You are here:Home/Research/ Byzantine Studies/ Project Grants/ The Survey of Thisve (Byzantine Kastorion and Multiperiod Complex of Sites): 2007–2009

The Survey of Thisve (Byzantine Kastorion and Multiperiod Complex of Sites): 2007–2009

Archie Dunn, University of Birmingham, Project Grant 2008–2009

Fig. 2: Western Boeotia (Dunn 2008–2009)
Fig. 2: Western Boeotia. Reported late Roman fortifications and loci of maritime traffic (top). Reported Byzantine and Frankish fortifications and loci of maritime traffic (bottom).


Since the British Academy’s Boeotian Archaeological-Geological Expedition began in 1978, Boeotia, located in central Greece, has become the noninsular Greek historico-geographical space that has been the most fully covered by teams applying multiperiod intensive interdisciplinary approaches to the archaeologies of settlement and landscape. Seven to eight surveys of this kind have been or are being conducted in there, including the present survey of Thisve in western Boeotia. Boeotia, meanwhile, has a relatively rich historical bibliography for all periods, beginning with editions of its Mycenaean Linear B tablets. The quality and quantity of the region’s historical and archaeological materials, particularly those derived from surveys, are now such that a project can be designed that engages with problems specific to the Byzantine-Frankish millennium without abandoning the multiperiod and interdisciplinary approaches, including landscape archaeology, that are helping investigators to unlock the pasts of some circum-Mediterranean regions.

In western Boeotia the Thisve Basin and its area of access to the Mediterranean, the adjoining Bay of Domvraina, are a case in point. A largely unrecorded multi-period site with clear evidence of ancient and Byzantine urbanizing phases at Thisve, a much better-recorded but unintegrated settlement archaeology of the wider Basin and Bay, and unmistakable signs of a rich, unpublished archaeology of long-term landscape-management and change around the Basin, are being studied and reanalyzed by a team from the Institute of Archaeology and Antiquity at the University of Birmingham, England, in collaboration with members of the Greek Archaeological Service, the American School of Classical Studies at Athens, and specialists from other institutes of research and higher education (see Acknowledgments).

Stages in the Archaeological Exploration of Thisve, the Thisve Basin, and the Bay of Domvraina

As elsewhere, archaeological exploration was preceded by an antiquarian phase in which educated travelers, collectors, or emissaries, notably William Leake, identified the site of Greco-Roman Thisve at the village of “Kakosi” (renamed Thisve in the twentieth century), as well as a monumental “dam” described by Strabo and Pausanias which extends from Thisve across the Basin. In addition, the Italian Pomardi  recorded in detail, between 1804 and 1806, a rock-cut burial chamber at one of Thisve’s ancient necropoleis, and the Austrian Archduke Ludwig-Salvator von Habsburg, before 1876, accurately illustrated the Basin and Bay. By the 1880s academic Classical Archaeology had struck Thisve under the banners of historical topography and epigraphy, damaging Byzantine monuments in the search for immured Greco-Roman inscriptions, and of architectural history, which by 1958 had produced a selective and idealized—as opposed to an actual-state—ground plan of part of one phase, probably Hellenistic, of Thisve’s fortifications.

In the 1960s to 1980s, stratigraphic excavation and/or actual-state site planning were carried out in and around the Basin at the sites of Kastro Khostiôn, Vathy, Kastro, Alyki (ancient Siphai/Tiphai), Kouveli, and Diporto/Makronisos, while intensive artifact-sampling strategies were applied across the Basin and Lower Acropolis of Thisve. The Basin’s and Bay’s spectrum of settlement sites of all periods including, prominently, the early Byzantine, was studied. However, study of the multiphase monumental archaeology of Thisve itself did not advance, and the significance of the Basin’s and Bay’s total settlement pattern, including its relationship to the ancient “dam” and other potentially connected water-management installations, was not addressed. Specifically, for the Byzantine phase or phases of the settlement pattern, the significance of the Bay’s offshore sites was considered, but not that of the total spectrum of sites excavated and surveyed in the 1960s–1980s. Furthermore Byzantine, medieval Western, and postmedieval historical sources were not exploited, nor was the potential relevance of a burgeoning bibliography concerning Byzantine and Frankish Central Greece explored.

The New Research Program: Archaeological, Historical, Palaeoenvironmental

The current project, initiated in 2004 after an exploratory phase, has sought to integrate four things:

  1. the older archaeological surveys of the Basin and of sites in and around the Bay (overwhelmingly Byzantine),
  2. a new comprehensive survey of the site of Thisve (preclassical–post-Byzantine),
  3. selective recording of the Basin’s preclassical(?)–post-Byzantine monumental archaeology of water management, and
  4. new studies of the Holocene evolution of the Basin.

The Byzantino-Frankish material will be discussed in the light of a comprehensive survey of the sources for the history of the region that has already been compiled by the principal investigator. The overarching aim is that the new (and renewed) fieldwork will, together with the published excavations at the Basin’s and Bay’s other major sites (Kastro Khostiôn, Kastro, and Siphai), constitute a case study that, for the Byzantino-Frankish millennium, will complement the findings of surveys conducted in other parts of Boeotia, but at the same time without neglecting other eras. Around the Basin and Bay for instance the bigger picture indicates that we have a diverse range of early Byzantine successors to the ancient polis, and the signs of a subsequent medieval revival of towns, emporia, and skalai, significant phenomena that the region’s other intensive surveys have not adequately captured or have not captured at all. Historical research by the principal investigator (see Publications) has identified the Basin’s principal Byzantine site (at the now re-named “Thisve”), the Bay itself, and one or two of its littoral sites in Byzantine, western medieval, Ottoman, and post-Byzantine sources. The principal site had by the ninth century acquired the name Kastorion, the medieval Greek name for the purple dye extracted as an imperial and fiscal monopoly from the murex purpurea group of marine molluscs, the shells of which constitute a large “industrial” site at Thisve (as reported in the nineteenth century). The site’s wealth of exposed diagnostic Byzantine and Frankish structures ensures that survey techniques have much to offer within the multistranded approach outlined above.

In addition to surveys of the settlement areas, accurate recording of the ancient “dam” is essential. A “medieval” phase was perceived by some travellers and archaeologists, but it also significantly predates its Principate-era mentions by Strabo and Pausanias. The one study, by the water engineer Eberhard Knauss, offers no model of the Basin’s hydrology/palaeohydrology, or of the operation of other environmental factors. The nature and extent of changing configurations of sediments, water inflows and water bodies—all reflected in the landscape, the distribution of sites, relationship of sites and monuments to sediments (perhaps too in the periodicity and scales of construction)—deserve to be factored into this case study for all periods.

Fig. 3: Ancient Thisve/Byzantine Kastorion: Pre-Classical to Byzantine and Frankish features (Dunn 2008–2009)
Fig. 3: Ancient Thisve/Byzantine Kastorion: Pre-Classical to Byzantine and Frankish features (selected).


  1. The integration of older surveys. Byzantine archaeologists Professor Tim Gregory and Professor Bill Caraher have digitized the written and photographic archives of the Ohio State University Thisve Basin Survey and Corinthian Gulf Islands Project (of which Gregory is principal investigator); checked the surveys’ samples in the light of the recent literature; checked their mapping of sites; and entered their survey units into the new project’s digitized base map of the Basin and Bay.
  2. After the evaluation phase, electronic distance measurers (EDMs) were used to insert into our digitized versions of the Greek army’s 1:5000 series of maps the topographical footprints of in situ exposed diagnostic features from the preclassical to Byzantino-Frankish periods, and of selected types of post-Byzantine–early modern features (three types of traditional productive and storage installations). We included, however, only the topography of ancient Thisve’s many rock-cut tombs. After we cleaned vegetation, EDMs were also used to create ground plans of all preclassical to Byzantino-Frankish monuments, selected elevations, and the same for selected types of later structures. They were also used to fill some gaps in the cartographic base. Ten monuments were selected for recording using a High-Definition Three-Dimensional Laser Scanner, which captures volumes and surfaces (vertical and horizontal) in their spatial relationships. Linear architectural outputs of all kinds can be generated including actual-state models viewed from any angle. The sites of four Byzantine monuments that stand partially preserved in open terrain, which were probably once inscribed within building complexes such as urban monasteries, were surveyed using one or both of the standard geophysical technologies (Resistivity and Magnetometry). In addition “Geophysical Survey Area 1” on the general site plan was subjected to sherd counting using its 50 x 50 cm grid as recording units for a potential control upon the geophysical readings.
  3. The integration of the surveys of the urban site and of the Basin, and greater consistency and higher resolution in the recording of relief, including the setting of Thisve’s “dam,” eventually made the variable accuracy of the army’s 1:5000–scale maps problematic, especially their inadequate record of pronounced relief within Thisve, and of the Basin’s very attenuated relief.  So a differential (high-resolution) GPS was used to create digital elevation models of Thisve and relief features around the “dam.” It was also used to capture the volumes of important features of Thisve’s Upper Acropolis, which were inaccessible to the laser scanner, and to create a topographical map of Thisve-Kastorion’s unrecorded fourth harbor-side site, Agios Ioannis. This approach to the site of Thisve-Kastorion integrates high-resolution mapping, architectural planning, geophysical exploration of buried architecture wherever possible, intensive artifactual survey of the only large exposed and high-visibility area of the site (the Lower Acropolis), detailed descriptions and photography of all Late Roman, Byzantine and Frankish features, inclusion of post-Byzantine features selected for their relevance to the themes of the overarching publication, and new supplementary records of aspects of the site’s preclassical to Hellenistic exposed features.
  4. In addition to the integration of the new survey of Thisve with the older artifactual survey of the Basin, there is a parallel, Basin-wide palaeoenvironmental survey that studies the archaeology of water management around the Basin. Exposed sections of the ancient “dam” have been planned (as yet incomplete), as well as the attenuated bed of a river that flowed toward it from the east: the Telmessos of ancient texts. A pilot series of shallow cores was taken for fossil pollen in seasonally flooded areas, but without useful results for the history of the landscape. One pilot series of sedimentary samples was taken from a deep man-made cutting (one of three scientifically valuable deep cuttings) by the “dam,” in order to run tests for evidence of long-term inundations of the Basin. And a survey is under way of the Basin’s vegetation up to 1,000 meters, to characterize this key aspect of Thisve-Kastorion’s naturally defined terrain.
  5. The comprehensive palaeoenvironmental survey will, it is hoped, take place in 2010–2011.

The Archaeological Survey: Provisional Findings


Fig. 5: Thisve/Kastorion, the Lower Acropolis (Dunn 2008–2009)
Fig. 5: Thisve/Kastorion, the Lower Acropolis: Ancient, Byzantine, Frankish, and post-Byzantine features.


Our survey of the unpublished Upper Acropolis revealed five or six phases of fortifications, and indicates that the diagnostically early Byzantine one was the first true enclosed citadel. We identified:

  1. fragments of a preclassical phase of “Cyclopean” masonry
  2. fragments of preclassical “Lesbian” masonry
  3. fragments of one or two classical and/or Hellenistic towered enceintes immured within the early Byzantine phase
  4. a massive phase of lime mortar-bonded opus incertum and spolia construction (typically late Roman/early Byzantine) encircling the hilltop including a forewall on the most vulnerable side, a typical feature of early Byzantine fortresses in the Balkans
  5. a middle Byzantine or Frankish tower inserted into the southwest corner of Phase (d) in masonry distinguished  by liberal use of terra-cotta fragments in interstices and as leveling courses.

There are now ground plans and, wherever they can be extrapolated from the GPS survey, top plans and elevations.

Our survey of the schematically published Lower Acropolis revealed details of the Classical/Hellenistic plan not reflected in the idealized ground plan published by F. Maier (“Die Stadtmauer von Thisbe,” Mitteilungen der Deutschen Archäologischen Instituts: Athenische Abteilung 73 [1958]: 17–25). The laser scanner was used to record internally and externally the middle Byzantine or Frankish tower with vaulted chamber and platform that was constructed on the ruins of the tower at the SE corner of the ancient enceinte. The medieval tower does not communicate with the ancient enceinte and so is typologically similar to the freestanding “Frankish towers” of central Greece whose interpretation is such a fertile topic of discussion. We now have a completely new ground plan of the Lower Acropolis, and elevations, top plans, and volumetric surveys of sections of it, chosen to illustrate all preserved design features. All exposed features of all periods situated within the walls have also been recorded (for which see below).

The Lower town: fragments of the classical to Hellenistic walls were traced around the modern village, including two well-preserved towers, but no evidence of post-Roman urban defenses were visible. The likelihood is that the early Byzantine acropolis (Upper Acropolis) was the only true post-Roman fortification. The Lower Acropolis is not enclosed.

Churches and Chapels

Fig. 6: Agios Loukas: Ground plan (all phases) after cleaning (Dunn 2008–2009)
Fig. 6: Agios Loukas: Ground plan (all phases) after cleaning; one of ten laser-scanned monuments.
Our survey found archaeological evidence of ten Byzantine, medieval, or early postmedieval churches and chapels. Archduke Ludwig Salvator described another, which can be located under a modern church, and the American classical archaeologists Rolfe and Tarbell reported the existence of “twenty-three” ruined Byzantine churches in 1889. Certainly the villagers report that other ruined churches were demolished before the Second World War, and they report identifiable acts of demolition that occurred much more recently. Numerous functioning modern chapels may stand on the sites of older churches, and are certainly believed to do so by the people of Thisve. But there are no relevant diagnostic exposed features.

Of the ten recognizable sites, all were planned using EDMs, while six were also laser-scanned, four surveyed geophysically, and one recorded using the differential GPS. We appear to have:

  1. one early Byzantine basilica (extramural);
  2. one complex, possible “transitional” early middle Byzantine church with later phases (Agios Loukas), almost certainly the episcopal church of Kastorion, whose dedication probably evokes Osios Loukas (entitled “Agios Loukas” in an eleventh-century fresco in the crypt of his nearby monastery);
  3. two inscribed cruciform, probably originally domed, triple-apsed churches of the middle Byzantine type;
  4. one single-aisled chapel without apse that closely resembles an excavated Frankish chapel at Aliartos in Boeotia;
  5. two single-aisled chapels with apses, Byzantine, medieval, or early post-medieval; and
  6. three sites of destroyed chapels or churches.

Dwellings (?)

On the Lower Acropolis, between the Byzantino-Frankish tower and the diagnostically Frankish chapel, the spolia-built foundations of numerous rectangular buildings (“Post-Roman buildings” on the plan “Lower Acropolis 2008”) were planned: mostly door jambs and corner posts. There is no trace of the other materials used. The same use of spolia occurs at the chapel, where however it is associated with lime mortar-bonded rubble masonry. This area was targeted by Professor Gregory’s artifact survey of the Basin. He reported finding glazed Byzantino-Frankish sherds there.

The Purple Dye-Extraction Site

Extensive deposits of broken shells, truncated by early modern structures and modern houses, and defined on the south side by the escarpment of the Lower Acropolis, but still measuring 65/70 meters on one axis and 30 meters on the other, were plotted and sampled. All are murex purpurea. The predominant type found is Hexaplex trunculus L, which produces the best-quality purple dye. Their transportation 5 km inland from the sea to a processing site overlooked by both the Byzantine acropolis of Kastorion and the Byzantino-Frankish tower is strongly indicative of a controlled industry, and indeed of the imperial monopoly of purple-dye production and consumption. Thisve is potentially a valuable site for the documentation of the Byzantine enokhê kogkhylês.


Ancient Thisve had two substantial rock-cut necropoleis, one abutting the west side of the city walls, the other abutting their south side. These were presented at the 9th International Congress of Byzantine Studies by Professor Kominis as “Palaeochristian,” but there is no archaeological proof of this dating. The tombs’ locations have been recorded topographically.

Post-Byzantine Kastoriya-Kakosi

Fig. 4: Post-medieval productive installations of the village of Kastoriya-Kakosi (Dunn 2008–2009)
Fig. 4: Post-medieval productive installations of the village of Kastoriya-Kakosi. One of the eight located lênoi.

Monumental Threshing Floors

The positions and outlines of fifteen cobbled threshing floors (alônia) have been recorded. They form two complexes on the western and eastern edges of the village, on escarpments that overlook the village. The complex within and around the Lower Acropolis is served by a cobbled way. Several more floors are probably buried under recent intrusive waste dumps at both sites. We are clearly dealing with controlled sites of “conspicuous surplus extraction.” Their sizes and numbers will inform discussion of the Basin’s wheat production in the light of the published Ottoman and early modern fiscal data.

Monumental Wine Fermenting Vats

So far the locations of eight monumental wine fermenting vats (lênoi) have been recorded, one of which was laser-scanned externally and internally. As with the alônia, the volume of premechanised production (an archaeologically based minimum only) will be of interest, given the topographically and pedologically defined space that is the Thisve Basin, and can be juxtaposed with fiscal data. At the same time, these are sites of “conspicuous” production and storage, decorated with shallow niches that once contained inscriptions and religious items. And like the alônia they are major elements of a postmedieval heritage that is threatened with destruction as much by neglect as by willful acts.

Thisve Basin


A tower reported in the area by nineteenth-century travelers, who did not provide useful indications of topography or date, was found at a spot in between two of Thisve-Kastorion’s small ports, identified by Archie Dunn as the “Vathys Limên” and “Iôannitzê” of middle Byzantine texts. It was clearly situated to overlook both maritime sites while being much closer to Agios Ioannis. Its location and Classical/Hellenistic design are good indicators that both ports were functioning in Antiquity. The structure has been robbed to its foundations. Such freestanding towers were sometimes reconditioned in the Byzantine or Frankish periods, but there is no archaeological trace today (of any kind) of its re-use. It has now been recorded topographically and architecturally with both EDM and GPS.

Monumental Cistern

At Agios Ioannis, originally at the water’s edge, a monumental freestanding cistern was recorded topographically and architecturally. It is a fine example of the late Roman and Byzantine tradition of cistern design and construction, which continued into early modern times. It illustrates the infrastructure of a medieval port and deserves to be registered and recorded in its own right as part of the post-medieval heritage.


Dunn, A. “Historical and Archaeological Indicators of Economic Change in Middle Byzantine Boeotia and Their Problems.” In Β’ Διεθνές Συνέδριο Βοιωτικών Μελετών, 2:755–74. Athens: Hetaireia Voiōtikōn Meletōn, 1995.

Dunn, A. “The Rise and Fall of Towns, Loci of Maritime Traffic, and Silk Production: The Problem of Thisve-Kastorion.” In Byzantine Style, Religion and Civilization: In Honour of Sir Steven Runciman, ed. E. Jeffreys, 38–71. Cambridge: Cambridge University Press, 2006.

Dunn, A. “The Collaborative Survey of Thisve-Kastorion and Its Natural Harbours (2004-5).” Bulletin of British Byzantine Studies 32 (2006): 33–39.

Dunn, A. “The Survey of Thisve-Kastorion (The Urban Site: 2006).” Bulletin of British Byzantine Studies 33 (2007): 35–39.

Dunn, A. “Thisve-Kastorion: Town, Territorium and Loci of Maritime Traffic (Report on Fieldwork Conducted in 2007).” Bulletin of British Byzantine Studies 34 (2008): 33–40.

Dunn, A. “Thisve-Kastorion: Town, Territorium and Loci of Maritime Traffic (Report on Activity in 2008).” Bulletin of British Byzantine Studies 35 (2009): 37–46.


Dr. Eugenia Gerousi, Director and Dr Nikos Kondogiannis, Assistant Director of the 23rd Ephoreia of Byzantine Antiquities; Professor Vasilis Aravantinos, Director of the Archaeological Museum of Thebes; the staffs of the 23rd Ephoreia of Byzantine Antiquities and of the 9th Ephoreia of Prehistoric and Classical Antiquities; Charikleia Koilakou, Director of the 1st Ephoreia of Byzantine Antiquities; the Director and staff of the British School of Archaeology, Athens; the Mayor of Thisve and Domvraina; the Harbour Master’s office of Thisve; the staff of the Industrial Zone of Thisve; Professors Tim Gregory and Bill Caraher of the American School of Classical Studies at Athens; the team, Dr. Chrestos Anagnostou (sedimentologist and malacologist), Dr. Margaret Atherden (botanist and palynologist), Marylin Cassedy (assistant), Kevin Colls (GPS, EDM, laser scanner), Christopher Hewitson (laser scanner), Dr. Photeini Kondyli (Geophysicist), Michael Lobb (laser scanner), Christopher Mavromatis (EDM), Caroline Sturdy (technical assistant); Dr. Tim Van der Schriek (University of Newcastle-upon-Tyne) for joining us to assist in design of the final phase of the palaeoenvironmental survey of the Basin; Harry Buglass and Graham Norrie, technicians of the Institute of Archaeology and Antiquity, University of Birmingham; and for grants in either or both academic years 2007–08 and 2008–09, Birmingham University, the British Academy, the British School of Archaeology (Athens), Dumbarton Oaks, the Institute for the Study of Aegean Prehistory, and the Russell Trust.