Butrint 6: Excavations on the Vrina Plain: Volume 2 - The Finds

By Simon Greenslade

Butrint 6 describes the excavations carried out on the Vrina Plain by the Butrint Foundation from 2002–2007. Lying just to the south of the ancient port city of Butrint, these excavations have revealed a 1,300 year long story of a changing community that began in the 1st century AD, one which not only played its part in shaping the city of Butrint but also in how the city interacted and at times reacted to the changing political, economic and cultural situations occurring across the Mediterranean World over this period. Volume II discusses the finds from the Vrina Plain excavations. This volume provides an insight into how the Vrina Plain community lived, worked and ultimately died and includes chapters on the medieval and post-medieval ceramics from the excavations, analysis of the human and faunal remains, environmental evidence, Roman and Medieval coins, a detailed study of the small finds as well as a discussion of the glass including a report on a number of glass cakes, ingots of raw glass associated with glass working that were found during the excavations. The volume also reports on five lead seals dating from the late 9th to the 10th century, an uncommon find but one which when considered with the contemporary coins suggests that for 100 years the Vrina Plain was Butrint.

• Language: English
• Publisher: Oxbow Books
• Release date: Aug 4, 2019
• ISBN: 9781789252187

Book preview

Background to the site and excavations

Simon Greenslade

The Vrina Plain lies to the south of the city of Butrint, across the Vivari Channel which links Lake Butrint to the Ionian Sea (Fig. 0.1). The pastoral aspect of the low-lying area visible today is largely a result of the implementation of a state-run collective agricultural policy by the communist government of the 1960s and 1970s, based on a model developed by the Chinese, their ideological allies at the time. With the institution of state farms at the villages of Xarra and Vrina, woodland which had covered much of the plain was removed and a grid of large irrigation channels was dug across the plain in order to drain the marshy area, thereby creating a usable space for crops and animal grazing.

Up until recently, archaeological investigation of the plain had been limited. Luigi Maria Ugolini began excavations at Butrint in the 1920s and although he noted the presence of a number of Roman ruins on the Vrina Plain, due to the area being densely wooded and marshy no real attempt was made at investigating them in detail.¹ Further work during the 1960s and the 1990s by the Albanian Institute of Archaeology was restricted to a number of investigations focused around the standing remains, along with some surface survey, small-scale excavation trenches and the recording of finds made during further land improvement.² However, from 1995 the Butrint Foundation conducted an extensive field survey programme across this area. Through a combination of field-walking and geophysical survey, extensive remains of destroyed masonry structures and associated ceramics were identified.³ This led to the conclusion that this may have been a suburb of Roman Butrint and if so, that it could be evidence of the colony created by Julius Caesar in 44 BC and re-established by Augustus after his victory at Actium in 31 BC. As a result, since 2002 a major archaeological assessment has been undertaken on the Vrina Plain in order to test this theory and to understand the extent and character of the archaeological remains.

Figure 0.1. Aerial view of Butrint with the Vrina Plain beyond

Figure 0.2. The Vrina Plain excavations

The excavations

Initially the excavations were centred on one of the large drainage ditches that criss-cross the plain. The course of the ditch chosen had been deliberately diverted around some upstanding remains, consisting of a number of buildings in varying states of preservation. In all, an area along roughly 250 m of the ditch was investigated. Following on from these trial excavations, a number of areas were looked at in greater detail to gain a better understanding of the context for some of the partially exposed structures (Fig. 0.2).

The monuments

One of these areas focused on what appears to be a series of honorific column monuments, which seem to have been built during the 2nd century AD (Fig. 0.3). By the beginning of the 5th century AD, the northern, southern and western sides of the monument were enclosed by a series rooms, with access to the monument now restricted to the eastern side. Only the southern room was fully exposed and was found to consist of a rectangular structure with annexed spaces to the east and west of it. This building appeared to have undergone a complex sequence of rebuilds and alterations over a very short period of time, including at one point the addition of nine internal masonry piers or buttresses along with an internal division wall to support a second floor. At the beginning of the 6th century AD the area was altered again when a small apsidal-ended building was constructed to the northwest of the monument, built partially over the northern buildings.

Figure 0.3. Plan of the monument area excavations

Figure 0.4. Plan of the Temple mausoleum excavations

The Temple mausoleum

The second area investigated was a rectangular structure, located just to the southeast of the monuments (Fig. 0.4). Its proximity to the aqueduct had led initially to the interpretation of this structure being a bathhouse or cistern dating to the early Imperial period.⁴ However, cleaning and recording of the building revealed that the structure had been built upon a podium and had been accessed from the west, where the robbed-out impression of a staircase was found fronting the building. Internally, the impressions of five rectangular structures were found at its eastern end. Earlier excavations had recovered various sculptural pieces across the site that appeared funerary in nature. As fragments of human bones were also found within the backfill of the 1980s excavations, the building was reinterpreted as a mausoleum that had been built to look like a raised temple.⁵ Subsequent excavations along the sides of the mausoleum revealed further traces of broken sculptural pieces as well as in situ fragments of the lower marble facing of the building, indicating that this would have been an imposing structure. New excavations were also carried out to the west of the building where part of an earlier north–south-aligned road fronting the steps of the mausoleum was exposed; on the opposite side of the road a pavement was found fronting the colonnade of a portico that would have faced the mausoleum. Over time this western area appears to have been gradually altered: a series of crudely built structures made of rough clay-bonded stone and tile walls was found built over the line of the proposed portico. Fragments of masonry from the Temple were found in some of the walls of these buildings.

Domus area

The largest and most intense area of excavation, however, was centred on the upstanding buildings (Fig. 0.5). Covering an area roughly 65 × 45 m, these excavations – carried out using modern stratigraphic techniques – produced over a thousand contexts and over six hundred small finds, including almost five hundred coins. Large quantities of Roman and medieval pottery have also been collected from the excavations. Together, these results have enabled the creation of an important picture of a changing landscape from the 1st century AD to the 13th century AD.

Figure 0.5. Plan of the domus excavations

The suburb: mid-1st-3rd century AD (Figs 0.6 and 0.7)

The earliest evidence for occupation appears to date to the mid-1st century AD when a series of houses of varying sizes was constructed across the site, some of which fronted onto a series of roads which connected the suburb to the main road from Butrint to Mursia, located to the east. Rather than being a haphazard expansion, the settlement was laid out from the start with the houses divided into separate plots. Shops fronted onto the roads with the residential part of the house set behind. During the 2nd century the occupation appears to have expanded. However, by the 3rd century, the nature of the occupation changed as the area seems to have come under the control of one individual, who constructed a large peristyle house across the site. To allow for the construction of this large structure the earlier buildings were either demolished, as in the case of the buildings found to the west of the site, or they were integrated into the new build.

Figure 0.6. The Vrina Plain settlement (mid-1st to early 2nd century AD)

The domus: mid-3rd to mid-5th century AD (Fig. 0.8)

Central to this new complex was a large open courtyard. Measuring c. 26 m east to west by c. 16.50 m north to south, the courtyard contained an ornamental water feature comprising two pools, one built inside the other. The courtyard was surrounded on all four sides by a portico, originally floored with a series of mosaic pavements of geometric design.

The northern part of the Vrina Plain domus seems to have been the public area, centred on a large apsidal structure. This appears to have been an audience chamber where the owner would have received his clients as well as his business associates. Measuring 25.30 × 13.50 m, the room was orientated east to west with the apse at the eastern end of the structure and the main doorway opposite it in the western wall.

Figure 0.7. The Vrina Plain settlement (2nd to early-3rd century AD)

The eastern area of the building appears to have been the private part of the house, restricted to the owner’s most trusted friends and confidantes. The eastern portico gave access to a series of elaborately furnished rooms that utilised the earlier buildings located here. This seems to have involved blocking in a number of doorways as well as laying a sequence of new mosaic floors.

The full extent of this domus is unclear. From the geophysical survey it would appear that it could potentially continue to the south where a further, much larger, courtyard has been identified, which was surrounded by a portico and various ranges of rooms, including what appears to be another apsidal hall and a triple-apsed triclinium.

The house seems to have been occupied until the middle of the 4th century, at which point it appears to have been abandoned. The reason for the abandonment is unclear though it is possible that it may have been as a result of an earthquake. By the beginning of the 5th century the site was reoccupied but only for a short period and by the mid-5th century the house was once more abandoned. However, this period of abandonment was only temporary as by the beginning of the 6th century the house was reoccupied.

Figure 0.8. The Vrina Plain settlement (mid-3rd to mid-5th century AD)

The basilica: 6th century AD (Fig. 0.9)

The focus of the new occupation was religious rather than residential, as indicated by the construction of a north–south-aligned apsidal basilica, floored with a series of highly intricate mosaics.

The basilica, measuring c. 19.21 m by c. 15.85 m, was built out from the southern wall of the apsidal chamber, which necessitated the removal of part of this wall. Cut through a levelling deposit, which had been spread across the central area, the new structure was built partially over the northern and eastern porticoes of the late Roman house and extended into the courtyard, with the apse of this building built over the northern niche of the inner pool. Internally, the space was divided into a central area and two flanking aisles defined by a series of seven piers on either side of the central space. Covering the central space was a mosaic pavement separated by a chancel screen.

The foundation of the church does not seem to have been an isolated venture; rather, the entire area surrounding the basilica underwent a form of regeneration in this period including the construction of a small bath-house within the northern rooms of the Roman cistern.

The site was occupied until the middle of the 6th century at which point it appears to have been abandoned. There are signs that the building may have been partially destroyed by fire. For the next three hundred years the site was deserted but in the mid-9th century the buildings were reoccupied, becoming the manor-house or aristocratic oikos of a Byzantine official, possibly the commander of Butrint.

Figure 0.9. The Vrina Plain settlement (early to mid-6th century AD)

The oikos: mid-9th to mid-10th century AD (Fig. 0.10)

The focus of the new activity was centred once more on the area of the basilica and the surrounding buildings. The deposits that must have filled the abandoned building were cleaned out and the building repaired. The arcades of the southern extension, which had collapsed, were patched up and the openings in the nave-arcades were blocked in. The step in front of the main north entrance into the nave was extended over the mosaic and, due to an enduring weakness in the area of the sanctuary, the original chancel screen was replaced by a more robust barrier, which may have served as a diaphragm-wall to secure a weak point in the roof. In the narthex and the western chamber a number of blocking walls were also added.

The principal room of the new house was situated over the former narthex; this was supported on a series of posts, the holes for which were fire-blasted through the earlier flagstone floor. Meanwhile, the religious nature of the site became focused on the nave and sanctuary. As a result of the blocking-in of the arcades, the aisles were used for more practical activities. The eastern aisle appears to have been used as a workshop associated with a large pottery kiln, constructed in the small room that opened-off the aisle, while the western aisle seems to have become a storage space. Once more the surrounding rooms of the earlier domus were reoccupied, while the area of the infilled Roman pools became a cemetery, possibly for the community or retainers under the official’s jurisdiction.

Figure 0.10. The Vrina Plain settlement (mid-9th to mid-10th century AD)

The discovery of 51 Byzantine coins spanning the period c. AD 820–950, along with five lead seals and a ceramic assemblage showing a revival in trade with southern Italy, suggests that by the 10th century AD the household on the Vrina Plain had become a centre of regional, as well as international, administration, with an important local official, archon, living and working there. This role seems to have been maintained up to the mid-10th century AD, after which the building was abandoned, seemingly due to a rise in the water level.

With the abandonment, much of the superstructure of the building appears to have collapsed and any usable material was removed. Although there are indications of small-scale industrial activity continuing within some of the rooms, while the apse of the basilica possibly remained a focal point as a small chapel, generally this period of use was short-lived; from the later 11th century on the buildings appear to have been abandoned, with several layers of ‘black earth’ deposits building up over the remains. Knowledge of the site seems to have remained, however, as at least nine burials have been located cut through the ‘black earth’ deposits across the area as a whole.

This volume

Through the use of modern stratigraphic excavation and detailed interpretation of the archaeological sequences in relation to the structural remains, work so far undertaken on the Vrina Plain has identified 16 phases of activity covering the period from the mid-1st century AD up to the modern day (Table 0.1), an archaeological sequence as impressive and important as any recovered elsewhere in Butrint and in some aspects anywhere in the Mediterranean. The stratigraphic and structural sequence from the excavations has been detailed in Volume 6.1: Excavations on the Vrina Plain. The lost Roman and Byzantine suburb. The aim of this volume, together with Volume 6.3: Excavations on the Vrina Plain. The Roman and late Antique pottery from the Vrina Plain excavations, by Paul Reynolds, is to further this study by presenting the work undertaken on the many finds which the excavations produced and provide a complementary resource to work in tandem with Volume 6.1, the archaeological interpretation of the site.

Table 0.1 Overview of the development of the Vrina Plain settlement: Phases 1–16

The initial chapter by Joanita Vroom discusses the medieval and post-medieval ceramics from the excavations. Covering a period between the 7th/8th and 12th centuries AD, the majority of the material from a total of 2134 diagnostic sherds date to the middle Byzantine/Norman period in Butrint, approximately the 10th to 12th century AD. This is then followed by two reports on the coins recovered from the excavations. Sam Moorhead discusses the 511 ancient and early Byzantine coins ranging from the 3rd century BC to the 6th century AD, while Pagona Papadopoulou reports on the 59 coins covering the 9th to 17th centuries. Papadopoulou also reports on the five lead seals dating from the late 9th to the 10th century, which, as she notes, are not common finds in archaeological excavations, a fact that attributes special importance to the group presented here. The next contribution is by Angela Soler, Carolyn V. Isaac, Jared S. Beatrice and Todd W. Fenton who discuss the human skeletons. The skeletons cover two periods, with five individuals buried between the 4th to 6th century AD and a further 21 individuals interred between the 9th to 13th century AD. The two burial phases are looked at individually and then as a whole before being compared to nearby sites, in particular the 9th- to 13th-century AD burials from the Triconch Palace excavations in Butrint, to gain an enhanced perspective of life on the Vrina Plain through time. Following this is the study of the Vrina Plain small finds by John Mitchell. Comprising a range of materials, the small finds have been ordered by type; silver, lead, copper, iron, metal slag, glass, worked ivory and bone, amber, ceramic, stucco, semiprecious stone, stone, mosaics and inscriptions in stone and ceramic. All small finds have been listed with their context and individual small find numbers as well as a brief description of the context, its phasing and dating. This chapter is supplemented by a discussion by Pippa Pearce on the methods used to conserve the small finds, highlighting a number of important pieces. The glass finds are discussed next: covering all periods, this work was initially undertaken by Sarah Jennings and subsequently, after Sarah’s untimely death, by Karen Stark. In all over 2000 pieces have been studied and catalogued. As well as glassware, the Vrina Plain produced a number of glass cakes – ingots of raw glass associated with glass working. These and a number of glass tesserae have been studied by Nadine Schibille, who has looked at the compositional relationship of the cakes with the tesserae and some of the glass vessels. The faunal finds are discussed by Richard Madgwick. Comprising an assemblage of 11,461 bone fragments from more than 300 contexts ranging in date from the mid-1st century AD to the 13th century and beyond, this contribution represents a summary of the dietary practice, husbandry regimes and human–animal relations during the Roman and late antique/medieval periods on the Vrina Plain. As part of the excavations, an environmental sampling strategy was undertaken, the results of which are presented in the final three chapters of the volume. Rena Veropoulidou discusses the archaeomalacological (mollusc) data recovered and explains how this initial study has the potential to provide a better understanding of everyday life, diet and craft production in a medieval community in southeastern Europe. In addition to the shells extracted from bulk sediment flotation samples, a number of mollusc shells were hand-collected during the course of the excavations. These have been looked at by Matthew Law and Richard Madgwick and confirm Veropoulidou’s archaeomalacological results. Finally, the archaeobotanical evidence from the Vrina Plain settlement is discussed by Alexandra Livarda. Although the preservation of the remains was limited, making identification of species difficult, the data does provide an indication of the range of available food plants and allows an insight into the subsistence base of the inhabitants of the settlement.

Notes

1Ugolini 1927.

2Budina 1971.

3Bescoby 2003; 2007, 95–118; Chroston and Hounslow 2004, 64–75; Hodges et al. 1997, 211–14; Pluciennik et al. 2004, 47–63.

4Çondi 1988.

5Crowson and Gilkes 2007, 126–31; Gilkes, Hysa and Çondi 2013.

1   The medieval and post-medieval pottery finds from the Vrina Plain excavations

Joanita Vroom

¹

Introduction

In this preliminary report I present a selection of medieval and post-medieval pottery finds from the 2004–08 excavations on the Vrina Plain in Butrint.² In total c. 2070 contexts were excavated during the years 2002–08, of which approximately 90 contexts had mostly medieval pottery finds. The assemblage includes a number of wares that are described below, ranging from glazed fine wares to medieval amphorae, the study of which is still ongoing.

Some pottery quantifications

The assemblage of medieval and post-medieval pottery fragments from the Vrina Plain excavations consists of a total of 2134 diagnostic sherds with a total weight of c. 87 kg. Of this total, handle fragments make up the largest part of the ceramic assemblage (37%), followed by body fragments (33%), rim fragments (19%), base fragments (8%) and, finally, lid fragments (3%) (Plate 1.1, Graph 1).

Graphs 2 and 3 (Plate 1.1) show the relative quantities by count of various wares from the total assemblage. In Graph 2 one can notice, for instance, that coarse wares (‘CW’) represent 50% of the total (1014 fragments in total, of which 155 include wares of light utility),³ directly followed by amphorae (‘AMP’) with 45% of the total. It is interesting to see that table wares (‘TW’) represent only 2% of the total, while other vessels (‘OTH’) make up 1%, and finally special finds (‘SP’) 2%. It is clear from this graph that coarse wares represent the largest group from this period. The large numbers of coarse wares and amphorae can also be seen in the significant difference in the number of open and closed shapes. Closed vessels represent 98% of the total assemblage, whereas open vessels represent only 2% (Plate 1.1, Graph 3).

Regarding the chronology of the material, pottery finds of the early medieval period (c. late 7th/8th to late 9th/early 10th century) represent 16% of the total assemblage (in Graph 4 described as ‘EMED’), pottery finds of the early medieval to middle Byzantine period represent 2% of the total (‘EMED-MBYZ’), pottery finds of the middle Byzantine/Norman period (c. 10th to 12th century) represent 78% of the total (‘MBYZ’), pottery finds that can be dated only in general to medieval times represent 2% of the total (‘MED’), pottery finds of the late medieval/Angevin/late Byzantine period (c.13th to 14th century) represent 1% of the total (‘LMED’) and pottery finds of the Venetian period (c. late 14th to late 18th century) represent only 1% of the total (‘EVEN’) (Plate 1.1, Graph 4). In short, these numbers show that the majority of the medieval fragments excavated on the Vrina Plain can be dated to the middle Byzantine/Norman period of Butrint, and that ceramic finds of the post-medieval period are rare in this part of the city.

In addition, a comparison of the frequency of local and imported wares is presented in Graph 5. The imported wares were clustered into four main groups: ‘Southern Italy’, ‘Northern Italy’, ‘the Aegean’ and ‘either Southern Italy or the Aegean’. Looking at the quantities of local and imported wares from the Vrina Plain excavations in Graph 5, it is clear that imported wares from Southern Italy were found in the largest quantity (49%), followed by locally produced wares (41%) and imported wares from Southern Italy or the Aegean (7%), whereas the imported wares from the Aegean (1%) and Northern Italy (0%) are almost negligible (Plate 1.1, Graph 5).

In short, we can conclude from these graphs that the pottery assemblage consisted mainly of handle fragments of locally made coarse wares and undecorated amphorae imported from Southern Italy. The majority of these vessels can be dated in the middle Byzantine–Norman period of Butrint (c.10th–12th centuries). The predominance of this era is also shown in Graphs 6 and 7, where the totals of coarse wares and amphorae are presented by period. The numerical data in these graphs show that middle Byzantine coarse wares (58%) and middle Byzantine amphorae (66%) were the most numerous in the Vrina Plain pottery assemblage (Plate 1.1, Graphs 6 and 7).

Figure 1.1. Profile of Glazed White Ware rim and stem fragments from contexts 3322 and 3333

The pottery finds

Glazed fine wares

The excavations on the Vrina Plain yielded approximately five sherds of so-called ‘Glazed White Wares’ (GWW). These include a rim fragment and a stem fragment of the same vessel, which could be either a small chalice or a lamp of Glazed White Ware I (-II?) (Fig. 1.1; Plate 1.2). The rim is impressed on the top and decorated with red staining. The shape of this vessel looks very similar to 8th/9th-century examples from excavations at the St. Polyeuktos church in the Saraçhane district in Istanbul.⁴ The two Vrina Plain pieces were found in different contexts (3322 and 3333), but both in association with coins of the 9th century (context 3333: one coin of AD 820–29; context 3322: two coins of AD 842–67).⁵

Figure 1.2. Profiles of Monochrome Green Glazed Wares and Sgraffito Ware fragments from contexts 3044, 1006 and 3325

Furthermore, one body sherd and two small handle fragments of Glazed White Ware II were found (in contexts 3110, 3225 and 3325 respectively), the last handle being covered on either side by a yellow and green lead glaze and with a central groove on the exterior part (Plate 1.3). This piece is probably a handle fragment of a cup, and can be approximately dated to the 11th–12th century.⁶

Glazed White Wares are a group of lead-glazed ceramics made in a whitish kaolin fabric. These wares were mostly produced at workshops in Constantinople between the 7th and 12th century.⁷ Glazed White Ware II was more widely distributed than its predecessor, Glazed White Ware I.⁸ To date, Glazed White Ware II has been found on various sites in the Aegean, Balkans, Turkey, Crimea, Albania, Italy, and even in Sweden.⁹

Apart from these finds of Glazed White Wares, a small body piece of Glazed Red Ware came from context 3225, a layer above the mosaic in the sanctuary of the basilica (Plate 1.4). The sherd is covered on both sides with a transparent lead glaze (in Italy this type of pottery is known as ceramica a vetrina pesante), and can be roughly dated to the 9th–10th century.

Figure 1.3. Profiles of Polychrome Lead-glazed Ware type ‘RMR’ from contexts 1066, 3003, 3482 and 3488

Later types of imported fine wares from the Byzantine world were also recovered in the Vrina Plain complex. In the apse of the basilica, for instance, some sherds of glazed ware with a painted decoration were recovered in context 3217. These were four fragments of a dish made of a fine orange fabric covered with a white slip in the interior and painted motifs in green and brown under the lead glaze. The pieces are a type of painted tableware of the middle Byzantine/Norman period in Butrint, also known as ‘Brown and Green Painted Ware’ (Plates 1.5 and 1.6). The sherds can be dated in the second half of the 12th century. Both the shape and painted decoration of this dish have similarities with contemporary pottery finds from Central Greece, for example, from Corinth, Chalkis and rural sites in Boeotia.¹⁰

A few small fragments of imported fine wares from the Aegean, which were found in the basilica on the Vrina Plain, can be dated in the same middle Byzantine/Norman period. They include a rim fragment of a Monochrome Pale Green Glazed dish (context 3322; Plate 1.7(A)), a base and two body fragments of Monochrome Green Sgraffito dishes (contexts 3044, 86 respectively: Fig. 1.2 and Plate 1.7(B)), as well as a base, a rim and a body fragment of Monochrome Yellow Sgraffito dishes with gouged lines on the interior (contexts 1006, 3325; Fig. 1.2 and Plate 1.7(C)). These last three fragments with gouged decoration could belong to the so-called ‘Incised Sgraffito Ware’ of the late 12th/early 13th century.¹¹

A few sherds of fine wares from late medieval times were also recovered at the Vrina Plain excavations. Among these are three base fragments of bowls and a handle fragment of a closed vessel (probably a jug) of so-called Polychrome Lead-glazed Ware type ‘RMR’ (contexts 1066, 3003, 3482, 3488), the last one with brown stripes painted on the exterior (context 3003; Fig. 1.3).¹² One of the three base fragments shows a painted central grid-iron medallion (in a brownish colour) on the interior of the bowl, imitating the grid-iron design of Proto-Maiolica of the Brindisi type group I (context 3488; Fig. 1.3). Similar lead-glazed imitations of this grid-iron motif have been found at excavations in Torre di Mare, the medieval site of Metaponte in Basilicata.¹³ In fact, the grid-iron medallion is one of the most common motifs in Southern Italy: not only in Proto-Maiolica but also in ‘RMR Ware’.¹⁴ All the fragments of Polychrome Lead-glazed Ware type ‘RMR’ from the Vrina Plain can be dated to the late 13th to mid-14th century.

Finally, one rim fragment from context 7308 has a white slip and a pale creamish-yellowish lead glaze (7.5 Y 8/1) on the inside of an open vessel. It is further decorated with four incised lines on top of the rim (Plate 1.8). This could be a piece of a so-called ‘Zeuxippus Ware Subtype’ or ‘Zeuxippus Ware Variant’. This is a type of incised glazed ware of the 13th–mid-14th centuries, which was widely distributed throughout the Eastern Mediterranean (and especially in northern and central Greece and on the western coast of Turkey).¹⁵

Unglazed fine wares

Context 3469 yielded nine fragments of an unglazed vessel in a fine orange fabric (Plate 1.9). It is covered with a reddish slip (2.5 YR 6/8) on the exterior. In addition, its shoulder is decorated with fine straight and wavy incised lines. The shape is of a closed vessel with a spout on one side (probably a jug), which also includes a filter with holes pierced through on the inside. One can often find such filters on Islamic jugs used for the serving or pouring of liquids (mostly water). The filter is designed to keep impurities or insects out.

A similar-looking jug with incised wavy lines, spout and filter has been identified from excavations in the city of Crotone, situated on the Ionian side of Calabria, although the shape is not exactly the same. This latter vessel comes from a ‘Late Roman and Proto Byzantine’ burial area on the north coast of Crotone; it is currently part of a private collection known as ‘Collezione E. Palopoli’. Due to its incised decoration and painted surface this jug has been roughly dated to the 7th century.¹⁶ Considering the combination of shape, fabric, incised decoration style and context, the vessel from the Vrina Plain can tentatively be dated in the second half of the 7th century, or perhaps early 8th century.

Unglazed coarse wares

The Vrina Plain excavations yielded many fragments of unglazed coarse wares in orange gritty fabrics with numerous white quartz inclusions which were locally produced. The majority of these are wheel-made. Only two rim-neck fragments of small cooking pots from context 3408 appear to be irregular, handmade examples, but these are smoothed on the outside (Fig. 1.4).

Two large pieces belonging to the same cooking pot made in one such gritty orange fabric (exterior: 5 YR 7/8; core: 7.5 YR 7/1) came from two different contexts (Fig. 1.5). The rim-handle fragment (from context 3718) has an everted rim of a wide diameter and rows of incised wavy lines on the rim and on the exterior upper body. The base fragment of this vessel (from context 3736) has a concave base with central button in the centre. The shape of this locally made cooking jar is similar to 7th/8th-century pots recovered in Southern Apulia, especially at Otranto. Its incised decoration, on the other hand, appears to be more analogous to similar early medieval decoration styles from the lower Danube region (or Dobroudja region) in northeastern Bulgaria, also known there as the ‘Dridu Balkan-Danube Basin Culture’.¹⁷

A similarly decorated piece belongs to an open vessel (probably a dish) (Fig. 1.6 and Plate 1.10). It has an everted rim with finger-impressed sides, which is decorated with incised irregular dots on top. Furthermore, it has two incised wavy lines on the exterior upper part. Its fabric is finer than the previous decorated cooking pot, and its fracture is smoother. Its shape seems closer to earlier dishes (late 6th/7th century) from Canosa di Puglia.¹⁸

Some distinctive rim-handle fragments came from contexts 3361 and 3732 within the pottery kiln located in the small room off the eastern aisle of the basilica of the Vrina Plain complex (Fig. 1.7). They belonged to closed vessels, mostly wide-rimmed cooking pots with globular bodies and two plain oval handles. Their fabrics are primarily soft, medium-coarse and orange in colour (5 YR 7/6); some contain lime and mica as well as many angular white quartz inclusions. These sherds can be dated to the 9th and 10th centuries based on their contexts and by comparison with similar fabrics and shapes found at Butrint (e.g. Triconch Palace excavations) and in Saranda.¹⁹ In addition, similar versions of these medieval cooking pots can be found across the Adriatic Sea, on sites in the Salento region (e.g. Previtero at Otranto, Apigliano) and in Sicily (e.g. Catania), as well as on mainland Greece (e.g. Sparta).²⁰

Figure 1.4. Profiles of two unglazed coarse ware rim-neck fragments of small cooking pots from context 3408

One piece of unknown coarse ware, which is definitely not locally manufactured, was found in context 3488 (Fig. 1.8). In fact, this handle-body fragment is made in a very red fabric (10 R 5/6) with many large golden mica flecks, suggesting it might come from southwestern Italy or Sicily. Furthermore, it has a peculiar handle tail; a similar flanged handle on a lid or clibanus/testo (a baking oven/lid) was noted from a context of the second half of the 10th/mid-11th century at Santa Maria del Mare in Calabria.²¹

The locally made cooking pots of the middle Byzantine period (c. 10th–12th century), on the other hand, are smaller in shape than their early medieval predecessors, and have a smaller rim diameter (Fig. 1.9). Most of them come from contexts 3000 and 3003. They are grooved on the exterior surface and sometimes on the exterior rim, but they are not decorated with incisions. Their fabric is gritty orange, with many white and light-grey quartz inclusions. They can have traces of fire use on the exterior rim and body, and on the interior upper wall.

Finally, one rim-handle fragment of a middle Byzantine local cooking pot from context 3218 can probably be dated slightly later (Fig. 1.10). This piece is related in shape to cooking pots from excavations in Nichoria (in the southwestern Peloponnese) which are dated to the late 12th/early 13th century.²²

Amphorae

As we have seen above, a large part of the medieval pottery assemblage from the Vrina Plain excavations consists of fragments of undecorated imported amphorae