The Leiden Ljubljana Ancient Cities of Boeotia Project 2009 Seasons

The Leiden-Ljubljana Ancient Cities of Boeotia Project 2009 seasons
JOHN BINTLIFF, BOZIDAR SLAPSAK, BART NOORDERVLIET AND JANNEKE VAN ZWIENEN, INGE UYTTERHOEVEN, KALLIOPE SARRI, MARK VAN DER ENDEN, ROB SHIEL & CHIARA PICCOLI
Abstract This article forms a preliminary report on the work undertaken in 2009 by the Leiden-Ljubljana Project in Boeotia, Central Greece. The main research was focussed on the ceramic and architectural surface survey at the ancient city of Koroneia, but additional work has been carried out on cities studied earlier by the Boeotia Project, Hyettos, Haliartos, and Thespiai in connection with the preparation of monographs on each one. Ceramic studies and land evaluation programmes continue alongside the archaeological fieldwork, and a further stage in the recording of traditional vernacular architecture has introduced advanced virtual reality into our sub-programme investigating the post-Medieval evolution of the Central Greek landscape. Keywords Boeotia surface-survey urbanism landscape vernacular architecture.
Introduction (John Bintliff, Bozidar Slapsak, Bart Noordervliet & Janneke van Zwienen) During August 2009 the joint Leiden-Ljubljana Project continued work at the ancient city of Koroneia, the city of Thespiai, the city of Haliartos, and revisited the city of Hyettos. Joint teams of staff and students from the respective universities were directed by John Bintliff (Koroneia, Hyettos) and Bozidar Slapsak (Haliartos, Thespiai). The assistant director was Professor Vladimir Stissi (Amsterdam), and the assistant field director was Dr Athanasios Vionis (Leuven/University of Cyprus). A ceramic study season was carried out in April under the direction of Professor Jeroen Poblome (Leuven) and Professor Stissi. Key staff members for the fieldwork at Koroneia and Hyettos were Bart Noordervliet and Janneke van Zwienen (Leiden), whose specialist research in digital recording and GIS was fundamental to the project. Emeri Farinetti (Leiden) was the database manager.1
1 Previous work has been presented in Pharos. The Journal of the Netherlands Institute at Athens: Bintliff & Slapsak 2006; Bintliff et al. 2007; Bintliff et al. 2008.
Pharos 17(2), 1-63. doi: 10.2143/PHA.17.2.2162462 2010 by Pharos. All rights reserved.
2 Koroneia
JOHN BINTLIFF ET AL.
As in previous seasons, surface survey of this large Greco-Roman city continued, using a standard 20 20 metre grid-unit for the recording of surface finds and surface architecture. By the end of the summer season, 581 survey quadrants had been completed, mostly in the ideal size, but a minority in other dimensions when survey conditions were too poor for full treatment. The total area covered is now 22.6 hectares of the city-hill (Figure 1). The exact boundaries of the maximum extent of the ancient town are only known at some points, and on this Figure we have shown two lines (in orange), which are the uppermost and lowermost imaginable for a strong Lower Town fortification. They represent 22 hectares and 53 hectares, respectively. On the basis of several pieces of evidence, including one definite and one possible cemetery, and the spread of built-up housing detected in our grid-units, we believe the actual maximum extent of the Greco-Roman city lay somewhere in between these two figures. Some tentative, more realistic alignments for the city wall are shown as dashed purple lines. On that basis we have perhaps covered one half of the city surface. At the end of the season, a large field which had been bulldozed by a local farmer yielded numerous traces of architectural fragments and an inscription, in an area (marked as a blue locality on Figure 2) where the evidence just cited had led us to suggest a city wall-line, and this may well be a periurban cemetery in the vicinity of a town-gate. The hypothetical wall-line here will necessarily need to be drawn slightly further west so as to place this field just outside and to the suspected cemeterys north. Additional support for defining the edge of the built-up town comes from the total ceramic density map for the grids covered so far, corrected for visibility of the ground surface (Figure 3). As a rough guide, intensive analysis at the city of Thespiai in Boeotia has indicated that the boundary here shown of around 21,000 sherds per hectare may mark city-periphery, which means that all the zones from blue-colour to higher could be urban, green and lower extramural, with the exception of some zones on the acropolis where bedrock has limited occupation deposits or where vegetation covers their visibility. The low values in the far south are clearly non-urban, and lie across a natural stream boundary. In the north-west and north-east, values above and below the guidelines, may suggest we are near the city edge, but caution is needed in using these values with precision, since the known Classical cemetery and the Frankish period village fall mostly into city density but we believe these to be outside the ancient town walls. Clearly a combination of quantitative and qualitative criteria will be needed. The new putative city-gate cemetery was not the only place where farmers have begun to opt for speedy but unprecedentedly destructive reorganisation of agricultural land inside the ancient town, the other being the agora. Fortunately our
THE LEIDEN-LJUBLJANA ANCIENT CITIES OF BOEOTIA PROJECT
2009
Figure 1. Map of the ancient city hill of Koroneia, with survey grids studied between 2007-2009. The upper and lower orange line mark the likely outer boundaries for the fortification wall around the Lower Town, whilst the purple line represents currently the most likely course
Figure 2. Ancient Koroneia, possible wall lines as in previous figure, with the addition of the location of a likely ancient cemetery located in 2009
2009
Figure 3. Density per hectare of surface ceramics at Koroneia, corrected for surface visibility
summoning the phylax (archaeological site guard) for the site soon brought an end to such damage. Amongst important discoveries from this season, in situ blocks from the Theatre gave some insights into its construction, whilst our architectural expert Dr Inge Uytterhoeven (Leuven) (see further this report), has suggested that the large bath at the highest part of the acropolis is a major public building, perhaps a bishops palace, datable from its tiles to the 5th-6th centuries AD. She also clarified the supposed church on the lower southern plateau of the acropolis as Late Antique clumsy housing with a heavy use of spolia. In the Lower Town, a massive Late Antique building could be a private mansion, less likely a church. The surface ceramics continued to appear in extraordinary quantities in almost all the grids examined, and show clear evidence on first inspection that different districts of the city had varying periods of use. Whilst Archaic finds indicate that large parts of the town were already active in the earliest urban phase, Classical finds are ubiquitous, and Late Hellenistic-Early Roman appear equally-so, although perhaps concentrating on the period till the Late Republic rather than in the early Imperial centuries. Middle to Late Roman material in contrast appears to be much more confined in its extent. So far Medieval finds are rare except around the Crusader feudal tower at the north-east base of the city hill, and there they are essentially of Frankish age. We have still then to find the presumed Byzantine predecessor to the dependent village of the 13th-14th centuries, although sporadic ceramic pieces of pre-Frankish medieval date have been found in the upper parts of the Lower Town by Dr Vionis. Little of Post-Medieval date is visible, supporting the Ottoman archival evidence for the removal of the Frankishera village to the modern settlement of Ayios Georgios some two kilometres south-west of the city during the late 14th-early 15th centuries. Also only rare finds of prehistoric pottery and lithics support older mention of Neolithic-Bronze Age activity at the site (see the report of Kalliope Sarri below); we have yet to find any settlement focus, although the overlay of literally millions of Greco-Roman habitation-debris sherds and roof tiles makes the discovery of prehistoric artefacts extremely difficult. These immense quantities of surface potsherds had already given rise to lively discussion between the survey team and the Project ceramic specialists, as to whether we were collecting the most representative sample for laboratory study. In particular, an apparent poverty of cookwares seemed to suggest discrimination in the collection-phase. In order to investigate these issues, Mark van der Enden (Finds Laboratory Supervisor, Leiden/Leicester) set up experiments to test the relationship of the collected sample to the total available on the site surface (see his report below). This involved comparing the make-up of our normal grid collection, a sample from each 20 20 m unit, with a series of 4 4 m squares placed
2009
within such grids, but from which all pieces were brought back. The main criteria evaluated were size, fabric and the main functional classes (cookware, other domestic wares and fineware). This is intended to form a separate publication by Mark van der Enden, but some of the chief provisional findings are worth recording here. Firstly, the representation of coarseware in the total collections was little different from that in the samples, if we exclude tiny fragments which are not gathered in the samples (the rule there, is a sherd must be larger than a fingernail). In fact in the total samples much coarseware was indeed in the sub-fingernail size, suggesting that one reason for lower than expected coarseware recovery is due to its far higher risk of disintegration than better-fired wares. Secondly, it appears likely that in regions like Boeotia, with excellent potting clay and traditional highquality products, cooking vessels appear to have been made using fabrics and hardness usually qualified as domestic ware. In addition it can be suggested that much more use was being made of grilling and other forms of food preparation rather than in formal cooking-vessels. An additional observation was made by Professor Stissi on the two comparative samples: as far as chronology is concerned, where quantities are considerable, as is the case at Koroneia, the information from body-sherds adds little to the information obtained from feature-sherds (rims, bases and handles), with the rare exception of prehistoric pieces, where their scarcity requires every possible piece to be retained. Bart Noordervliet and Janneke van Zwienen registered the study-grids and their qualitative and quantitative data on GPS-linked handheld computers (they are responsible for all the mapping data presented in our Figures), as well as continuing the complete fine-detail mapping of the city surface using a Differential GPS device (measurements on average every 7 metres see Figure 4 for the latest map of points measured-in). The three-dimensional model of the city-hill (Figures 5-6 are from the most up-to-date version), is the basis for their ongoing-analysis of possible street and path routes for ancient inhabitants, as well as the layout of residential and public zones. All architectural pieces are being recorded to centimetre-accuracy, and are being studied by Dr I. Uytterhoeven (see her report below). Haliartos This ancient city was surface-surveyed in the mid-1980s by the earlier Boeotia Project, under the direction of Professor Anthony Snodgrass (Cambridge).2 In connection with the final publication monograph for this site, the Ljublana team under the direction of Bozidar Slapsak, began geophysical survey across the city
2
Cf. Bintliff & Snodgrass 1988.
Figure 4. Map of the points recorded for surface topography and elevation by Differential GPS in 2008-2009
2009
Figure 5. Digital elevation model of Koroneia, produced from the DGPS measurements mapped in Figure 4, vertical view
10
Figure 6. Digital elevation model of Koroneia, produced from the DGPS measurements mapped in Figure 4, oblique view
surface. It is hoped that over the next few years as much as possible of the town will be studied in this fashion, since balloon and light-aircraft photography by the same team has already shown that a gridplan with well-marked streets and houseblocks can be observed from altitude. Since the city was destroyed by the Roman army in the 2nd century BC, and was subsequently only partially reoccupied, this townplan is likely to reflect the Classical-Hellenistic city with little later modification. Thespiai The second monograph to be published, of the older Boeotia Project, will cover the 1980s surface survey of the city of Thespiai, carried out by John Bintliff and Anthony Snodgrass.3 As part of the checking of our knowledge of that large site (some 100 hectares at its maximum), one major unknown element remained to be clarified. Amongst a small number of standing ruins on the modern site surface, a striking faade and rectilinear associated wall-plan in the far east of the town remained undated. It had long been assumed to be a Late Roman, Early Christian basilica. In the period since our surface survey, however, the monument, clearly once-excavated but never attributable to a known archaeologist, has suffered the
3
Bintliff & Snodgrass 1988.
2009
11
indignity of becoming a dump for masses of field clearance stones by local farmers. Not only were we still unsure of the chronology of this major building, but it was clearly being rapidly destroyed by this very recent damage. In response to a request, the Byzantine Ephoreia intervened, and with the assistance of a labourforce provided by the mayor of the modern village of Thespiai, and the technical support of the Ljubljana team, the monument was cleared of its dumped stone and through subsequent clearing, revealed in suitable detail to obtain at least the main lines of its history (see Bozidar Slapsaks report below). Although further research may nuance this picture, at this point the likely history of the building(s) appears to Professor Slapsak as follows: it was undoubtedly excavated at the end of the 19th century AD by the French epigrapher P. Jamot, in search of inscriptions, then left till the 1990s as an open hole in the ground. The oldest phase is a small church of Late Roman times, which is followed by a larger Middle Byzantine successor. These however are slighted by a giant platform, whose faade had been mistakenly taken to be one end of these churches, but in fact this now appears to be a large multi-storey rectangular structure. It may be that the Byzantine nave was remodelled to accompany this solid block. This final complex structure appears to be Frankish in date, and some Gothic architectural fragments were recovered. If the Late Antique and Middle Byzantine churches can be associated with the final phase of the city and the subsequent village of Erimokastro in its eastern extremity, respectively, the Crusader structure can be linked both to the Frankish-era continuation of the eastern village but more significantly to a papal letter, which indicates that an Italian branch of the Latin monastic order of Premonstratensians was given the feudal control over the Byzantine village in 1210 AD following the Fourth Crusades conquest of Greece.4 It is probable that the last phase of the building represents a tower-(and church?) complex as the tangible presence on the site of the new lords of the village-community of Erimokastro. One other vital new element in our understanding of the ancient city of Thespiai came from the expertise of Michaelis Kambanis, representing both the Byzantine Ephorate and Leiden University. Michaelis Kambanis recognizes in architectural members located earlier by our project from the Late Roman fortified enceinte or Kastro, in the centre of the ancient town, pieces of church architecture of the 8th and possibly even 7th centuries AD, a time of Slav occupation of the Boeotian countryside. This, taken with Dr Athanasios Vionis identification of rare ceramic finds of the same period from several points scattered across the ancient city-site, allow us to suggest that the town survived in occupation probably in
Koder & Hild 1976, 275.
12
dispersed village-form, from the end of Antiquity into the Middle Byzantine revival of the 9th-12th centuries. Hyettos The city-site of Hyettos and a section of its countryside, originally surveyed in the late 1980s and early 1990s,5 was revisited this season in preparation for publication. GPS measurements, which were obviously not available during the prior survey, have been used to check and, where needed, correct the absolute location of smaller sites identified in the ancient towns vicinity. An attempt has also been made to check the locations of a number of significant architectural finds within the urban area itself. As it turns out however a large number of those have either been taken away or at least shifted, making it difficult to relocate them based on the old measurements and photographs. However, using more stationary landmarks such as roads, churches and, on one occasion, an old well, the old grid system could be reconstructed and updated with the new information. In the countryside north of Hyettos, part of the citys dependent countryside which had been intensively-surveyed in the late 1980s and early 1990s was revisited by the directors of that project, John Bintliff and Professor Anthony Snodgrass, together with Bart Noordervliet, to relocate the rural sites previously discovered and place them in a GPS-controlled absolute location. Meanwhile, also in preparation for the publication of Hyettos and its landscape, volume 3 planned in the Boeotia Project monographs, Dr Popi Sarri re-examined all the prehistoric finds from this area of countryside (see her report below). Prehistoric Ceramic Research With the aid of an INSTAP grant, Kalliope Sarri (assisted by a student, Ray di Caccio) conducted a detailed re-examination of the prehistoric sherds from Thespiai city, and the Hyettos countryside sites, as well as working for the first time on some parts of Koroneia town. The work on older surveyed sites was in preparation for their final publication (see her report below). Vernacular Architecture Project Following on a series of earlier recording of historic domestic houses by the Boeotia Project,6 a new phase of this activity was inaugurated in 2009 by Chiara
5 6
Bintliff et al. 2000. See the contribution by Verweij in Bintliff et al. 2007, 37-42.
2009
13
Piccoli as part of her preparation for a PhD study at Leiden. With the aid of a Robotic Total Station and suitable photographic image-processing software, she undertook 3-dimensional reconstructions of ruined houses in the villages of Mazi and Evangelistria, near the project base, as well as collecting local ethnohistoric information from residents regarding the use and furnishings of these houses, one of which is believed to date back to the early 19th century (see her report below). Given the rapid destruction of pre-modern traditional houses in Greece, this work of recording and visualisation is a form of heritage-work which can catch the last manifestations of a way of life likely to date back to the late Middle Ages in the case of the dominant longhouse accommodation.7 More complex multistorey houses of the late 19th and early 20th centuries also however represent important historical monuments as they can be associated with major transformations in the Greek countryside through rising commercialism, globalisation and economic and political stability. Chiara Piccoli was assisted with technical support by Bart Noordervliet and Janneke van Zwienen, and with the ethnographic interviews by Dr Vionis. Eric Dullaart (Leiden) gave helpful advice on the Robotic Total Station. Architectural remains at Koroneia (Inge Uytterhoeven) From the 5th until the 9th of April 2009 an architectural survey was conducted at Koroneia. Architectural remains, both in situ and architectural fragments out of their original contexts that had been localised during previous survey campaigns on the site, were measured, identified and photographically recorded (I. Uytterhoeven), while their position was accurately defined by GPS (J. van Zwienen and B. Noordervliet). In total 26 in situ wall sections, 116 architectural fragments, two fragmentary floors and ten collapsed vault fragments were studied. The larger part of the recorded building blocks were made of local limestone. With some exceptions most of these architectural fragments were undecorated and rather roughly finished (see Figure 7 for an indicative plan of the major areas of the ancient city mentioned in this article and for the location of some of the illustrated architectural pieces, marked by their figure numbers.) The architectural remains recorded at Koroneia can be classified in various functional groups. Defensive architecture A first group of architectural material consisted of remains of the fortification walls of Koroneia. Apart from several wall stretches that were found in situ, individual displaced blocks were recorded.
7
Sigalos 2004.
14
Figure 7. General map of Koroneia, showing the location of the major sectors of the ancient city, as well as specific findspot for some of the architectural pieces discussed by I. Uytterhoeven in this article with their figure numbers indicated (J. van Zwienen)
Although it can be presumed that the acropolis wall originally consisted of a double wall curtain, filled with stone material, in all cases only a single row of blocks has been preserved in situ and evidence for a second (inner) curtain wall is not recognizable at the surface at present. Both the wall blocks in situ and the individual dislocated building elements registered were large-sized polygonal limestone blocks (up to 1.00 m long) with one or more worked faces. Numerous individual building blocks, including several of those that had fallen down from the western slopes of the acropolis, were provided with a straight-cut corner in order to fit the stone with the adjacent block (e.g. Figure 8). Thus, according to the type and dimensions of the individual blocks the Koroneia fortification wall was constructed
2009
15
Figure 8. Polygonal building block of the fortification wall with one straight-cut corner
Figure 9. In situ fragment of the fortification wall in polygonal masonry at the north-east side of the acropolis
Figure 10. In situ fragment of the fortification wall in polygonal masonry at the south side of the acropolis
Figure 11. In situ fragment of a wall in polygonal masonry with incorporation of the bedrock on the upper slopes of the Lower Town
in polygonal masonry8 with courses (ca. 0.45-0.55 m high) of well-fitting building stones (e.g. Figures 9-10). Besides, in correspondence with the terrain situation the natural bedrock was sometimes incorporated in the artificial polygonal wall, as attested in a wall stretch on the upper slopes of the Lower Town, perhaps part of its early fortification (Figure 11). Apart from the actual acropolis area similar polygonal masonry was used for a north-east to south-west running wall section of seven wall blocks in situ that was identified further southwards along a modern path/road. This wall segment must have enclosed the southern part of the inhabited city area, as a Lower City wall (Figure 12).
8 Following the definition of polygonal masonry in Scranton 1941, 45: The proposal, therefore, is to use the word exclusively for masonry the blocks of which have varying numbers of straight nonparallel sides, usually more than four, which meet at clear-cut angles.
16
Figure 12. In situ fragment of the fortification wall in polygonal masonry surrounding the southern part of the inhabited lower-city area
Although dating of the fortification walls of Koroneia on the basis of their building material and technique is difficult, parallels of regular polygonal masonry suggest a construction date in the Classical period.9 Public architecture A number of architectural blocks recorded in the field may be identified as elements of public buildings. These include for instance a fragment of a large fluted column shaft (flutes of 0.10 m wide; Figure 13), a large threshold (max. preserved length: 2.34 m; Figure 14), a fragmentary dentil frieze (0.43 0.37 0.30 m; Figure 15) and several large limestone ashlars. The largest concentrations of large architectural fragments were found in the supposed Agora area on the eastern middle slopes, and near the theatre on the eastern side of the hill. Apart from the architectural fragments out of position, some walls, such as a 2.37 m long stretch in ashlar masonry combined with rubble of ca. 1.50 m wide (Figure 16), may have been part of constructions with a public function, given their large dimensions.
9
Cf. Scranton 1941, 54-69; Fossey 1988.
2009
17
Figure 13. Fragment of a large fluted column shaft
Figure 14. Large threshold
Figure 15. Fragmentary dentil freeze
Figure 16. In situ fragment of a presumably public building in ashlar (and rubble) masonry Figure 17. Collapsed vault fragment in brick and mortared rubble on the southern edge of the acropolis
Finally, the numerous collapsed vault fragments in brick and mortared rubble found on the far northern sector of the acropolis, which were already noticed by travellers in previous centuries, may have belonged to a public building (Figure 17). Since several of the rubble-brick fragments were still partially lying over each other, it is highly probable that they are still in the position in which they originally collapsed. The length of the individual collapsed fragments varied
18
between 1.20 m and 4.80 m. The bricks used in the vault and carrying arches had dimensions of 0.26/0.27 0.26/0.27 0.030/0.035 m and 0.36/0.40 m (dimensions of mortar joints: 0.03-0.045; 0.06 (H)/0.017-0.03 (W)). Although one should be cautious when using brick dimensions as a chronological indicator and although further research on the topic is still needed, parallels known from stratigraphical contexts in Greece point towards a late antique construction date.10 On the basis of the preserved surface remains the building has previously been identified as a possible church or bath building. Nevertheless, a residential function cannot be excluded. The location of the building at one of the highest and best visible points of the city, from which one had, moreover, excellent views on the surrounding landscape, may have formed an important criterion in the choice of the building plot for a large elite residence. Domestic architecture Other architectural fragments may be related to domestic contexts. This is for instance the case with a limestone door jamb (0.31 0.45 m) recorded in the south-eastern area of the town (Figure 18) and a fragmentary limestone basin (Figure 19). In addition, based on the survey material collected during previous campaigns, small stone walls built on the citys (south-)eastern slopes could be interpreted as in situ remains of dwellings. Furthermore, at the south side of the acropolis, in an area where in the recent past excavations were carried out, several north-south running walls were studied, as well as perpendicularly constructed east-west walls. These walls formed smaller, rectangular units that presumably functioned as rooms of houses or shops. The walls in the old excavation trench were preserved up to a maximum height of 0.60 m, had a width of ca. 0.60-0.70 m and were constructed in mortared rubble with inclusion of brick fragments and numerous spolia, such as ashlars and a column shaft. A moulded altar base (0.49 0.64 0.34 m) located in between two of the walls must equally have been re-used in a wall construction, seen from the mortar traces on several of its faces (Figure 20). A similar incorporation of column shafts and re-used ashlars in mortared rubble walls was also recognisable higher up the slope in a north-eastern direction (Figure 21). The occurrence of numerous spolia in the walls of this and the previously excavated area suggests for both zones a construction date in Late Antiquity. Apart from this, an opus spiccatum fragment in brick that was recorded along the modern road to the south-east of the acropolis, as well as the related pink
10
Cf. Aupert 1990, 632.
2009
19
Figure 18. Limestone door jamb
Figure 19. Fragmentary limestone basin
Figure 20. Late antique domestic area with mortared rubble walls incorporating spolia
Figure 21. Late antique rubble wall including spoliated columns
mortar floor substrate, may have been part of the pavement of a (service) room in a (late) dwelling (Figure 22). Industrial installations: presses Finally, several very large architectural elements could be identified as fragments of press installations, either used for the production of oil and/or wine.11 All these elements were made of limestone. They included entirely and fragmentarily preserved press basins (e.g. Figure 23:, with height 0.49 m, exterior diameter 1.34 m; Figures 23, Figures 24-25, with height 0.94 m, diameter at bottom, with height 1.55 m; Figures 27-28, with height 0.43 m, exterior diameter 0.685 m; Figures 29-30, with height 0.78-0.81 m, exterior diameter 0.73 m), and a rectangular beam press11 For the identification of installations related to oil and wine processing, cf. Brun 1993; Foxhall 1993; Carington Smith & Wall 1994, 373-374.
20
Figure 22. Opus spiccatum fragment
Figure 23. Press basin
Figure 24. Fragmentary press basin
Figures 25-26. Press basin
2009
21
Figures 27-28. Press basin.
Figures 29-30. Press basin
Figure 31. Beam press-weight.
weight (0.80 0.64 0.28 m; Figure 31). Since similar press installations stayed in use for centuries, it is not possible to accurately date the press and weight elements. However, they probably form attestations for the ruralisation of the city centre in the late antique or Byzantine periods, notably one of the largest, on the acropolis.
22
Our preliminary detailed architectural survey at Koroneia gave us a first idea of the type of material occurring on the site. A confrontation of the already obtained and future architectural survey results with other surface data, such as pottery and small finds, as well as the plotting of the recorded architectural features on the Koroneia site map will certainly reveal more important information about the use of the various parts of the town, and, in general, on its rich occupational history. The Work of the Ljubljana Architecture, Aerial Photography and Geophysics team in 2009 (Bozidar Slapsak) In 2009, the Ljubljana team, besides processing data from previous campaigns in Thespiai for imminent publication and assisting the epigraphic team in documenting new inscriptions found in previous years, focused on recording of the site of the ancient theatre in Thespiai, carried out groundtruthing of aerial photographs taken over the site of Haliartos for georeferencing and mapping surface architecture, and started geophysical prospections on the lower city of Haliartos. We also assisted the Ephorate of Byzantine Antiquities in cleaning and documenting the site of Barbakas/Makri Ekklesia at the eastern outskirts of the ancient city of Thespiai. Some thirty new inscriptions found by the Ljubljana team at Thespiai during surface architectural recording are being studied by Fabienne Marchand and Albert Schachter, with the consent of the Ninth Ephorate of Prehistoric and Classical Antiquities. They include funerary texts from the Classical (one possibly late Archaic), Hellenistic and Roman periods; dedications to Apollo and Artemis, to the Great Mother of the Gods, to the polis; what appears to be the beginning of a list of tax farmers; a partial victors list (of the Mouseia); a bilingual dedication by Domitian; a boundary stone of the sanctuary of Theos Tauros (the deified Titus Statilius Taurus); a ceramic standard measure incised (in cursive script) with the name of the agoranomos; and a text mentioning the festival Panamia. Exact findspots are provided by our team, and a list of stones which remain to be transported to the Museum has been presented to the Ephorate, including most of the inscriptions and some important anepigraphic stones. The newly identified site of the ancient theatre of Thespiai reported here in our 2006 report, and in Teiresias,12 has been surveyed for a digital terrain model by Uros Kirn in order to make a detailed analysis of its surface morphology. The initial measuring position was defined by GPS by Bart Noordervliet of the Leiden team, and checked against other standpoints across the site previously set by Andrej Bilc, while individual measures were taken by Total Station on the grid of 1 1 m on average. A series of aerial photographs by the Hellenic Military Geographic
12
Slapsak 2007, 17-20.
2009
23
Service, covering the period between 1945 and 2005, was acquired to study recent changes on the site. By combining information from the photographs and the DTM, and by considering local geology, we hope to be able to explicate the taphonomy of the site and provide evidence beyond doubt for the crucial role of anthropic agency in the shaping of the site, and for its fitting the design of an ancient theatre. The site of Makri Ekklesia is located at the east edge of the urban area of ancient Thespiai south of the Kanavaris stream, just west of the Barbakas spring. Both represent somewhat of a landmark in the area: Barbakas is a notable fountain structure by the track to Pyrgos and the Thespiai Polyandrion, built mainly from ancient spolia and noted by travellers, together with the springs of Dendra, Ayios Ioannis and Megali Vrisi to the west.13 In its present form, it is to be dated to the Turkish period, with some post WW II repairs in concrete. It has been abandoned since the 1960s, but remains in vivid memory of the local people as the place where the villagers of Leondari would come for water and to wash their clothes. In the late period of its use at least, it would get water also from a spring further up the hill, some 350 m to the south-west: the last stretch of the aqueduct is still visible behind the fountain. The water from the spring was captured in basins across the track to the north-east and used for irrigation of the fields to the north, while some 75 m further to the south-east, there used to be a wheel-driven well, the remains of which are still visible due north by the track to Pyrgos. Makri Ekklesia, on the other hand, is the ruin starting some 20 m west of Barbakas and stretching for some 30 m in a westerly direction. Before cleaning, walls were visible in the western part of the site to the maximum height of 1.8 m above ground, while the eastern part was covered with loads of rubble amassed there from the neighbouring fields. The ruin has been the source of stone for the locals, and object of excavation, clearly also by Jamot in the 1880s: he extracted a number of inscriptions there noted under that name in epigraphic publications.14 However, in his archaeological notes and articles, he gives no observations on the structural remains there. The toponym refers to some long church the memory of the patron saint has faded among the locals, and there appears to be no record in the church archives. In early epigraphic publication though, Ayios Loukas is noted as a findspot of inscriptions in the vicinity of Barbakas,15 along with at least one other ruined chapel (a lump of concrete in a field to the north-east of Barbakas may indicate the site of one such chapel): identification with Makri Ekklesia is a possibility, to be verified further. The special interest of the site lies in the fact that it is situated in the core of the
13
Ulrichs 1863, 82ff.; he mentions, besides Varvaka, Ayios Ioannis and Megali Vrisi, a spring named Tis Itias(of the willow), possibly to be identified with Dendra (trees). 14 Cf. Roesch et al. 2007, passim. 15 Site names referred to and clearly related to Makri Ekklesia are Barbakas, the church of Ayios Loukas near Barbakas, and the ruined chapel near Barbakas.
24
surface distribution of Byzantine, Frankish and Ottoman ceramics as detected by the Bintliff-Snodgrass Thespiai urban survey.16 The two ruined structures would thus be the last remaining architectural markers of the Medieval and post-Medieval village, which developed on the eastern outskirts of the abandoned ancient city south of the river Kanavaris, parallel to another half of that village which lay around the surviving church of Ayios Athanasios north of the stream. The last inhabitants of the one remaining village in the plain referred to by the travelers as Lefka17 moved to Erimokastro (Thespies) and Kaskavelli (Leondari) on the hills immediately above and to the north, between 1805 and 1806.18 Arguably, the late village or what remained of it in the first years of the 19th century is to be located in the general area of Barbakas/Makri Ekklesia, within the zone of the southerly of the two Medieval post-Medieval ceramic scatters identified in the 1980s BintliffSnodgrass city surface survey seasons. In 2009, the 23rd Ephorate of Byzantine Antiquities decided, upon the initiative of the Demos of Thespies, to clean the site of the debris amassed there since the late 19th-century investigations, and agreed that the Ljubljana team assist in the process. As this work will be the subject of a specialist publication by the Ephorate we shall confine our remarks here to the provisional results from the cleaning as it affects our knowledge of the ancient city in its final phases and of the subsequent village occupation. An Early Christian basilican church of some size was replaced by a Middle Byzantine cross-in-square church, then by a tower of presumed Frankish-Crusader Age. The Late Roman monument will have been on the fringes of the extramural settlement of the city, whilst the Middle Byzantine successor was presumably one of several churches in the two-focal contemporary villages known collectively as Erimokastro in our oldest text referring to the settlement, a papal letter of the early 13th century.19 That letter, by referring to the handing-over of the village to the monastic order of Premonstratensians from southern Italy, may be closely connected to the Frankish tower, which probably formed the focus of the colonial management of the village and in the basement of which would be stored the tithes from villagers. The spring of Barbakas was preliminarily cleaned of the rubble, garbage and vegetation: the functional level in front of the spring has not been reached, however.
Bintliff 1996, 4ff. and fig. 10. E.D. Clarke visiting in 1801, W.M. Leake in 1802, and E. Dodwell in 1805 all refer to the existing village of Lefka. 18 Upon his 1806 visit, Leake remarks that the three cottages he saw in 1802 were now abandoned; travellers texts are being studied by A.M. Snodgrass for imminent publication of the urban survey in Thespiai. 19 Koder & Hild 1976.
17
16
2009
25
Prehistoric finds from Boeotian surveyed areas: an update (Kalliope Sarri) During the 2009 season the study devoted to the prehistoric finds had the character of a general updating of our knowledge on the local pottery groups, in order to confirm the spatial distribution of such early finds, both in the landscape and within particular sites. Since the Boeotia project has been active in research for many years, being assisted by various specialists, the changing concepts and the growth of available data have made it necessary to coordinate changing concepts in order to create an updated, uniform terminology for the pottery description of older and recent finds. Our aims for the 2009 study season can be summarised as follows: a. The inspection of new material coming from the ongoing survey at Koroneia, looking for diagnostic prehistoric finds. b. An updated analysis on the finds from older survey seasons (Hyettos, Thespiai) in the light of recently published results or ongoing projects. d. The shaping of a common language of terms and methods for the study of prehistoric surface finds from Boeotia. For this purpose we used the prehistoric reference collection, created during the Tanagra survey, as well as the rich Neolithic pottery assemblage from the Thespiai Magoula, which makes up for the missing Neolithic periods from the Tanagra area. e. The refinement of dating methods in order to extract more information from the finds, i.e. by developing various levels of diagnosticity and new fabric definitions. The issues mentioned were investigated site by site in four different surveyed areas of Boeotia: the Koroneia hill, the rural area of Hyettos, the Thespiai Magoula and the Tanagra off-site region.20 Koroneia The prehistoric presence at Koroneia is referred to by previous studies. The collections of J. Fossey and the British School point to habitation during the Neolithic (N), Middle Helladic (MH) and Late Helladic (LH) IIIA and B periods.21 More evident are traces of prehistoric habitation (N, MH and LH) on the nearby lowland site of Kalami.22
We would like to thank the Institute for Aegean Prehistory for granting funds to enable this study season and Mr R. di Ciaccio for his valuable assistance. 21 Fossey 1988, 326, n. 19. 22 French 1972, 10, 13, 29, 31-33, 36, figs 5-7, 11; Mountjoy 1986, 103; Fossey 1988, 336-337, Syriopoulos 1994, 149-150, 387, 637.
20
26
The field seasons 2007 and 2009 at Koroneia, which covered a large area of the ancient city hill,23 produced just a handful of prehistoric finds, and of insecure date. Amongst the total amount of 300 sample bags from the 2007 season, only two possible Mycenaean sherds and a few sherds of prehistoric coarse fabrics were found. Amongst the total amount of 86 findspots of the 2009 surveying season on the east slope of the city hill, we were able to recognize only one LH IIIC skyphos fragment, some pieces of chert and two grinding stones. Since these finds are not particularly diagnostic, the prehistoric evidence from the surveyed areas of Koroneia is still very weak. The fieldwork at Koroneia, as well as the study of the whole 2008 season collection, will be continued in 2010; hence it is nescessary to find some more prehistoric evidence from other areas of the urban survey. Hyettos A series of sites located in the rural landscape of the ancient city of Hyettos (Figure 32) have been re-studied. The survey in this region which was conducted during the 1989-1990 seasons24 located a number of prehistoric sites, consisting of concentrations mainly of N and BA finds on the upland plain and adjacent hill land around Hyettos. Some preliminary comments on the chronology made just after the fieldwork must now be revised and updated in the light of new publications and research projects on the Greek Mainland. These have modified significantly the picture of the local prehistoric pottery classes, especially for the FN and EH periods,25 evidence for which we can now find in Boeotia, in the region of Hyettos and in a series of sites in the Asopos valley.26 The Hyettos finds served as the test material for this years effort to improve our dating approaches. After years of studying surface finds, we have come to realise that answering the question of a periods presence with a simple yes or no was insufficient. Recording of some prehistoric features in our crude undiagnostic material could have a decisive effect on the statistical evaluation. Therefore, a four-scaled diagnosticity factor for the dating was introduced: 1 = firm dating, 2 = quite possible, 3 = possible and 4 = questionable. This method has proved quite useful for avoiding random similarities between fabrics of different periods, has increased the accuracy of our results and assisted the recognition of some new local pottery classes. We give the following short guide to the prehistoric sites of the Hyettos region after our last study season: two areas showed a remarkable presence of prehistoric
23 24 25 26
Bintliff et al. 2007, fig. 6. Bintliff 1992. Douzougli 1998; Phelps 2004; Alram-Stern 2007; Sampson 2008. Bintliff 2005, 551-555.
2009
27
Figure 32. Hyettos rural; sites, with the ancient city shown as a white survey grid (E. Farinetti)
28
habitation: Grid area O18 and site CN 3 (Figure 32). Area O18, located to the north of the acropolis of Hyettos, turned out to be an important area for the MH period as well as for the transitional period from the MH to the Mycenaean period, the latter being scarcely identifiable within survey collections. All diagnostic samples of this period belong to finewares of excellent quality. The middle, classical phase of MH is attested by fragments of Grey Minyan goblets and kantharoi (Figure 33: 3-6),27 whilst the transitional MH-LH period is identifiable by fine burnished low-stemmed goblets (Figure 33: 7)28 and craters. A semi-coarse rim fragment with plastic rope decoration below the rim (Figure 33: 1) can be dated to the period from Final Neolithic to EH.29 The second site with a strong prehistoric habitation is CN 3, a deserted medieval village, located to the north-west of the city of Hyettos30. Some EH I red-slipped and burnished sherds were collected here but also one possible coarse FN rim. There is also evidence of obsidian tool production. The MH period is attested by semi-coarse fragments of diagnostic shapes, like Grey Minyan kantharoi and goblets, dated to the MH II-III periods (Figure 33: 3-4). The LH period is indicated here by some sherds of wheel-made burnished wares. A few sherds of highly polished handmade coarse wares date probably to the Neolithic period. FN coarse straw-tempered sherds were collected from the Roman villa site CN 7, from site CN 9 and from the grid area O19, which is located to the north of site CN 3. Sherds of EH gritty fabrics were collected on the site CN 5, located approximately 1.5 km away from the city to the north-east of the surveyed area, as well as from the adjoining area P 20. At site CN 11, located at the same distance from the city but to the north-west of the surveyed area, we were able to identify some MH Grey Minyan sherds. In the area N19 to the north-west of site CN 3, MH sherds and some sherds of possible FN crusted wares were attested (Figure 33: 2). The representation of the Mycenaean period seems to be very low outside the city. In the grid area P18, which is located at the north foot of the acropolis, a fragment of a LH IIA squat rounded alabastron with well-preserved painted decoration of ivy and rock patterns was found (Figure 33: 9).31
For the local Boeotian Grey Minyan, see Goldman 1931, figs 185, 187 nos. 3-4; Sarri 2010, 110-124, 90-100. On the chronology of the transitional low-stemmed goblets, see Dietz 1991, 166, fig. 51. 29 Phelps 2004, 116 fig. 61 nos. 1, 5-6. This is equivalent to the LN IB period according to Sampson; cf. Sampson 2008, fig. 127 no. 1468. In Ayia Triada in Aetolia similar pottery was found in an early phase of EH I: Dietz & Moschos 2006, 42 fig. 11 no. 121. 30 Bintliff 1992, 23. 31 For the shape see Mountjoy 1986, fig. 19 no.1.
28
27
2009
29
Figure 33. Prehistoric sherds from the Hyettos off-site area
30
Figure 34. Thespiai Magoula site map (E. Farinetti)
Thespiai Magoula One of the main aims of the 2009 study season was a preliminary examination of the Neolithic pottery collected on the Magoula of Thespiai, one of the most significant Neolithic sites in Boeotia, investigated by the Cambridge-Durham team in the 1980s. The site was known already from pioneer research in the early 20th century.32 An assemblage of Neolithic figurines was published by G. Bass33 and a small group of sherds by Caskey.34 In the Aegean Bronze Age Gazetteer the Neolithic, EH I, EH II-III, MH and LH IIIA-B periods were all attested.35 The new survey material derives mainly from the summit and the west slope of the Magoula which have not been investigated within the formal city survey grid system of 1984-85, but which were separately sampled as a sub-site within the urban survey at Thespiai (Figure 34). The periods represented are Middle, Late
32 33 34 35
Caskey 1951, 289. Bass 1959. Caskey 1951. Hope Simpson & Dickinson 1979, 248; see also French 1972, 12, 14 fig. 3.
2009
31
and Final Neolithic. Since the study is still ongoing, it is not yet clear if the Magoula was inhabited during the Early Neolithic period. The surface material is fairly rich and consists of well-preserved sherds and larger fragments of plain and decorated pottery of excellent quality, as well as of fragments from household vessels. During the 2009 season we focused on the preliminary dating of the assemblage in comparison to the Neolithic pottery classes known from Thessaly and the Peloponnese. The comparison with the Neolithic pottery from Halai36 and the pottery coming from the Magoula Balomenou in Chaeronea, displayed in the new exhibition of the Chaeronea Museum,37 helped us towards a more refined understanding of the local pottery fabrics and technologies. Concerning the recognition of type fossils, we were able to observe that the MN period is dominant, represented by much pottery with red painted decoration assigned to the Neolithic Urfirnis (Figure 36: 1-2)38 and to the red-on-white, so-called Chaeroneia Ware.39 The LN phase is also well represented by sherds of the black-polished early LN Larissa ware (Figure 36: 3-4),40 as well as by a large number of matt-painted sherds (Figure 36: 5-6).41 The FN is represented by sherds of coarser wares, some of them bearing plastic rope decoration (Figure 36: 7, 9).42 Beside the recognition of established stylistic groups, which concern mainly the fine-tempered decorated pottery, we were able to define local standardised ceramic technologies, such as surface treatments and the use of various tempering materials. In this way we attempted to establish a local classification guide mainly based on fabric, which could be linked in terms of terminology to better investigated regions like the Argolid.43 By analysing the rich and diagnostic Neolithic material of the Thespiai Magoula we hope to refine the pottery classification and the dating of other sites with fewer and less diagnostic surface finds. Thespiai Neolithic fabrics Concerning the materials used as tempering, we can distinguish two major types of inclusions: the straw-tempered and the lime-tempered (the latter often showing
36 We would like to thank warmly Professor J. Coleman for showing us the Halai study collection and for his valuable advice concerning the ceramic sequence of the Neolithic period. 37 We are also indebted to Professor C. Perls for her valued advice on the Neolithic finds, as well as to Dr I. Fappas and Mr N. Liaros for their assistance in the Chaeronea collection. 38 Phelps 2004, 17, 61. 39 The fabric is also known as A3b ware after Wace & Thompson 1912, 199-200 fig. 140 a-e; cf. French 1972, 6-7. 40 For the chronology of the fabric, see Gallis 1987. 41 Phelps 2004, 87. 42 Phelps 2004, 116. 43 Douzougli 1998; Vitelli 1993, 1999, 2007.
32
large lime inclusions).44 The surface treatment can be slightly rough or polished, the first comprising two very distinctive groups, brush-burnished (Figure 35: a) and facet-burnished (Figure 35: b). A similar effect is given to the vases with rippled burnish from Lerna.45 On the surface of the brush-burnished there are clear traces of a brushing implement that has been moved in different directions. The facet-burnished sherds on the other hand show parallel marks of a flat and hard burnishing tool. The highly polished examples are usually fine-tempered and fired to red or buff and sometimes they are given a very thin slip. Coarse polished examples are, however, also quite numerous. A particular kind of highly polished pottery recognised in the assemblage (Figure 36: 3-4) belongs to the black polished A5g46 or Larissa ware.47 A beading of the surface is attested in two cases (Figure 36: 4).48 The application of a slip is mainly observed on fine fabrics with red- or creamy-surface but also on a semi-coarse dark grey variety. The painted pottery can be attributed to various known styles dated to the MN and LN periods. However, the majority of the sherds preserve only small parts of the decoration, preventing a more accurate dating. Both lustrous and matt-painted decoration are represented. The colours of the paint used are black, brown, red and red-orange. There are also bichrome examples with red and black decoration. A specific MN pottery class, usually defined as Urfirnis, is well represented in the assemblage.49 Different styles of this class are easily discernable. A monochrome brown-orange subgroup which we named highly polished shows close similarities to Vitellis MU-Monochrome Urf ware.50 Another fine-levigated and hard fired fabric, the Red-Brown Streaky Slipped might also be related to the same category.51 The most distinctive Urfirnis fabric is the so-called Corinthian Urfirnis pattern painted.52 In Thespiai this class is abundantly represented and appears sometimes covered with a solid or streaky reddish-brown paint over one side of the pot, whereas the other side is decorated with a solid or semitransparent paint (Figure 36: 2). The LN matt-painted class (Figure 36: 6), corresponding to the Thessalian Tsangli and Otzaki wares,53 constitutes a large group within the assemblage.54
44 45 46 47 48 49 50 51 52 53 54
Vitelli 2007, 7, table 1.2, 75-78. Vitelli 2007, CD Photo 8, 37. Wace &Thompson 1912, 15. Hauptmann 1981, 75-76; Gallis 1987. Cf. Vitelli 2007, CD Photo 43. Phelps 2004, 44-47. Vitelli 2007, CD Photo 25-27. Vitelli 2007, CD Photo 17b. Weinberg 1962, 179-182; Phelps 2004, 56-59; Vitelli 2007, 105-106. Hauptmann 1981, 24-33 fig. IV, 1-9. Phelps 2004, 87-88; Vitelli 2007, 111; Sampson 1993, 67-83 figs 22-27; Sampson 2008, 112.
2009
33
Figure 35. Thespiae Magoula. Burnishing techniques
Open and closed vessels are represented in approximately equal numbers during this period. The bichrome variety of the matt-painted ware is also attested.55 The surface finds of the Thespiai Magoula contribute very usefully to an outline of the Neolithic production of Central Greece, known from Chaeroneia,56 Elateia,57 Orchomenos,58 the Sarakenos cave59 and the Corycian cave.60 The site follows a development similar to many of the other contemporary sites: after an apparently missing EN period, continuous habitation is attested on the Magoula from the Middle Neolithic to the Chalcolithic/Final Neolithic period. A statistical analysis and a refined study of the pottery classes, planned for next years season, should give us a full picture of the pottery from this significant Neolithic site. Koroneia 2009: surface survey sampling experiments (Mark van der Enden)61 During initial analysis in April 2009 of the pottery from the city of ancient Koroneia (which is mostly Hellenistic/Roman in date), conducted under the auspices of Professor J. Poblome, Professor V. Stissi and Dr P. Bes, it had become apparent that few sherds of cooking ware fabrics were picked up during field survey. Preliminary inspection of the Koroneia assemblages showed an abundance of slipped
55
Kunze 1931, 43 fig. IV, 1-2; Touchais et al. 1981, 150-154 figs 30-31; Sampson 1993, 83; Vitelli 2007, CD Photo 50; Sampson 2008, 157-159. 56 Soteriades 1908. 57 Weinberg 1962. 58 Kunze 1931. 59 Sampson 2008. 60 Touchais et al. 1981. 61 In close cooperation with Professor J. Poblome, Professor V. Stissi and Dr P. Bes.
34
Figure 36. Thespiae Magoula. Neolithic sherds
2009
35
wares. We also noticed that these sherds were quite sizeable and were dominated by large diagnostic fragments. The few cooking wares attested in what are assumed to be domestic contexts, suggested that either the material is not there62 or that we are missing it. We came to believe that besides the obvious processes of deposition and postdeposition which possibly might have had a more severe effect on cooking wares, a bias in our sampling strategies was a real possibility. The fields encompassing ancient Koroneia are so rich in material and yield such staggering amounts of pottery, that it is a daunting task indeed to select a representative sample from each standardized grid of 20 20 m. Furthermore, as cooking ware fabrics are generally more brittle then their fine-ware counterparts, they break more easily into small pieces. These small fragments are easily swamped by the thousands of plain and slipped wares covering our collection units, not to mention the hundreds and hundreds of tile fragments. The eye of a student is also very likely to be more easily attracted by a sizeable slipped rim fragment of a plate, than a small body sherd of a cooking pot. All in all, we have strong reasons for believing that our sampling methods systematically underrepresent the cooking wares in our assemblages. In order to verify the presence of this supposed bias and its effects, we planned a sampling experiment for the 2010 season.63 This experiment was specifically designed to check if our normal sampling methods where systematically biased against cooking wares. The field team was asked to give extra emphasis to the recognition and collection of sherds of cooking ware fabrics. It was hoped that an increased attention to cooking ware fabrics would raise their participation in our ceramic sample. In addition, we conducted a sampling experiment. In order to compare the results of the standard sample collections from our 20 20 m grid units, with the finds actually lying on the surface, a total collection sampling strategy was devised. Within five standard 20 20 m grid units (from which the usual general sample collection was made) we also created 4 4 m total collection units, from which all the surface material was picked up, subsequently analysed and kept. Items smaller than a fingernail where excluded from analysis. Even though a 4 4 m unit is by no means representative for the pottery present in a 20 20 m unit, we feel it does provide an indication of what is really on the ground in a designated
62 Possibly the result of patterns of discard and/or re-use, or post-depositional processes. It is also possible that our expected ratio of cooking ware does not correspond in the case of Koroneia to actual patterns of use. 63 Many thanks to Professor Bintliff for his encouragement and assistance. Thanks as well to Leiden students Leon Theelen, Xerena Jue and Manon Degenkamp for their enthusiastic participation and assistance.
36
area compared to what we normally sample. Moreover, the densities commonly encountered in ancient towns preclude total collection studies from complete grid units. The total collection units provide us with the opportunity to see if more cooking wares can be identified in comparison to the quantities normally recovered from our standard collection units. The question was: when all the material was brought home from a designated area, would a more balanced assemblage appear? In total, five total collection units were taken from their respective standard 20 by 20 m grid units (city units: 358, 359, 360, 468 and 525; cf. Figures 37-44), the results of which were compared. The analysis of the total collection units and their parent units (the 20 20 m city grids) was conducted on a preliminary level; only rough chronological, morphological and functional classifications were made. It was decided that in order to form a first impression of the results and possibly adjust procedures in the field, a crude division was necessary between tiles (slipped and plain), slipped ware, plain ware, cooking ware and coarse ware. In this way it would be possible to measure the percentage of cooking ware present, in relation to the total assemblage. Not only the percentage of cooking ware but also the relative proportion of the other categories like slipped and plain wares could be established and compared. 2010 collection experiments: results In what follows, the analysis of the total collection units will be presented and contrasted with the material coming from the standard collection grids. The pie charts (see the Appendix to this article on page 59) present the data, both numerical and in percentage, of both standard collection units and total collection units (designated as unit number + suffix TC). Furthermore, additional analysis documented the number of diagnostic sherds present in both standard collection and total collection unit (except for units 360 and 360 TC). These results are presented in additional pie charts. This preliminary analysis of results will take the form of a general discussion of the data, followed by some provisional conclusions. The data will be discussed per ceramic category (e.g. slipped ware, cooking ware etc.). Cooking ware The percentage of cooking ware present in the standard collection units varies from 2 to 17%. However, the 2% in unit 359 seems to be an exception. Unit 358 registers 17% of cooking ware, units 360 and 468 have both 13% of cooking ware and unit 525 11%. If contrasted with the total collection units, we can see that the percentage of cooking ware varies there from 9 to 17%. The 17% of unit 358TC
2009
37
Figure 37. 358TC (Photo Leon Theelen)
Figure 42. Cooking ware 359TC
38
Figure 43. Slipped ware 360TC
Figure 44. Coarse ware 358TC
forms the upper end of the range. It can be deduced from the pie charts that the percentage of cooking ware for the other total collection units ranges from 9% for unit 525TC, 10% for unit 359TC, 11% for unit 468TC to 16% for unit 360TC. The general impression one gets, if both the total and normal city grid units are compared, is that the cooking ware percentages are comparable. Seen in percentages, there is no great boost to cooking ware in the total collection units. What is interesting is that in the first three total collection units, the percentage of cooking ware present is higher than that recovered in the normal standard grid units. This changes, however, with the last two total collection units, where we can see that the percentage of cooking ware stemming from the standard grid units is slightly higher. This is very likely the result of our improved collection strategies which emphasize the picking up of cooking ware in the standard sample collections. It also seems logical that generally a more representative picture will emerge from a 20 20 m grid than from a relatively small 4 4 m unit. Slipped ware The percentage of slipped ware present in the standard collection units varies from 28 to 61%. In the first three standard units the following figures were documented: 53% for unit 358, 61% for unit 359 and 49% for unit 360. This contrasts with the last two standard units: 28% for unit 468 and 31% for unit 525. The results for the total collection units are as follows: 33% for unit 358TC and 359TC, 36% for unit 360TC, 15% for 468TC and 13% for 525TC. These results seem to justify our growing fears about a substantial bias in our sampling methods. It appears that slipped ware is generally overrepresented in our standard sampled assemblages. Comparison with excavated assemblages seems to indicate an overrepresentation of around 20%.64
64
For Halos, see Reinders & Prummel 2003; for Halieis, Ault 2005.
2009
39
The decrease in percentage of slipped ware, documented both in the total collection units and in the standard units from grid to grid, seems to indicate that there was generally less slipped ware on the surface in one particular area and does not provide evidence for an improvement in our collection methods (in which we would have deliberately picked up less slipped ware, in favour of other categories). If compared to excavated material, however, an assemblage including circa 30% slipped ware seems reasonable. It might thus be that the total collection units in this case actually underrepresent the amount of slipped ware present in the 20 by 20m units. As explained before, this is just the result of the spot where one happens to position the 4 4 m unit. The last two standard collection units seem to provide a more balanced picture, with respectively 28 and 31% of slipped ware. Plain ware The percentage of plain ware present in the standard collection units varies from 20 to 46%. In the first three standard units the following figures were documented: 28% for unit 358, 21% for unit 359 and 20% for unit 360. This contrasts with the last two standard units which give the following results: 45% for unit 468 and 46% for unit 525. The results for the total collection units are as follows: 41% for unit 359TC, 53% for unit 359TC, 45% for unit 360TC, 74% for unit 468TC and 73% for unit 525TC. Again we can document an increase in the last two collection units, both standard and total. We believe we have here a similar situation as with regard to the slipped ware. The increase of plain ware in the last two collection units, both total and standard, seems to represent a genuine increase in the amount of plain ware present. Again we feel here that the standard units provide the more balanced picture, one which can be more easily compared with excavated domestic assemblages. Coarse ware The percentage of coarse ware present in the standard collection units varies from 1 to 16%. The results are as follows: 2% for unit 358, 16% for unit 359, 14% for unit 360, 1% for unit 468 and 3% for unit 525. The total collection units provide the following picture: 9% for unit 358TC, 4% for unit 359TC, 3% for unit 360TC and 0% for both unit 468TC and 525TC. It seems that once again the standard units provide us with a better sample of the steady but minor presence (in percentage) of storage vessels and pithoi, as is for example attested in excavated assemblages. Experience from the field shows us that in each and every standard collection unit we normally encounter fragments of pithoi and storage vessels. Obviously, as these fragments are often still quite sizeable, one has more chance to encounter them in the normal city units than in a total collection unit of 4 4 m.
40
Diagnostics versus non-diagnostics65 In addition to comparing the assemblage composition of the different units, we have compared the rate of diagnostics versus non-diagnostics for four standard units and their total collection counterparts. It was our intention to verify if by focusing only on diagnostic material a similar picture would emerge as when all sherds were collected. We had the feeling that simple body sherds even if they have some identifiable features, like slip, do not add much to the overall picture. In our samples in particular, it seemed likely that large amounts of slipped bodysherds, easily distinguished by student collectors as potentially datable, might produce their overrepresentation66 and also make finds processing problematic, because of the huge bulk of material that is collected. The results of a comparison of diagnostics versus non-diagnostics, as displayed in the pie charts, show that the standard units exhibit more diagnostics in each ceramic category then their total collection counterparts. The assemblage of the standard collection units looks also quite balanced and compares better with the results of the entire assemblage when also the non-diagnostics are taken into account. It also seems that when we just focus on diagnostics, the overrepresentation of slipped ware is less visible in the standard collections. Once again it seems that in the last two units there is a real decrease in the presence of slipped ware and an increase in the amount of plain ware, which explains the lowered participation of slipped ware in the total collection and standard units. It can thus be suggested with caution that focussing on diagnostics only seems to produce a similar picture as to when both diagnostics and non-diagnostics are collected. Body sherds do not provide much added information, in this particular case their inclusion does not seem to alter the chronological picture. Their inclusion seems, with regard to the composition of an assemblage, only to inflate the participation of slipped ware. The material from the five collection units analysed seems to span a relatively short time-period. Most sherds seem to belong to the Hellenistic and Early Roman eras. A focus just on collecting diagnostics becomes problematic, however, when more periods are involved. It is then questionable if diagnostics alone will be enough to provide a representative picture, and there is a real danger of missing out on certain periods, or underrepresenting them (e.g. prehistoric or medieval periods). Preliminary conclusions The results based on the total collection units indicate that with regard to cooking ware, they do not give a dramatically different picture than that obtained from the
65 66
Diagnostic in this context refers to rims, bases and handles. Non-diagnostics are body-sherds. As mentioned earlier there is a distinct tendency of field walkers to focus on these wares.
2009
41
standard grid samples. Generally, both total collection units and the normal city grid units have a comparable percentage of cooking ware. The question then is whether 11 to 13% of cooking ware is enough for a functional domestic assemblage? When compared to excavated domestic assemblages this does not seem to be the case. There cooking ware generally makes up around 25/30% of an assemblage.67 The next question then is, what happened to the cooking ware of Koroneia? It seems unlikely that we are still missing it, considering the total collection units and the increased attention of field surveyors. It has been suggested that at Koroneia a lot of shapes, for example jugs, which elsewhere are executed in what are generally classified as coarse/cooking ware fabrics, are made in plain ware fabrics.68 This might also explain the large percentage of plain ware in the assemblages, although this is to be expected. The amount of diagnostic plain ware is unfortunately small. The reason why we so far have relatively limited amounts of cooking ware at Koroneia remains still somewhat elusive. However, the answer probably needs to be sought in a combination of locally specific patterns of production, use, discard, deposition and post-depositional processes. In summary, it seems that where the first three total collection units provide a more balanced assemblage, the situation is reversed when the last two total collection units are considered. In their case, the normal collection units provide a more balanced picture. This, after all, taking into account the expected variation within and between grids and total collection units, might be the result of our increased focus on picking up cooking ware and our realisation that we tend to overrepresent slipped ware. The 2009 field season saw the analysis of five total collection units. This is obviously a very small sample, and in order to balance out the variation that exists within each individual 20 20 meter unit, more total collection units in new grids are desirable in order to test our preliminary observations. We therefore hope to continue similar experiments in the near future in order to continuously advance our understanding of sampling strategies and techniques and their influence on data patterning. Boeotia 2009 soils and land use report (Robert S. Shiel) Northern Tanagra region As much of the area around Tanagra city and the villages of Kleithi and Ayios Thomas had been investigated last year,69 the centre of attention this year was
67 68 69
See Van der Enden 2008. Pers. comm. Professor V. Stissi. For comparable evidence at Corinth, see Slane 2003. Shiel in Bintliff et al. 2008, 53-58.
42
moved to the area north of the motorway. This area can be divided into four parts: 1. the Ritsona basin and its western extension, 2. the northern uplands, 3. the broad coastal plain and 4. the badlands (Figure 45). The Ritsona basin and hills The basin lies north of the E75 Athens-Thessaloniki motorway and contains about 18 km2 of cultivated land lying between 160 and 240 m (Figure 46). To the west and north rise substantial limestone hills reaching over 880 m. The basin is well watered by numerous streams issuing from the limestone, and the land use is mixed, but cropping is dominated by cereals and grapes with some areas of pistachio and only small groves of olives. There is relatively little uncultivated land or woodland other than along streams. Soil is mapped as rendzina,70 but much is strongly reddened and is similar to terra rossa; soil surfaces seem stable. Grapes were seen to be wilting, indicating that even good soil and available water are insufficient in strong sun and wind. Like other basins containing reddened soils (for example Pavlos in northern Boeotia), this appears to be an area of stable freely drained deep soils and, in this case, with a better supply of water than at Pavlos. The surrounding limestone hills have a few small isolated fields in valleys corresponding to the eroded rendzinas and terra rossas described by Zvorykin and Saul71 and there appears to be occasional flooding, and probably soil movement, into the Ritsona basin from some of these valleys, judging by the woodland surrounding some of the streams, particularly from the west. The proportion of such field areas ranges from zero to the north-west to about 25% to the north-east. Outside these areas the land offers only browsing for sheep and goat flocks. The limestone has little vegetation cover, some of which had been recently burned, though there was some pine woodland on the lower slopes. Outside the arable area the vegetation would only provide very poor grazing for small numbers of goats except around Avlida where there is more vegetation. The poor vegetation cover is typical of the hills to the north of Kopais where the vegetation is much sparser than on the south side. Mykalessos within the basin was examined by Fossey who noted Ritsona was surrounded by a fair-sized fertile basin72; clearly nothing much has changed. This is an attractive area offering not only a substantial area of good soils but adequate ground and stream water to grow crops and a surrounding area of woodland and grazing as a source of fuel, animal products and manure from the animals. It seems possible that some land has been lost by erosion from the fringes
70 71 72
Zvorykin & Saul 1948. Zvorykin & Saul 1948. Fossey 1988, 80-85.
2009
43
Figure 45. Satellite image showing the four major subregions of northern Tanagra
Figure 46. View to the northwest across the Ritsona basin
Figure 47. Badlands area looking north from the motorway
of the basin and the lost soil has deepened that within the basin. It is also possible that the removal of woodland has resulted in soil loss, possibly by vertical erosion into the limestone basin from the surrounding hills. The overall change with time though does not seem to be large. The badlands To the southeast of Ritsona basin is a large area of badlands. These are deeply dissected with steep valley sides and, though vegetated, show all the evidence of
44
continuing and recent as well as ancient dissection of a large plateau of soft sediments lying between 140 and 220 m (Figure 47). The total area is about 12 km2 with large valleys trending predominantly east the southeast part is much more heavily dissected than the north-west. Some of the area to the northwest has been contour terraced but there is no evidence of subsequent cultivation. Apart from one or two fields on the surviving level plateau tops, used for olives or for tillage, there is no evidence of recent usage. This is the largest area in Boeotia of active erosion that appears to have led to the land being abandoned. Soils are mapped as rendzina by Zvorykin and Saul but are not differentiated from other areas of soil in surrounding areas which are cultivated and appear more stable. The vegetation cover over most of it is steppe grassland and maquis. It provides, no doubt, grazing for goats and sheep that is better than on the limestone hills, but overgrazing must be avoided to prevent renewed erosion. The underlying causes of the instability of this area remain to be investigated. It would seem that large amounts of sediment have been lost to the east, but there is no clear evidence of where this has been deposited, though the area to the south of Vathy is remarkably flat. Fossey does not refer to any former settlements in this area73 and it may be that it has never been cultivated or that any attempts have led to such severe erosion that use has been abandoned before settlement evidence could accumulate. The coastal plain To the east of the badlands is an extensive plain from Vathy to Dilesi covering some 25 km2 and lying between sea level and 140 m. Some of this area may have benefitted by erosion from the badlands described earlier, and the soils are classified as alluvial by Zvorykin and Saul (Figure 48). This area is, however, stable and intensely cultivated much of it has been recently urbanised and is laid out in narrow strip fields. Though it is crossed by numerous streams these are narrower than at Ritsona and less well wooded any areas of woodland appear to be land abandoned by their owners. The whole area has a more arid appearance. Apart from small areas of cereals about 30% the remainder is olive groves, though many of the plots are only wide enough for a single row of trees. There are no rock outcrops south of the hill at Vathy and the soil is less reddened than at Ritsona. It would seem that the Ritsona basin is benefitting substantially from the presence of the large hills nearby, which may attract a larger rainfall in an otherwise very dry area, plus there is a better chance of ground or stream water continuing to be present for longer into the summer as water drains through the limestone. The coastal plain is lower and flatter and distant from the hills on
73
Fossey 1988.
2009 45
Figure 48. Soil map of the northern Tanagra region after Zvorykin & Saul (1948). Number 8 is the alluvium eroded from the rendzina (2), while 6 is the reddened rendzina of the Ritsona basin; the red and blue broad vertical lines show the area of eroded soil to the northwest, while 4/5 are the karst limestone areas of the limestone hills
46
Evvoia and, as a result, suffers from the hot dry conditions of the Tanagra plain. This is likely to have been a permanent feature of the climate of the area and hence, although soils are good and the area easy to manage, the lack of rainfall restricts the range of crops that can be grown barley is likely to be more successful here than wheat. Many of the olive groves were noted as well aged and suggest that olive growing has been a long standing feature of the area. Fossey notes that this coastal area has been intensely settled since at least Mycenaean times and that at modern Paralia Avlidos ancient masonry was buried at a depth of 1.5 m.74 This last comment may offer support to recent losses of sediment from the badlands, which may have been initially water eroded and then redistributed by wind erosion. The presence of harbours etc. with a large agricultural hinterland and the city of Tanagra nearby may be the reason for the dense settlement of this plain, rather than the good quality of the soil, bearing in mind the dry climate.75 Koroneia In 2008 the area adjacent to the two main routes to modern Koroneia and to Ayia Triada villages had been examined, so this year the survey was extended to the west of the road to Ayia Triada (Figure 49: 1 and 2) and also to the area around the ancient acropolis (Figure 49: 3). From Ayios Georgios a metalled road ran west towards a chapel at the foot of the steep slopes to the west and north respectively. The southern boundary is the west-east streamline to the south of Ayios Georgios. The vegetation was predominantly olives which were well established and of some age. The land was all terraced on moderate slopes with a deep soil overlying flysch where this was exposed. The soil was slightly stony, with stone and rocks increasing towards the steeper slopes. Where rock had occurred within fields it had been pushed out of the way but had not been cleared off or used. The terrace steps were about 1 m high without stone reinforcement. Soils appeared to have been largely rendzina with a varying degree of reddening, but due to the earth movement in terrace construction any profile characters had been strongly altered. Weeds and grass had grown well and on most of the plots the grass had been cut (Figure 50). Approaching the steeper areas the soil cover over the limestone became less complete and the land use changed to maquis vegetation providing grazing with large goateries. There was little bare rock exposed and large areas of continuous soil with grass or maquis were available The demarcation between the olives and maquis is very abrupt and within the olive groves there is only woodland along narrow streamlines and one small wood of less than 0.2 ha in an area
74 75
Fossey 1988, 66-67. Shiel in Bintliff et al. 2008, 53-58.
2009
47
Figure 49. Areas examined to the west and south of Ayios Georgios (1 and 2) and around the ancient acropolis of Koroneia (3)
Figure 50. Terracing with olives to the west of Ayios Georgios
48
of some 70 ha. Streams had vegetation associated with wet conditions but the only stream running was close to Ayios Georgios and may have consisted of waste water. The land use potential of this area is much better than the current olive production and Fosseys comment that the area around Koroneia was used for cereals76 could apply today. The use of the land for olives seems to be based more on social or economic forces rather than the potential of the land. Area 2 has the road to the church of Ayios Taxiarchis. This is an area of rolling country and once again the vegetation is predominantly olives on terraces. The soils are less reddened than near Ayios Georgios and are more stony, but the majority is loose surface stone sometimes this has been gathered onto the terrace faces. The pattern of use is similar to area 1 but with more maquis and with steep slopes of the surrounding limestone hills having more rock outcrops (Figure 51). On the whole these are younger olive groves than near Ayios Georgios. Apart from a few arable fields on level land close to the main road from Ayios Georgios to Ayia Triada and on level land close to Ayios Georgios, there are no arable crops being grown, a complete change from Fosseys description of the area. From Ayios Taxiarchis to Ayia Triada land use in the whole of the rolling country below the limestone mountains is over 85% olives with the rest maquis. The same is true of the area from Ayios Taxiarchis north to the streamline delineating the southern boundary of area 1. The total area of the olive groves in area 2 is some 540 ha. The only change in this area is that as the slopes become steeper close to Ayia Triada, more of the terraces have stone faces, some of which revetting has collapsed. These appear in some cases to be older olive groves than those further down the valley. The only stream that had running water was that coming from Ayios Georgios, but the dry streams all had vegetation such as plane trees and bamboo associated with water being available at least at depth if not at the surface. The area around the Koroneia acropolis is, like areas 1 and 2, predominantly used for olive growing. However the proportion of scrubby maquis is much higher, there is more steppe grassland and many of the olive groves are relatively young this fits well with Fosseys comment that the area was cultivated for cereals when he visited it. This previous use is evidenced by the very rounded tile fragments which are found in large banks over parts of the area, especially as here in the waster zones of the citys ancient potters quarter (Figure 52). There is some terracing and the terraces often have stone faces built with stone and tile from the ancient buildings. Most of the soil is rather grey rendzina with a large content of small natural stone, if it does not contain stones from buildings. It is more rolling than the land nearer Ayios Georgios and some of the slopes are relatively steep though not as steep as those near Ayia Triada or
76
Fossey 1988, 324-330.
2009
49
modern Koroneia. To the west of the acropolis some 70% of the area is regular olive groves which are generally well established but not old. The remainder is the woodland along the main stream draining the valley and areas of scrubby maquis, often on areas which have rock outcrops. Across the stream between the acropolis and Ayios Georgios to its south-west is the only substantial Figure 51. Olive groves in area 2, looking south remaining area of cereal production in the valley. To the north of the acropolis is a very regular area of olive groves with over 90% of the land area occupied thus; there are odd fields of cotton and the area of these increases towards former Lake Kopais. To the south-east is an area of well established older olive groves on terraces, pocket terraces and flat planting. The acropolis itself has the lowest proportion of used land some 50% is scrub and grassland while the remainder is olive groves. The only substantial unused area is to the south of the acropolis where there is a small monastery; the hill on which it sits is entirely maquis scrub. Area 3 therefore provides today a substantial area of olive grove but Figure 52. Terrace on the acropolis at Koroneia clearly in the past has been used largely facing north, showing grey soils and large amounts for cereal growing. Like areas 1 and 2 of rounded tile fragments from pottery producthe soil is not a limit to production tion wasters and, though irrigation is impractical on the slopes, the soil is relatively deep, much has been terraced and the climate is relatively moist. Water availability is best on the level land close to Ayios Georgios and on the south fringe of Kopais. When the lake was in existence more water would have been available, though the lowest lying land is known to have been flooded at some times. The only permanent stream is the one from Ayios Georgios that flows past the acropolis and waters the more level lower land; it is being used for this purpose at present. Presumably there would have been no land
50
available for farming on the acropolis during its occupancy but the surrounding area between the edge of Kopais and the monastery to the south, and bounded by the two main roads up the valley, and which could be cultivated, is some 180 ha. To the east of the road to modern Koroneia village, the pattern of land use continues up the valley similar to that to the west. Below the acropolis is a broad triangular area of level grey rendzina with predominantly olive production but with some 15% of arable crops, such as cotton, mostly close to the old road from Livadhia to Aliartos. The total area is small, 110 ha, as the land rises steeply to the east of the road to Koroneia, and north of the acropolis the area of olives to the east of the road is usually only 600 m wide and narrows towards Koroneia. In this region there is a high proportion of scrub and maquis on the steep slopes although where there is deeper soil and less sloping land it is used for olives; much has been terraced. Near Koroneia itself the land use pattern changes to a system of much smaller fields which have probably been hand-cultivated in the past and have a natural shape rather than the rectangular larger fields lower down the slopes. The area of larger fields is about 60 ha and of smaller ones 20 ha. To the west below Koroneia the pattern of small fields is continued, covering over 50 ha and cut by two wooded streams descending to the main drainage system. Separated from the lower valley by the hill and monastery described above there is a broad area of olives in larger fields on rolling ground to the west of these small fields and occupying about 120 ha. These fields are often isolated between larger areas of uncultivated land about 50% of the total. To the west of this is a strip facing the Ayia Triada road some 700 m wide and occupying some 130 ha of olives in intermediate sized fields with little maquis or scrub, and with about 10 ha of cereals on level land close to the Ayia Triada road. It would seem that the larger fields are probably modern clearance of land which may not have been used before, while the intermediate sized fields are relatively older but were probably animal cultivated rather than the hand cultivated tiny fields below Koroneia. All of these have more of the stony greyish rendzina soil. The recording of Greek vernacular architecture (Chiara Piccoli) During the 2009 summer season of the Boeotia Project, a survey has been carried out that aimed to record the last remains of traditional houses that is, from the Late Ottoman period onwards in two villages of the Boeotia region, Mazi and Evangelistria. The study of these houses is related to the field of so-called vernacular architecture, a term that was first used in the middle of the 19th century to define traditional rural buildings of the pre-Industrial period.77 The urgency of
77
Sigalos 2004, 12.
2009
51
the recording is motivated by the fact that nowadays there is a rising tendency in modern Greece to demolish old buildings and replace them with new ones. Since traditional houses are quickly disappearing, a large body of local knowledge about Greek culture, history and past ways of living is being swept away along with them. The work here briefly presented, represents an advance on our previously published method of recording such structures through two-dimensional photography pasted onto a three-dimensional AutoCAD house-model.78 The current approach relies on a Robotic Total Station, allowing a three-dimensional laserbased recording, and on far more elaborate graphics computer programmes than were available for the previous experiment in the Boeotian village of Ayios Georgios (see below). Due to recent developments in recording architectural evidence, pictures and maps do not suffice to keep a full image of the houses to be documented. For documentation purposes, three-dimensional models are taking over the role of pictures and handmade drawings in architecture and archaeology. They have in fact a huge potential for both of these disciplines, because they do not merely add a supplementary aesthetic factor, but they are also useful tools for interpretation, visualisation and communication of the recorded data. Especially when we come to interpretation, the addition of volumetric characteristics allows us to obtain a much deeper understanding of the recorded entity. 3D-models of buildings are also now proving to be very useful in focusing on the diverse shapes of the formal spaces within a house, taking us further into investigating the function and mental significance of such subdivisions. A next advantage is that three-dimensional reconstructions are also important as tools of visualisation. Firstly, they allow us to summarise many different types of information within a single model with a multi-scalar approach. Secondly, a 3D-model permits further manipulation of data, that is not static, as are notes and sketches drawn on paper, but that can be updated and transformed. Thirdly, their digital nature allows an easier circulation of data within the scientific community, thus stimulating a more dynamic debate on interpretations and methodologies. Finally, 3D-representations are a very effective means of communication, since they are appreciated by the public of nonspecialists, as immediately comprehensible visions of an ancient past. The Boeotia project has always encouraged the use of new and advanced methodologies to be experimented with in the field. In this case, the recording of traditional houses represented a good opportunity to introduce three-dimensional documentation within the project. A first attempt in this direction was made in 2007 during the summer season of the Boeotia project by Joep Verweij and a team composed of five members. The location was the village of Ayios Georgios, near
78
Verweij in Bintliff et al. 2007, 37-42.
52
to the former city of Koroneia, where two out of four surviving longhouses (the typical peasant houses with a broad faade) were recorded. However, as Verweij notes in the report of his work, the results, although representing a development on the flat documentation previously used for vernacular architecture, give still an overall sensation of bi-dimensionality, whilst there is also no integration between the recording of the structure and the 3D-reconstructions of the house interiors. In fact, only the walls of the houses that are reconstructed in AutoCAD are actually in 3D, whereas the faades and all the other inside and outside features are visualised by means of the 2D-pictures. Moreover, since the pictures that were taken inside the houses typically contained a number of objects that were placed in front of the wall, once these pictures were applied as texture on the 3D-model, they caused such objects to be flattened on the 3D-walls, thus creating an effect of distortion. Methodology The documentation of the houses in 2009 was performed with the purpose of collecting as accurate and complete as possible data that could provide a comprehensive image, not only of the structures themselves but also of the way of living. The houses to be recorded were located in villages and are part of private properties; moreover, some of them are still used (often not for housing but as a storage place), therefore a fast method had to be applied in order not to interfere with the everyday life of the inhabitants. These circumstances increased the difficulty of the recording, as speed had to be coupled with a high level of accuracy. Given these premises, the best way to reach our goal appeared to be a wide-ranging approach that makes use of digital recordings to collect and process the data, combined with interviews with the local inhabitant and any other available databases.79 The device that was chosen for documenting the architectural features of the houses is the Robotic Total Station (RTS). The RTS was recently acquired by the Faculty of Archaeology of Leiden University. Moreover, since the RTS is reflectorless, it was indispensable for the recording of the upper parts of the buildings. The RTS was used to record the three-dimensional coordinates of the most significant points of the house (e.g. edges of the building, windows, doors; Figure 53); the result of the recording session is a low density point cloud that provides the reference points from which to reconstruct the basic geometry of the building. While taking the measurements with the RTS, photographs have also been taken to be used as textures for the 3D-model after a process of rectification.
79 Dr Athanasios Vionis was extremely helpful with local interviews and the background of traditional life ways.
2009
53
Where present, also elements of the interior of the houses such as furniture, fireplaces, cooking tools and everyday objects have been photographed and recorded. In order to gain a broader understanding of the disposition of the furniture in the houses and their use, the owners of the houses have been interviewed, providing thus a deeper comprehension of the history of house Figure 53. Example of points recorded with the and household. These interviews conRTS; House in Mazi firmed the existing information regarding the division of space between animals and people in the longhouses and added new detail about the position and arrangement of beds, tables and chairs within the main room of the house. To process the data gathered during the survey, a hybrid approach has been applied that exploits techniques such as photogrammetry, computer aided design programmes (AutoCAD) and graphics computer programmes (Autodesk 3ds Max). The result of the recording with the RTS is a low density point cloud that has to be exported from TopSurv, the software that handles the data recorded by the RTS. The points were then processed in AutoCAD in order to extract the geometry and create the basic 3D-model of the recorded house (Figure 55). The 3D-model created in AutoCAD was then imported into 3ds Max, which allows a better management of 3D-objects. In 3ds Max, further three-dimensional features of the building have been added such as a characterisation of doors, windows and roof tiles in order to create a real three-dimensional structure. The same software allows for texturing the model by using the photographs that were taken in situ and rectified by means of PHoToPlan. The application of graphics computer programmes aims both to reconstruct in three dimensions the recorded building and to recreate the interior where possible. During the field work, in fact, the owners and the people from the village have been interviewed in order to collect their memories about what kind of furniture was present inside, where it was placed, and what kind of objects were used in the everyday life. A virtual environment can then be created that guarantees immediate understanding of the houses and that can be used as an interpretative tool for further comparisons and research. The application of this methodology proved to be very useful in the case of the longhouse that was spotted near to the church of Mazi (Figure 54). The structure, which used to be the house of the grandmother of the current owner, was still used as a storage place and a new building was joined onto it. From the memories
54
of the owner the date of construction could be placed around the early decades of the 19th century. When we arrived, the longhouse was about to be demolished in order to make room for a new building; this fact increased the urgency of compiling a complete and accurate documentation. The house was in fact very poorly preserved: part of the roof and of the walls had collapsed and Figure 54. Longhouse near the church in Mazi rubble was lying around; nevertheless, part of the original furniture such as the fireplace and a small closet were still in their original place, and everyday objects and cooking tools were kept inside; moreover, the partial collapse of the roof had made visible its phases of construction. This house therefore offered an unrepeatable occasion to record all these elements, which along with the important contribution given by the memories of the owner who used to live there when he was a child helped us to recreate in a virtual environment the aspect of the interior of the house as it used to be. Photographs were taken of the interior, including objects such as cooking pots, and tripods that had been used to support the cooking pots on the fire. When possible, these elements were measured and recreated in the 3D-model of the longhouse, aimed at a reconstruction that corresponds to reality as much as possible. Figure 56 represents the inside of the room according to the surviving items and the owners memories: he remembered, in fact, that there was a double bed at one side of the fireplace that was made of wooden boards that were supported by stones; whereas at the other side a couple of mattresses which were usually used by the children of the family had been placed on some straw on the ground. The closet, a low table and some small stools completed the furniture of the room. The owner explained that the furniture of the house did not have a fixed location in the room, but could be moved to other positions if the need arose. Furthermore, the childrens bed could be set down only during the night, thus saving space for daily activities. It has also been possible to collect further information about the way of life in the villages from a small museum that is housed inside the village school of Evangelistria. The exhibition displays old furniture and pictures and clothes and textiles that were collected from the inhabitants of the village, thus rescuing these testimonies of an almost forgotten way of living. The visit to the museum was very useful to get a more complete idea of the objects that were used in everyday life and to recreate them in the 3D-model. For example, the textures that have been
2009
55
Figure 55. The longhouse near the church in Mazi; 3D-model
Figure 56. Longhouse in Mazi; detail of the interior
used for mattresses and blankets in the virtual reconstruction of the above mentioned longhouse in Mazi (Figures 55-56) come from the pictures taken of pieces of textile that were exhibited in the museum as samples of handmade fabric. Discussion of the results Within the three weeks of the archaeological campaign, ten surviving buildings in Mazi and Evangelistria were entirely recorded with the above mentioned techniques. Two to three hours were needed for the complete recording of a house (including travelling time to reach the location and setting up of the device). It has to be noted that the time spent for the recording depends on the experience of the operator: since the procedure of recording is rather standard, a considerable saving of time is observed when the work is performed routinely. In our case, the recording also served the purpose of training students in using the device, therefore a balance had to be struck between speed of recording and educational purposes. Subsequent processing of the data and reconstructing the basic 3D-model took one morning for each house. However, such processing will also vary according to the amount of detail needed and to the experience of the computer operator. In conclusion, the method applied has proved feasible in terms of time spent in the field for the recording (which allowed us also not to heavily interfere with the daily life of the owners, who were anyway usually very pleased to be of any help), and time available for subsequent processing procedures, all in relation with the intended final results. The downside of using the RTS is that common to any laserbased device: in order for the RTS to precisely record the points in the space, no obstruction has to exist between the source of the laser beam and its aim. When the recording is performed in a village, the operator has to deal with fences, vegetation and cars passing by, all of which may hinder the functioning of the RTS. In this
56
case, having more than one recording position will usually solve such problems, although the time spent for the recording will then be slightly longer, as the RTS has to be taken down and reassembled. It is worth adding that it is still of use to take a few manual measurements to supplement the device-based recording. J.L. BINTLIFF Leiden University j.l.bintliff@arch.leidenuniv.nl B. NOORDERVLIET AND J. VAN ZWIENEN Leiden University bart@noordervliet.net jannekevz@gmail.com K. SARRI kalliope.sarri@gmail.com B. SLAPSAK Ljubljana University, Slovenia bSlapsak@yahoo.com I. UYTTERHOEVEN Leuven University Inge.Uytterhoeven@arts.kuleuven.be
M. VAN DER ENDEN Leicester University markvanderenden@hotmail.com C. PICCOLI Leiden University chiara_piccoli@hotmail.it
R. SHIEL Newcastle University R.S.Shiel@newcastle.ac.uk Acknowledgements
We wish to acknowledge the always helpful support of Professor Vassilis Aravantinos of the Thebes Ephorate, the staff of the Netherlands Institute at Athens, several members of the Byzantine Ephorate, and the generous hospitality of the bishop of Thebes and Livadhia for our project accommodation at Evangelistria. References
ALRAM-STERN, E. 2007. Das Chalkolithikum in Sdgriechenland. Versuch einer chronologischen und topographischen Gliederung. In: F.C. Reinholdt & J. Weilhartner (eds), Stfanov Aristeov. Archologische Forschungen zwischen Nil und Istros. Festschrift fr Stefan Hiller zum 65. Geburtstag, Wien, 1-10. AULT, B.A. 2005. The Excavations at Ancient Halieis II: The Houses. The Organization and Use of Domestic Space, Bloomington IN. AUPERT, P. 1990. L evolution des appareils en Grce a lepoque impriale, BCH 114, 593637. BASS, G. 1959. Neolithic figurines from Thespiai, Hesperia 28, 344-349.
2009
57
BINTLIFF, J. L. 1992. The Boeotia Project 1991: survey at the city of Hyettos, University of Durham and University of Newcastle upon Tyne Archaeological Reports, Durham, 23-28. BINTLIFF, J.L. 1996. The Frankish countryside in Central Greece. The evidence from archaeological field survey. In: P. Lock & G.D.R. Sanders (eds), The Archaeology of Medieval Greece, Oxford, 1-18. BINTLIFF, J. L. 2005. The Tanagra Project: investigations at an ancient Boeotian city and its countryside (2000-2002), BCH 128-129, 541-606. BINTLIFF, J.L., M. KUNA & N. VENCLOVA 2000. The Future of Archaeological Survey in Europe, Sheffield. BINTLIFF, J.L. & B. SLAPSAK 2006. The Leiden-Ljubljana Ancient Cities of Boeotia Project. Season 2006, Pharos XIV, 15-27. BINTLIFF, J.L., B. SLAPSAK, B. NOORDERVLIET, J. VAN ZWIENEN & J. VERWEIJ 2007. The Leiden-Ljubljana Ancient Cities of Boeotia Project. Summer 2007-Spring 2008, Pharos XV, 18-42. BINTLIFF, J., B. SLAPSAK, B. NOORDERVLIET, J. VAN ZWIENEN, K. WILKINSON & R. SHIEL 2008. The Leiden-Ljubljana Ancient Cities of Boeotia Project. 2008 season, Pharos XVI, 31-60. BINTLIFF, J.L. & A. SNODGRASS 1988. Mediterranean survey and the city, Antiquity 62, 57-71. BRUN, J.-P. 1993. La discrimination entre les installations olicoles et vinicoles. In: M.-C. Amouretti & J.-P. Brun (eds), La production du vin et de lhuile en Mditerrane. Oil and Wine Production in the Mediterranean Area, BCH Supplment XXVI, Paris, 511-537. CASKEY, J.L. 1951. Neolithic sherds from Thespiai, Hesperia 20, 289-290. CARINGTON SMITH, J. & S. WALL 1994. A Late Hellenistic wine press at Knossos, BSA 89, 359-376. DIETZ, S. 1991. The Argolid at the Transition to the Mycenaean Age. Studies in the Chronology and Cultural Development in the Shaft Grave Period, Copenhagen. DIETZ, S. & I. MOSCHOS 2006. Studies in the Bronze Age of Aghia Triada. A summary. In: S. Dietz & I. Moschos (eds), Chalkis Aitolias I. The Prehistoric Periods, Monographs of the Danish Institute at Athens 7, Athens, 38-63. DOUZOUGLI 1998. Aria Argoldov. Xeiropojtj keramik tjv Neterjv Neoliqikv kai Xalkoliqikv peridou, Aqna. FOSSEY, J. M. 1988. Topography and Population of Ancient Boiotia I, Chicago. FOXHALL, L. 1993. Oil extraction and processing equipment in Classical Greece. In: M.-C. Amouretti & J.-P. Brun (eds), La production du vin et de lhuile en Mditerrane. Oil and Wine Production in the Mediterranean Area, BCH Supplment XXVI, 183-200. FRENCH, D.H. 1972. Notes on prehistoric pottery groups from Central Greece (privately circulated). GALLIS, K. J. 1987. Die stratigraphische Einordnung der Larissa-Kultur: eine Richtigstellung, PZ 62, 147-163. GOLDMAN, H. 1931. Excavations at Eutresis in Boeotia, Cambridge MA. HAUPTMANN, H. 1981. Die deutschen Ausgrabungen auf der Otzaki-Magula in Thessalien III. Das spte Neolithikum und das Chalkolithikum, Bonn. HOPE SIMPSON, R. & O.T.P.K. DICKINSON 1979. A Gazetteer of Aegean Civilisation in the Bronze Age I. The Mainland and Islands, SIMA 52, Gteborg. KODER, J. & F. HILD 1976. Hellas und Thessalia, Wien.
58
KUNZE, E. 1931. Orchomenos II. Die Neolithische Keramik, Abhandlungen der Bayerischen Akademie der Wissenschaften, Philosophisch-historische Abteilung, Neue Folge 5, Mnchen. MOUNTJOY, P. A 1986. Mycenaean Decorated Pottery. A Guide to Identification, SIMA 73, Gteborg. PHELPS, B. 2004. The Neolithic Pottery Sequence in Southern Greece, BAR International series 1259, Oxford. REINDERS, H.R. & W. PRUMMEL 2003. Housing in New Halos. A Hellenistic Town in Thessaly, Greece, Lisse. ROBINSON H.S. & S. WEINBERG 1960. Excavations at Corinth 1959, Hesperia 29, 246-253. ROESCH, P., G. ARGOUD, A. SCHACHTER & G.VOTTRO (eds) 2007. Les inscriptions de Thespies, Lyon, dition lectronique. SAMPSON, A. 1993. Skotein Qarrounin. To splaio, o oikismv kai to nekrotafeo, Aqna. SAMPSON, A. 2008. The Sarakenos Cave at Akraephnion, Boeotia, Greece I. The Neolithic and the Bronze Age, Athens. SARRI, K. 2010. Orchomenos IV. Orchomenos in der mittleren Bronzezeit. Bayerische Akademie der Wissenschaften, Philosophisch-historische Klasse, Abhandlungen, Neue Folge 135. SCRANTON, R.L. 1941. Greek Walls, Cambridge MA. SIGALOS E., 2004. Housing in Medieval and Post Medieval Greece, BAR International Series 129, Oxford. SLANE, K. W. 2003. Corinths Roman pottery. Quantification and meaning. In: C.K. Williams & N. Bookidis (eds), Corinth. The Centenary 1896-1996, Athens, 323-335. SLAPSAK, B. 2007. The Leiden-Ljubljana Ancient Cities of Boeotia Project 2006, Teiresias 37/2, electronic journal. SYRIOPOULOS, K. 1994. J prostorik katokjsiv tjv Elldov kai j gnnesiv tou elljniko qnouv, Aqna. SOTERIADES, G. 1908. Prostorik aggea Xairneiav kai Elateav, ArchEph, 65-95. TOUCHAIS, G., L.C. COURTOIS, E. DIMOU & C. PERLS 1981. Le matriel nolithique. In: cole Franaise d Athnes (ed.), L antre corycien I, BCH Supplment VII, Paris, 95-182. ULRICHS, H.N., 1863. Reisen und Forschungen in Griechenland, Halle. VAN DER ENDEN, M. 2008. Rediscovering the Hellenistic Landscape. The Study of a Cinderella period, M Phil thesis Leiden University. VITELLI, K.D. 1993. Franchthi Neolithic Pottery 1. Classification and Ceramic Phases 1 and 2, Bloomington & Indianapolis. VITELLI, K.D. 1999. Franchthi Neolithic Pottery 2. The Later Neolithic Ceramic Phases 3 to 5, Bloomington - Indianapolis. VITELLI, K.D. 2007. Lerna. A Preclassical Site in the Argolid V. The Neolithic Pottery from Lerna, Princeton. WACE, A.J.B. & M.S. THOMPSON 1912. Prehistoric Thessaly, Cambridge. WEINBERG, S.S. 1962. Excavations at prehistoric Elateia 1959, Hesperia 31, 158-209. ZVORYKIN, I.A. & SAUL, P.J. 1948. Soil Map of Tanagra area, http://eusoils.jrc.ec.europa. eu/ESDB_Archive/EuDASM/images/maps/download/grec_x3.jpg
2009
59
APPENDIX The pie charts on the following pages belong to the article: Koroneia 2009: surface survey sampling experiments (Mark van der Enden) on page 33-41.
60
2009
61
62
2009
63

The Leiden Ljubljana Ancient Cities of Boeotia Project 2009 Seasons

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Leiden Ljubljana Ancient Cities of Boeotia Project 2009 Seasons

Uploaded by

Copyright:

Available Formats

The Leiden-Ljubljana Ancient Cities of Boeotia Project 2009 seasons

JOHN BINTLIFF ET AL.