This article is available at the URI as part of the NYU Library's Ancient World Digital Library in partnership with the Institute for the Study of the Ancient World (ISAW). More information about ISAW Papers is available on the ISAW website.

Except where noted, ©2015 The Authors; text distributed under the terms of the Creative Commons Attribution License
Creative Commons License

This article can be downloaded as a single file

ISAW Papers 10 (2015)

Preliminary Report on Early Byzantine Pottery from a Building Complex at Kenchreai (Greece)

Sebastian Heath, Joseph L. Rife, Jorge J. Bravo III, and Gavin Blasdel

Abstract: This paper presents the results of preliminary study of Early Byzantine pottery from a large building near the waterfront at Kenchreai in southern Greece. Kenchreai served as the eastern port of Corinth throughout antiquity. The building was first excavated in 1976 by the Greek Archaeological Service, and it has been investigated since 2014 by the American Excavations at Kenchreai with permission from the Ministry of Culture under the auspices of the American School of Classical Studies at Athens. The pottery is characterized by the presence of many Late Roman Amphora 2 rims as well as stoppers and funnels. This indicates that the building had a role in the distribution of regional agricultural products during its final phase, which is dated to the very late sixth or early seventh centures A.D. by African Red-Slip and Phocaean Red-Slip tablewares. A wide range of lamps, glass vessels, and other small finds has also been recorded. Results to date are preliminary but ongoing work may allow further precision as to the chronology and use of this building.

Library of Congress Subjects: Kenchreai (Greece); Pottery, Roman; Pottery, Byzantine; Economic history--Medieval, 500-1500.



In 1976 the Greek Archaeological Service under the direction of Kalliopi Kristalli-Votsi, then Ephor of the former 4th ΕΠΚΑ of the Argolidocorinthia, and her assistant, Elisavet Chatzipouliou, excavated a large building at Kenchreai, the town that served as the Aegean port of the major city of Corinth. The building is located very close to the ancient harbor, alongside the National Highway to Epidavros (figs. 1 and 2). Attention was drawn to the site following plans for construction by the landowners at the time, whose surname Threpsiades now serves as the name of the excavated area. The Threpsiades property is a declared archaeological site under the oversight of the Hellenic Ministry of Culture, and the finds recovered in 1976 are stored in the Isthmia Museum.1

As is usual for excavation in a Mediterranean urban environment, pottery makes up the overwhelming bulk of the assemblage. This material is remarkable for a large number of regional and imported amphoras, the latter in much smaller amounts, along with numerous stoppers for the regional vessels and a small group of funnels. The pottery is the main focus of this preliminary report. There is, however, also a rich variety of small finds, including a large group of Late Roman to Early Byzantine glass, including cylinder-blown window panes, trail-decorated flasks with funnel- to U-shaped mouths of a type common in the Byzantine to Islamic Middle East, and possible evidence for glassmaking. These diverse and copious finds suggest that the building complex on the Threpsiades property played a role in the movement of regional agricultural products, and perhaps of consumer goods as well, during the very late sixth to early seventh centuries A.D.2 They provide important new evidence for local vitality, long-distance exchange, and dynamic communication in the northeastern Peloponnese during the Early Byzantine period.

Figure 1: Location of Kenchreai within the Corinthia. Reproduced courtesy of the American Excavations at Kenchreai under the Creative Commons By-ND-NC 4.0 license.

A summary of the work in 1976 by the Archaeological Service work was published in 1984.3 That report noted the presence of many amphoras and provided basic observations about the building and a massive destruction event involving burning and structural collapse. To date, however, no full study has appeared. In 2013 Ms. Kristalli-Votsi transferred her rights to the site and its finds to Joseph L. Rife, Director of the American Excavations at Kenchreai. All work at the site and on the finds takes place with the permission of Ministry of Culture and the oversight of the Corinthian Ephoreia under the auspices of the American School of Classical Studies at Athens. In May 2014, Sebastian Heath and Phyllis Graham, with the assistance of Gavin Blasdel, began study of the material in the Isthmia Museum. Rife and Jorge J. Bravo III, Co-Director of the American Excavations at Kenchreai, began comprehensive re-evaluation of the building on the Threpsiades property in June 2014. The results of these new investigations offer a firmer basis for determining the date of activity in the complex than was previously published.4 Moreover, preliminary quantification of amphora rims provides evidence for assessing the role of Early Byzantine Kenchreai in the regional and Eastern Mediterranean economy. The numbers appearing below are drawn from processing of 93 crates of pottery that hold well over 50,000 sherds. Quantification was done in 2014 and study continued in 2015. While this ongoing work is the beginning of a multi-year process that will lead to full publication, we believe that a timely announcement of our research may be useful to readers and that any responses will benefit us as we move toward more complete results.

The Site of Kenchreai

The port town of Kenchreai lies at the western end of the Saronic Gulf at the point where the shore is easily reached from Corinth, which lies some 10 km inland and to which Kenchreai was connected by road during the Roman era. The long, irregular ridge of Mount Oneion extends inland just south of the site, and the coast immediately to the north has low cliffs that would have made landing and loading or unloading trade-vessels difficult. Between these coastal stretches, the small, curved bay and beachline today provide access to the water, and movement inland is relatively easy. It is this landward topographic aspect that contributed to the placement of Corinth’s eastern harbor, a role that Kenchreai may have served throughout antiquity. While Hellenistic and earlier remains at the site are not well preserved, excavations begun in 1963 under the general direction of Robert Scranton of the University of Chicago for the American School revealed the extensive remains of the Roman port.5 Two massive breakwaters erected probably under Augustus facilitated the reception of maritime traffic and sustained seaborne commerce for centuries. The moles defined clear northern and southern limits of the port and turned an exposed landing into a deep, protected harbor with good anchorage and a long quay.

Figure 2: Plan of Kenchreai with the Threpsiades site circled. Reproduced courtesy of the American Excavations at Kenchreai under the Creative Commons By-ND-NC 4.0 license

The literary testimony for Kenchreai during the Early to Middle Roman periods attests to local prosperity, social, cultural and religious diversity, and a bustling economy. Strabo (Geographies 8.6.22) wrote of Kenchreai that the Corinthians, “use [it] for trade from Asia.” Ovid (Tristia 1.10.9) reported that he boarded the vessel Minerva there and then disembarked at Tempyra on the Thracian coast before proceeding on foot to his exile in Tomis. Kenchreai is briefly mentioned in the New Testament, first as an early stopping point in Paul’s journey from Corinth to Syria via Ephesus (Acts 18:18) and then as the home of the deaconess Phoebe (Romans 16:1), whom Paul commends to the Roman church. In his brief description of the port-town, the second-century travel writer Pausanias (Description of Greece 2.2.3) mentions sanctuaries and a bronze statue of Poseidon at the harbor.6 Kenchreai is also the setting for the last book of Apuleius’ Metamorphoses (Book 11), in which the protagonist Lucius is initiated into the local cult of Isis. These passages illustrate Kenchreai as a well-connected port and a prosperous community during the first and second centuries A.D.

It was the report of Pausanias and the visible presence of ancient architecture that motivated the start of excavation in 1963. The initial publication of the site aimed to assimilate the extant architectural remains to the temples of Aphrodite and Isis cited by Pausanias, but subsequent scholarship has called these identifications into question, and they are not now widely held.7 The early phase of the American Excavations in the 1960s also uncovered extensive architectural and artifactual evidence for the Roman port’s commercial life. More recent fieldwork has focused on the necropolis and residential quarter on the ridge northeast of the harbor.8

Urban development continued at Kenchreai during and beyond the third century. The local community witnessed substantial renovation and expansion of the portside structures during the fourth century. In the late fourth century, one or more seismic disasters led to the abandonment and filling of a fountain court on the south mole. This space was temporarily being used for the storage of crates with over 120 glass panels in opus sectile depicting harbors, Nilotic scenes, and cultural celebrities.9 The difficulties indicated by the loss of these panels, however, were not the end of Roman Kenchreai. A church was built on the south mole around the late fifth or early sixth century, and another early church is located roughly 1 km up the coast in the port’s northern suburbs. These structures did not stay in use for very long and seem to have fallen into disrepair by the late sixth or early seventh centuries, perhaps around the time of the unknown destruction event at the Threpsiades complex under consideration in this article. The immediate circumstances of Early Byzantine Kenchreai’s further decline in the seventh to ninth centuries are unclear, as are the circumstances that lead to a renewal of activity in the Middle Byzantine period. Today the toponym Kechries (Κεχριές) refers to a small village lacking any municipal identity or business and consisting largely of weekend and summer homes.

Towards a Chronology of the Threpsiades Complex

The Threspiades property is located ca. 35 meters due west of the mid-southwest part of the modern public beach, which represents roughly the shoreline of the Roman harbor. It is situated at the northwest corner of the intersection of the highway to Epidavros (Ethniki Odos 10) and the road through the village of Kechries (Odos Apostolou Pavlou).10 Although constrained by existing property lines and modern roads, the excavators in 1976 uncovered multiple large rooms flanking wide halls and an ornate peristyle, all enclosed in a trapezoidal arrangement. The building had a second story indicated by a well-built staircase. This spacious plan and its interior decor suggest that the complex was not built solely as a work-space; perhaps its original use was as a lavish private residence or a monumental public building. In any case, it is clear that the building had developed mainly into a work-space in its final phase before abandonment and its gradual burial under colluvial sediment.

At the current stage of study, the date of the earliest phase of the building remains uncertain, though either Middle or Late Roman seems most likely. Moreover, while it is clear that the fabric and use of the building were disrupted by a catastrophic destruction, stratigraphic study of the Threpsiades complex has not yet revealed when this destruction happened. Two scenarios seem most likely. In the first, destruction struck the site at some point in the sixth century and damaged the impressive private or public complex, after which point it was reoccupied for more utilitarian or industrial use that ended in the seventh century.  In the second, destruction struck the site at some point in the early seventh century and brought an end to the utilitarian or industrial final use of a formerly impressive private or public complex. It is also possible that the area suffered multiple episodes of destruction.

Although a full and precise chronology of the Threpsiades complex remains elusive, current investigation allows the final phase of activity in the building to be assigned confidently to the very late sixth to early seventh centuries. This date is mainly based on red-slipped tablewares. A nearly complete African Red-Slip (ARS) Hayes form 105 (Fig. 3) with no interior stamped decoration gives a strong terminus post quem of 580.11 Another large fragment of a Hayes 105-type base along with similar, less-well preserved base sherds strengthen this chronology, as do small ARS Hayes form 99c rim sherds. In addition, three well-preserved ARS Hayes II lamps with late discus designs are not inconsistent with the ARS table vessels. We have also recorded a Phocaean Red-Slip (PRS) Hayes form 3f of the sixth century, five PRS Hayes form 10c rims (PRS) and nine Hayes form 10a rims (examples illustrated in Fig. 3).12 While the imported tableware assemblage only supports limited precision on the order of decades the ARS does suggest that the deposit closed by the early seventh century. The material from multiple deposits at Corinth published under the rubric “Assemblage 4” within a sequence of Late Roman material is relevant here. ARS forms 107 and 109 are present at Corinth but not identified in the Threpsiades assemblage, which can suggest a closing at Kenchreai before 600.13 Among more recently published early seventh-century assemblages, the material from the so-called “Bishop’s House” at Nicopolis is a useful parallel, particularly because ARS Hayes form 105 also contributes to establishing its date, as does a coin of the emperor Phocas (602-610).14 Moreover, discussion of the absence from Nicopolis of ARS Hayes form 109, which as noted also does not occur in the Threpsiades complex, further suggests that overly precise dating from ceramic forms is problematic in the face of competing chronologies and uncertain stratigraphy.15

Figure 3: Phocaean Red-Slip Hayes 10a (KTH 1), Phocaean Red-Slip Hayes 10c (KTH 3), and African Red-Slip Hayes 105 (KTH 186). Reproduced courtesy of the American Excavations at Kenchreai under the Creative Commons By-ND-NC 4.0 license.

Care is warranted in establishing the chronological limits of the total assemblage from the building and in assessing how much of it was associated with the final phase of activity. Some but not all of the original recording system for the stored artifacts is recoverable. There has been mixing of material between plastic storage crates—this has proven that the bottoms of such containers break if overloaded—and not all tags were readable at first inspection. Furthermore, we stress that the data appearing in this article were collected during an initial survey of the material. Almost all sherds remain unwashed, so that full processing of the assemblage will reveal more detail. Future clearing of spoilage at the site will undoubtedly contribute more evidence to the finds already stored at Isthmia. Finally, it may be possible to assess more exactly the stratigraphic associations of the finds, in which case our understanding of the depositional sequence and the building’s chronology may shift.

Despite these concerns, we can now characterize certain basic features of the assemblage. It displays a striking degree of consistency. As will be discussed below, the regionally produced Late Roman Amphora 2 (LRA2) predominates. Except for a small number of Hellenistic and perhaps earlier sherds and a small number of Early and Middle Roman sherds, we agree with the original excavator that much of this material is associated with the final phase of occupation. Thus, as this important group of pottery is studied and further cited, it should be kept in mind that it may include a small but uncertain number of sherds that predate Late Antiquity. These would affect the accuracy of any quantification of sherds dated to the late sixth and early seventh centuries.

Preliminary Quantification of Amphoras

Late Roman transport amphoras form the most abundant component of the ceramic record in the Threpsiades complex. Preliminary quantification (fig. 4) of 501 rims from the well-known series of Late Roman Amphoras 1 through 5/6 provides a firm basis for situating Kenchreai in both regional and Eastern Mediterranean economic networks.16 It is the case, however, that essentially all the quantification undertaken at this stage of study has been by rim counts. In his review of quantification methods, J. T. Peña described such data as “not statistically valid.”17 While recognizing the shortcomings of this approach, we do believe that the overall impression offered by Figure 4 and Table 1 is useful in discussing the relative amounts of these amphoras present in the complex during Late Antiquity. Moreover, the first two seasons of work have identified amphora forms other than these most widely recognized ones, which will be briefly discussed below.

Figure. 4: Quantification of 501 Late Roman Amphora 1, 2, 3, 4 and 5 rims
Table 1: Quantification of 501 LRA 1-5 rims
Type Count Percent Origin
LRA2 361 72.1 Aegean
LRA1 51 10.2 Cilicia/Cyprus
LRA5 46 9.2 N. Palestine
LRA4 39 7.8 S. Palestine/Gaza
LRA3 4 0.8 SW Anatolia

Large and Small Late Roman Amphora 2

The round-bodied, high-necked form Late Roman Amphora 2 makes up approximately 72% of the rim count at the Threpsiades site.18 Slane and Sanders have appropriately warned against simplistic usage of this term and we intend it here in a very generic sense.19 As they note, multiple Aegean regions produced this form, a point reinforced by Reynolds’ analysis of assemblages at Beirut and Butrint as well as by other studies.20 In the context of this article it is important that the Argolid, which is linked to Kenchreai over the mountainous passes of the southeastern Corinthia, is recognized as a major producer of LRA2 vessels. A kiln-site producing the form near Porto Cheli in the southern Argolid is dated by the presence of ARS, PRS and a coin of the emperor Phocas (602-610).21 That region’s workshops produced vessels with an identifiable pinkish fabric and a beige surface.22 The fabric of many of the LRA2 rims, other diagnostics, and body sherds from the Threpsiades excavations is likely of or closely related to this Argolid variety, which has also been identified at Corinth. A further indication that Kenchreai is within one of the multiple active regions of LRA2 production and use comes from the Eastern Korinthia Archaeological Survey (EKAS), whose members recognized that combed body sherds were among the most abundant Late Roman artifact categories.23 When one considers LRA2 broadly, the recent publication of kilns at Dilesi in Boeotia adds to the increasing firm evidence that production was widespread.24

There is diversity among the LRA2 vessels from the Threpsiades complex. Among those rims measured, we have noted a range of maximum exterior diameters of 9-12 cm. Very large vessels have rim heights just over 5 cm and neck heights just over 9 cm. Rim heights of less massively constructed LRA2 vessels are just under 4 cm but the neck height remains around 9 cm. The rims on the largest vessels are often slightly offset from the neck. We have not refit body sherds, but we note the predominance of deep wavy combing. It is of interest that some LRA2 body sherds bear black characters or symbols. These markings are distinct from the red dipinti that are not uncommon on LRA2 vessels. These preliminary observations suggest directions for future study. In sum, the Threpsiades group of LRA2 amphoras is large enough to encompass differences, but those differences fall within a range that is well-known for the form.25

In addition to the many typical LRA2 amphora rims, we counted 222 rim sherds of distinctly smaller vessels. Figure 5 shows a comparison of the rim and neck profiles of the full-sized or “large” and “small” types. The smaller version has a less substantial rim that is of more even thickness, and there is often a distinct exterior demarcation as neck turns to rim. The distinction is subtle, presumably because the smaller version was very closely inspired by the larger, but familiarity makes the differences discernible, even between small rim and handle sherds. Further study may show that the “LRA2 Small” form is best considered a fractional container, meaning that it was produced at a regular and measured proportion of its larger mate.26

Beyond Kenchreai, smaller variants of LRA2 are known from Argos, Corinth and Pyrgouthi, among other sites.27 An approximate parallel for the small Threpsiades vessels appears in the late-sixth-century “Assemblage 3” at Corinth, which consists mostly of material from an underground chamber near the Baths of Aphrodite; the rim of that vessel appears thicker than ours in the published profile.28 There is an unpublished parallel at Athens, which is illustrated in an excavation notebook that the Agora Excavations makes available online.29 The Sikyon Survey Project has also published an LRA2 rim of small dimension. Other published profiles of smaller LRA2 vessels do not appear to be very close matches.30 A category of “LRA2-related” so-called amphorette known from Dichin, Bulgaria is very different.31 This brief survey may be enough to suggest that the Kenchreai “LRA2 Small” vessels and related forms are an LRA2 companion whose distribution far outside the Corinthia and nearby regions may be limited. At Kenchreai the form presumably had a role in surplus-oriented economic activities. At this early stage of study, however, we stress that the exact use of these small containers is unclear, and indeed they may have had multiple uses.

Figure 5: Comparison of full-size LRA2 (KTH 5) and (KTH 6) “LRA2 Small” profiles. Reproduced courtesy of the American Excavations at Kenchreai under the Creative Commons By-ND-NC 4.0 license.