gained an Orthodox bishop by the twelfth century, one of several new bishoprics established by the bishop of Thebes to serve a rapidly growing rural population in later MB Boeotia, very much supporting other evidence we have for general revival in the Aegean from the ninth and tenth centuries onwards.
Only hints of possible EB wares at Askra suggest continued use of the site after its LR florescence, but restudy is in progress. Provisionally more help comes from archival sources. The replacement of ancient Askra with Byzantine Zaratova indicates a dominant linguistic presence of Slav-speakers, the place-name “of the mountain” suiting this upland basin enclosed on three sides by imposing Mount Helicon. The toponym points to the settlement of Slavs in an open, non-defensive site surrounded by highly fertile land, exactly at the period where our ceramic evidence is still ambiguous for continuity of site use at Askra. Boeotia was recovered by the Byzantine army by the ninth century. The subsequent Hellenization and Christianization of the Slavs in Greece explains why by the end of the Middle Ages the village had acquired a Greek
Orthodox name, Panaghia (signifying the Virgin Mary). The precise overlying of the Medieval over the Greco-Roman village might hint at a merger of surviving local peasants and a dominant incoming Slav community, which we hope to test through re-examination of the survey ceramics.
Haliartos, a Greco-Roman city also in Boeotia, offers similar evidence (Bintliff 2000a—b). Destroyed by the Roman army in the second century BC, it never revived as a nucleated settlement during Antiquity. Our total urban survey found scanty Medieval activity, but when we surveyed around and between the houses of the adjacent, modern Haliartos town, we collected plentiful finds of Middle Byzantine, Frankish, and Early Ottoman times. The MB extramural village blossomed into a populous community in the thirteenth to sixteenth centuries AD, with the name Harmena or Charmaina. The reasons for postulating an EB settlement rest again on that community’s name: it appears to be a Slav name
For a church (P Soustalpers. comm.), associated with earlier rather than later Slav settlements. Indeed toponyms may offer complementary evidence to archaeology elsewhere in Greece: the Slav
Languages underwent major changes from the tenth century, but most of the Greek villages with original Slav names belong before this transformation (M. Kiel pers. comm.).
Thespiae offers a further parallel. The Late Roman town is replaced by two villages with the Greek name Erimokastro, in place by MB times on the far eastern periphery of their urban predecessor, according to texts and abundant surface finds. There is now strong reason to support continuity of occupation at the city, if only at village level, during the intervening EB period, both by A. Vionis’ recognition of sporadic EB ware amongst the sherd finds from our 1980s city survey, but also by M. Karambinis’ discovery that fragments of church architecture from the LR Kastro also belong to the “Dark Age.”
The Boeotian evidence resonates with the wider picture: “most of the cities of the Balkan area ceased to exist in the late sixth and/or the seventh century, and social life changed dramatically” (Gregory 2006, 158), but their village replacements are very common. A relatively small number of larger or more strategically vital towns did survive in Greece through the EB era, and we shall present their evidence later in this chapter in the context of their subsequent MB revival.
World History
