Skip to Content

The Survey of Thisve (Byzantine Kastorion) in Boeotia, Greece

Archibald Dunn, University of Birmingham, Project Grant 2012–2013

This survey was designed as a constructive response to the repeated focus of archaeological surveys around Boeotia and its borders in central Greece on sites that were significant in the Greco-Roman era and in Late Antiquity, but which were all without exception insignificant in the Byzantine era, and—in this part of Greece—the Frankish (late medieval) period too. Methodologically excellent systematic surveys of the extensive sites of ancient Aliartos, Askra, Hyettos, Koroneia, Plataia, and Tanagra have had the unintended consequence of providing no case study yet of the continuity or revival of the town or city in question after Antiquity, even in one of the more prosperous districts of Byzantine Greece and in one of the most intensively surveyed districts of modern Greece, an area which, since 1978/79, the then-new research vehicle of the “intensive interdisciplinary survey” has been using as a test-bed. While excavations at Thebes (sometimes the capital of a middle Byzantine province, and always the seat of an archbishopric and the principal town of Boeotia), are increasingly informative, the study of Byzantino-Frankish Boeotia as a whole, like that of any district of the Byzantine world, requires the study of a range of settlements. Thisve, the site of Byzantine Kastorion and Byzantino-Frankish Castorium, is a potentially instructive case in several ways, as already discussed by the principal investigator when reporting in 2009 on fieldwork co-funded by Dumbarton Oaks in 2007–09. Key aspects of its potential value need therefore only be summarized here: the identification of major early Byzantine, middle Byzantine, and Byzantino-Frankish phases in the monumental topography of the site (fortifications, churches, and chapels); the identification of extensive deposits of murex purpurea shells that had been transported inland to the edge of the settlement for the extraction of the imperial purple dye (which is called in several middle Byzantine sources kastorion); and confirmation that the three neighboring fortifications, although Classical or Hellenistic in origin, all have significant early Byzantine phases associated with Late Roman (i.e., early Byzantine) pottery. The principal investigator’s historical research makes the case that Greco-Roman and early Byzantine Thisve (or Thisvai) is the previously unlocated site of the middle Byzantine and Frankish episcopal town of Kastorion/Castorium; that its harbors can be identified in the Vita of Osios Loukas as sites connected with the western Mediterranean; and that the middle Byzantine revival of the settlement (which developed from kômê to bishopric) was to some extent connected with the famous if archaeologically unlocated silk industry of middle Byzantine Boeotia. The name “Kastorion” however, although still known in early Ottoman times, was forgotten thereafter. The settlement in fact acquired an Albanian name (“Kakosi”) in the late Middle Ages, at a time of significant Albanian settlement in central Greece, a name which was only replaced by “Thisve” in the 20th century.

Although the post-Roman settlement’s phases have been the primary focus of fieldwork, other grants held before, during, or after receipt of the grants from Dumbarton Oaks allowed us to contextualize, and thus evaluate more effectively than would otherwise have been possible, the trajectory of Byzantine and Frankish Thisve-Kastorion. Equally valuable to this contextualization process has been coordination with Professor Timothy Gregory of Ohio State University, author of an intensive ceramic survey of the Plain and Lower Acropolis of Thisve and of surveys of one of the harbors of Thisve and of offshore sites. While he and his colleague Professor Bill Caraher of North Dakota University have reevaluated and re-presented their data, other grants allowed us to record the four pre-Byzantine phases of the fortifications of Thisve systematically; to study another of its harbors; to record methodologically instructive post-Byzantine agricultural installations in which Thisve abounds (lênoi and threshing floors); to record at the surface and by remote sensing the evidence of Thisve’s ancient monumental water-management installations; and to model the sedimentary evolution of the Plain of Thisve itself, and therefore the relationship, stratigraphic and diachronic, between a system of ancient dams and the floodplain that they were designed to control for the city. The enigma of the fate of this system after Antiquity is a subject for future research, but our archaeological surveys, Professor Gregory’s surveys of one of the harbors and of the Byzantine offshore settlements in Thisve/Kastorion’s great natural harbor (today the Bay of Domvraina), and the Byzantine texts which the principal investigator has been able to connect to these various sites, all suggest that, in contrast to the other Greco-Roman urban settlements of Boeotia surveyed so intensively since the 1970s, the trajectory of Thisve after Antiquity, despite the probable failure of its ancient floodplain control system, was one of revival and prosperity until the late Middle Ages.

Fig 1: Thisve/Kastorion, phases of the upper acropolis.
Fig 1: Thisve/Kastorion, phases of the upper acropolis.

Data-processing has continued, since the completion of these programs of fieldwork in 2012, with the support of Dumbarton Oaks in 2012–13. The ground plan of the Byzantine Acropolis (“Upper Acropolis”) has been revised in the light of supplementary fieldwork in 2011–12. Our survey has demonstrated that the ancient city of Thisve/Thisvai had no acropolis as such. As was often the case at Greco-Roman cities, the enceinte incorporated a natural acropolis but did not separate this high point from the settled areas by means of an inner wall. We showed that a major phase of construction on this natural acropolis (Phase 4), which is built in a lime mortar-bonded opus incertum typical of the early Byzantine period, henceforth divided the site between settlement and citadel. At its weakest point in this phase this new citadel was given an inner and outer wall, again a relatively common feature of early Byzantine fortifications. Analysis of masonry allows us to distinguish a later phase (Phase 5) in which the interstices between the undressed stonework (of the façade at least) is laced with brick and tile fragments. But this more medieval phase is a much less significant intervention than the early Byzantine one. This early Byzantine phase will be reevaluated in the light of the pottery associated with the early Byzantine adaptations of the three other ancient fortifications of the area by our colleague Dr. Evi Daphi of the Ephoreia of Byzantine Antiquities.

The Differential GPS survey of the whole site, conducted by Mr. Kevin Colls (then of the Birmingham University Field Archaeology Unit; now of Staffordshire University) has been processed to produce the digital terrain model into which the locations and plans of all recorded monuments and features (Ancient, Byzantine and post-Byzantine) will be integrated. All monuments and other features are already interconnected by conventional topographic surveys which have been geo-rectified but these only convey detailed relief in the immediate surroundings of selected monuments and features.

Fig. 2: Thisve/Kastorion, Agios Loukas, overview from the east
Fig. 2: Thisve/Kastorion, Agios Loukas, overview from the east

Data for two of the six monuments selected for 3-dimensional laser scanning have now been processed by Mr. Jonnie Godfrey. Plans, elevations (longitudinal and cross-sections), and angled overhead views have so far been generated for the church of Agios Loukas and for the anonymous church at “Locus 1” in our gazetteer of monuments and features. We have now recorded archaeologically 14 sites identifiable as Byzantine or Byzantino-Frankish churches or chapels. For the ruined church known today as “Agios Loukas” (typically the title given to Osios Loukas on Byzantine and post-Byzantine frescoes and mosaics in this part of Greece) the survey reveals a three-aisled church of middle Byzantine proportions, with evidence of a parekklesion on its north side, a significant tomb with evidence for an arcosolium in the narthex, and an extremely careful revetment of the central apse composed of fine dressed blocks taken from ancient public buildings. It is clear that most of this façade is below modern ground level. Spolia are used throughout the monument as orthostats which may well have formed the bases of immured stone crosses, a treatment of the façades of several middle Byzantine Greek churches. This is the most important of the identifiable churches of Byzantine Thisve and it might have been the seat of the bishop of Kastorion/Castorium.

Fig. 3: Thisve/Kastorion, church at locus 1, ground plan.
Fig. 3: Thisve/Kastorion, church at locus 1, ground plan.

Mr. Jonnie Godfrey has also processed the laser scans of a typical cross-in-square (inscribed cruciform) church (Locus 1) to generate a ground plan, elevations and sections. This church has three semicircular apses like another probable inscribed cruciform church at Thisve (“Agia Triada”). Like all the Byzantine or Byzantino-Frankish churches at Thisve, Locus 1 makes great use of Greco-Roman spolia in its construction, including as orthostats in the construction of the piers of the central crossing. There is no stratigraphy as the structure stands upon exposed bedrock, so an early Frankish date cannot be excluded.

The Project Grant from Dumbarton Oaks has also enabled the creation by Mr. Harry Buglass of revised maps of the archaeological topography of the wider area and the preparation of many illustrations of the site’s Byzantine and Byzantino-Frankish phases for interim and final reports.