Topography of the site2

1The village of Prodromos is located 5 km northeast of Karditsa, which, together with Trikala, is one of the two largest modern cities on the western Thessalian plain. The eastern edge of the village is dominated by the high artificial mound (magoula) of Agios Ioannis, on the top of which stands the post-Byzantine church of the same name, as well as the “Konaki” of Ekrem Bei, an Ottoman manor house of the 19^th century AD.

2Magoula Agios Ioannis covers an area of around 0.85 ha and has a maximum height of 12 m at its northern part. It stands at roughly half-distance between two small rivers, Kalentzi to the west and Leipsimos to the east, which are both south-north flowing tributaries of the Pineios river (fig. 1) The north and the northwest sides of the magoula are precipitous, with great hypsometric variation, while the south, east and west sides form a gentle natural slope (fig. 2). The east and southeast edges were bordered in the past by a seasonal torrent, which ran north to south, and by one of its branches, flowing east to west. Both torrent beds today have partially been filled in. The immediate environs of the mound of Agios Ioannis thus constituted an almost level riverside area, which before the improvement works of the 1970s, suffered from frequent flooding. Furthermore, it is known that there were scattered marshy areas around the hill, which consequently significantly reduced the land available for settlement and agriculture in all periods3.

3This circumstance brings us to the supposition (although no modern geological investigation has been conducted) that Magoula Agios Ioannis formed, even in the remote prehistoric past, a natural elevation suitable for the first humans who decided to settle permanently in the area. This could explain, in part, the great height of the mound, in comparison with other neighbouring settlements, as it may be the result of uninterrupted occupation.

4Recently a large part of the south and southwest flanks of the hill have been taken up by building plots, as part of the modern village plan. The erection of a new school building, in one of the plots on the southwest side of the magoula, provided the opportunity to carry out a rescue excavation, which was completed in two seasons on a tight schedule: the first from February to September 2007 and the second from March to May 20084. The excavation took place in a rectangular trench measuring 50 x 20 m, equivalent to the 1000 m² footprint of the new building, with the long axis oriented east-west.

Previous research in the area around Magoula Agios Ioannis

5Prodromos Karditsa is well known in the bibliography for the Neolithic period, from the rescue excavations conducted by G. Hourmouziadis at the beginning of the 1970s at two further prehistoric settlements, located about 1.5‑2 km north-northeast of Magoula Agios Ioannis5 (fig. 1). These excavations revealed building remains and graves6, and provided a great amount of information on building techniques, as well as on the development of pottery production and figurine production during the Early Neolithic period. During the same decade Hourmouziadis also sunk a stratigraphic test pit into the top of the hill, the results of which have still not been published. Only the faunal and archaeobotanical remains of the Early Neolithic have been studied, and published, by P. Halstead and G. Jones7. Their article mentions that the stratigraphic sequence at Magoula Agios Ioannis also included deposits from the Bronze Age and historical period.

6During another rescue excavation conducted in 1991 in a field about 650 m southeast of the magoula, traces of a floor and walls of post-framed houses, quantities of pottery and stone and bone tools were discovered, dating to the Late and Final Neolithic/Chalcolithic periods8.

Stratigraphy – Structural Remains – Absolute Chronology

7The ground in the area excavated in the years 2007‑2008 follows the slope of the hill, with a slight to medium inclination from north-northeast to south-southwest. In the past there were some attempts to level the area, especially in the southwest, by covering parts of the site with great amounts of modern rubble. During the course of the recent demolition of the old building and the removal of the rubble, bulldozers created at this spot a large, deep cut.

8The restrictions imposed by the structural requirements of the new building and the time constraints for the completion of the excavation, obliged us to attain a depth which varied in areas between 2.10 and 2.30 m from the highest point of the trench. Nevertheless, the archaeological deposits continued to a much greater depth, and sterile level without traces of human occupation was reached at no point.

9Taking into account the complementary remains from the different trenches within the excavated area, we can reconstruct the stratigraphic sequence as follows (fig. 3).

101. Below the modern rubble layer, we distinguished (especially in the southern part) a layer of dark brown sandy clay with disturbed deposits from the Late Byzantine, Ottoman and modern periods, the depth of which fluctuated according to the slope. The uniform sequence is interrupted by a large number of elliptical or round pits, of differing depths and diameters, which could be used for the support of large vessels or for storage. In addition, deep cylindrical shafts (wells) were uncovered, similarly scattered throughout the area (fig. 4). Both types of structures were used, as the pottery from the fill indicates, during the Byzantine and Ottoman periods, although some of the wells were probably dug in the Archaic, Classical or Roman periods (see below). These structures had been cut through the lower levels disturbing much of the deposit from the prehistoric period. From the upper part of one of these pits in the southeastern part of the trench comes a charcoal sample submitted for ¹⁴C dating. This sample gave a calibrated age of 1020‑1160 AD (DEM-2036)9, thus confirming the date of use of these scattered multi-function pits.

112. A yellowish sandy layer follows with remains from the Archaic, Classical and Roman periods. In this layer, the depth of which varies again according to the natural slope, the remains of a few structures (stone walls), working areas, and interments were uncovered. In particular, in the middle of the north section of the trench, traces of a clay kiln were revealed, which dates to the Roman/Late Roman period and was probably used for firing bricks10. In the central part of the south section an isolated stone wall was uncovered, which can be roughly dated to the Archaic/Classical period, and also a disturbed extended burial, a simple interment of an adult, dating to the Classical Period.

123. At least three layers have been distinguished with remains of prehistoric occupation. On account of the slope of the modern ground surface, these levels are found at a depth varying from 0.80‑1.20 m in the north-northeast to 1.80‑2 m in the south-southwest. They can be described as follows (ordered from top to bottom)11:

13a) First is a brown-black to chestnut-brown and in places light-coloured sandy clay layer. This was not uniform or distinct everywhere, first, because it did not have the same thickness across the full extent of the excavation, and secondly, because in the north part the upper levels had been removed by the bulldozer (see supra). Thus, the structural remains, pottery, stone and bone tools, along with pits containing destruction debris, were largely confined to the southern part.

14In the southwest part of the excavation, part of a wattle-and-daub house with an apsidal end on the north was revealed at a depth of 2.05 m (fig. 5). It is oriented north-south and its preserved dimensions are 5.80 x 6.90 m. To the south it has been disturbed, probably in modern years, and might actually continue under the south edge of the trench. Inside the narrow, curvilinear foundation (or drainage) trench of the apsidal building, post-holes with traces of wood charcoal were discovered at varying distances apart. A sample taken from one such post-hole gave a date between 2260‑1980 cal BC (GrN-31175)12. At a distance of 3.50 m from the northern extent of the house, part of a dividing wall of mud brick plastered with clay was preserved. South of it was a circular hearth whose foundation was made of fragments of large storage vessels, and further to the south part of a floor was discovered. On it, and also a little to the east of the foundation trench, carbonised seeds were collected (probably almonds and acorns), which have been dated by radiocarbon. The first sample (DEM-2034) gave a date to the years between 2205‑2030 cal BC and the second (DEM-2035) between 2200‑1975 cal BC. There is almost complete correspondence in the dates of all samples, further confirmed by the fact that analysis was made in two different laboratories.

15A little to the east of the apsidal building, a floor made of hand-smoothed clay was uncovered at a depth of 1.85 m. Part of a fallen burnt beam was uncovered along the whole length of its south side, as well as the base of a circular clay hearth to the southwest. Sample DEM-2037, from the burnt beam, which was submitted for ¹⁴C determination, produced a calibrated date of 2290‑2045 BC, i.e. in exactly the same years as the other three. To the north of the floor and at a depth of 1.65 m was an oval clay structure, plastered with many layers of clay over a substructure of sherds, measuring 2 x 1.40 m with a north-south orientation (fig. 6).

16All the pottery from these areas is characterized by sherds of coarse vessels of light brown or brown-black clay, with or without inclusions, covered with a dark slip. They come from large pithoid or basin-like13 vessels, as well as from jars with plastic decoration of parallel wavy lines (fig. 7: a, b). Several grey and black (Minyan) sherds and whole vessels – mainly two-handled cups14 and cups with raised handles or parts of bowls15 – were also found (fig. 8‑10). As far as painted decoration is concerned, only a single matte-painted cut-away-neck jug was found with linear motifs on a burnished red surface (fig. 11)16. Finally there were monochrome cups and small bowls of dark red or brown clay with inclusions. The pottery, therefore (fig. 12), is consistent with the absolute chronology of this level, which should be placed at the end of the Early and the beginning of the Middle Bronze Age17.

17b) A layer of brown-red sandy clay soil mixed with many fragments of daub and heaps of burnt debris was next. It originated from the demolition and burning of various buildings and clay structures, which date to the Late Neolithic period. This level is thicker than those mentioned before (at certain points reaching 0.70‑0.80 m) and can be traced over the entire extent of the excavation, without its depth being consistent, since it follows, as has already been said, the land’s slope. It is certain that it included at least two phases of occupation, as demonstrated by the existence of destruction levels with marked traces of burning within this layer. Further study of the pottery in conjunction with absolute dates could well show the existence of additional phases of construction.

18The most characteristic structural remains, which were discovered in the layer, are as follows.

19In the southeastern part, piles of daub, fragments of floor, post-holes and circular hearths were revealed at a depth which varies between 1.50 and 1.90 m. The total area of these scattered remains is around 7 x 6 m and is oriented northwest-southeast. Sample DEM-2067 comes from a charred post found at a depth of 1.54 m, in a destruction layer with broken pottery and stone tools, beside a circular hearth. This produced the latest of the dates for the Neolithic period, at 4230‑3985 cal BC.

20Approximately in the middle of the northern part of the trench, at a depth of 1.25 m was the largest concentration of daub fragments and clay structures, in an area of approximately 7 x 4.5 m along a northwest-southeast orientation. This must mark the location of one of the houses of the settlement (fig. 13). Charcoal sample DEM-2059 came from the destruction layer, around a large open vessel at a depth of 1.50 m. Its ¹⁴C determination gave a date of 4770‑4550 cal BC. From the area of this house, between the piles of daub and the horseshoe-shaped structures (platforms), two further samples of wood charcoal fragments were collected, one from a post-hole. Their dates agree completely: the first sample (GrN-31174) gave a date of 4720‑4470 cal BC and the second (DEM-2068) a date of 4710‑4545 cal BC.

21To the southwest of the accumulated piles of fallen daub, among the remains of the destruction layer, two graves were found (fig. 13). These were simple internments in shallow pits, probably contemporary with each other. They are extended burials on the same orientation, that is to say northeast-southwest, but with the heads in opposite directions. The graves were unfurnished, but certainly later than the level in which they were found, since they were not found beneath its floor but were clearly cut through it. The date of the bones from one of the graves (Lyon-6469) falls in the interval 1885‑1690 cal BC, confirming the explanation given above, and placing both graves in a late stage of the Middle Bronze Age (MH III).

22Pottery from this layer forms the overwhelming majority of finds from the whole excavation. Undecorated and coarse vessels with thick and medium walls predominate. The most common shapes (fig. 14) are large or medium-sized pithoid vessels18, undecorated bowls of various types and sizes19, bucket-shaped vessels20, or cooking-jars and jugs. Rarer shapes include cut-away-necked jugs with a horizontal tubular spout (“feeding bottles”), whose surface is covered with a red burnished slip (fig. 15), amphorae with vertical handles from the rim to the shoulder21 (fig. 16), tall undecorated fruitstands in pale brown clay with nipple-shaped projections on the handles22, askoid vessels, bowls of different sizes, and jars. The decorated pottery is painted, incised and impressed (pointillé), or uses plastic decoration. Painted decoration is usually black-on-red, ranging from brown to brown-black on a light red, brown-red or orange background (fig. 17). The motifs of impressed (pointillé) decoration are linear as well as curvilinear, spiraliform or stepped23; this kind of decoration is sometimes used on vessels covered with red burnished slip24. Nipples are a common form of plastic decoration. On jars or large open vessels there are horizontal bands of impressed or relief decoration, wavy bands above the handles25, and even a few schematic applied anthropomorphic figures (fig. 18)26.

23The date DEM-2067 (last part of the 5^th millennium BC) dates the upper levels of this stratum, in which a few building remains may be placed, to an advanced stage of the Rachmani phase, something that should be clarified with further systematic study of the pottery. The largest part of the underlying levels, in which most of the material remains were found, can be placed in an early stage of the Final Neolithic period. More evident is the existence of many similarities with the pottery of the lowest, i.e. the earliest, phase at Pefkakia (“Unteres Rachmanistratum”, phase Rachmani I), and furthermore with the pottery of the Rachmani phase at Magoula Otzaki27. It must be noted that not all the categories of pottery of the Classical Dimini phase are present here. On the other hand, none of the shapes or other features could be placed in a more advanced period, such as those found at Petromagoula28, or recently at Mikrothives in Eastern Thessaly29, although pottery of this kind was found in the 1991 excavation to the southeast of Magoula Agios Ioannis Prodromos (supra, p. 383).

24c) Finally, a layer of clayey light brown to grey-green soil with charcoal fragments was identified, especially in the eastern section of the trench, in which the oldest (in relative terms) clay and stone artefacts were found. The remains of this period are few, particularly as excavation did not generally continue to this depth. Only a small section of floor with a few post-holes was found in the northern part, along with pottery and stone tools, and a circular pit. From the destruction level which consisted of burnt daub, broken vessels and scattered traces of burning, a sample of wood charcoal was collected at a depth of 2.15 m. This sample (DEM-2110) proved to be one of the earliest from the site and gave a calibrated date of 5480‑5340 BC. The rare pottery in this level consisted mainly of sherds with brown, dark-brown to brown-black decoration on a light coloured, light brown to brown-yellow surface, as well as a few sherds of grey-on-grey and red-on-cream decoration. The typology of the pottery and the date of this sample both suggest, rightly in our opinion, the dating of this level to the transition from the Middle to Late Neolithic.

25Overall, architectural features consist mostly of burnt remains of destroyed houses, which were built exclusively with clay on a timber frame. The wattle-and-daub technique (torchis) was the principal one used for construction, whereas the use of stone as a building material is completely absent30. The floors were of trodden or smoothed clay, while the roof and walls were made of plaited branches, reeds and/or planks coated with clay, as indicated by the many clay fragments with impressions. Clay fragments with a white coating were recovered in one of the houses. Also characteristic are the various clay structures, both in and outside the houses, which fulfilled the needs of food preparation and storage. These constructions include those for heating, such as ovens, hearths and other compound structures, “platforms”, “cupboards”, benches, and clay-lined pits for the storage (?) of various products31.

26As a result of the many disturbances and the nature of the remains (seen also in the plan, fig. 4), it is not possible to restore the details or the ground plans of the houses, nor their precise orientations, with few exceptions. In consequence, we do not have a clear and complete picture of the internal arrangement of the settlement. However, the pattern in which the daub fell, in combination with the post-holes and the remains of floors, indicates that the houses were quadrangular in shape (with the exception of the apsidal house of the Bronze Age) and had different kinds of additions or open-sided roofed areas. The differentiation between inside and outside space is equally difficult. The houses were only a few metres apart, with different activities being carried out both inside and in-between the inhabited spaces. The accumulation of blades and flakes of flint and obsidian allows the identification of certain areas as used for the manufacture or secondary working of chipped stone tools.

Conclusions

27On the basis of what is presented above, the following statements can be made:

1. There is almost complete correspondence between the different radiocarbon dates obtained for the two prehistoric periods (Neolithic, Bronze Age). The distribution of the samples, both vertically and horizontally, across almost the entire area of the excavation contributes to this conclusion, as does the dating at two different laboratories. This correspondence seems also to confirm the relative chronology suggested by the pottery sequence. It must be emphasized, however, that an up-to-date study of the pottery in this region (Western Thessaly) is completely lacking for both periods.

2. It appears from the various aspects of the material culture (construction techniques, pottery, stone and bone tools) that the site was continuously occupied with the same characteristics from the Middle to the Late Neolithic, which can be traced until almost the end of the 5^th millennium BC. This is suggested both by the pottery and the earliest sample (DEM-2110) dated to the middle of the 6^th millennium BC. It should be mentioned however that more recent evidence from the excavation suggests that the occupation could even go back to the Early Neolithic32.

3. Building and other remains that could be safely dated to the 4^th millennium BC are completely absent. However, at the 1991 excavation, the site of which is located quite close to Magoula Agios Ioannis, both pottery and building remains were found which seem to date to this period (unpublished).

4. After its abandonment at the end of the Late Neolithic/Chalcolithic this site was resettled again at the end of the Early-beginning of the Middle Bronze Age. In addition to the two burials and a few pottery finds from surface collection dating to the end of the Middle Bronze Age, there are also a few scattered sherds from the Mycenaean period, which show continuity into the Late Bronze Age.

5. After the abandonment at the end of the 5^th millennium BC, it appears that in some parts of the site layers accumulated to different depths. This can be explained as a result of the slope of the land, and of the natural erosion which followed the abandonment.

6. As a consequence of the previous point, the remains of the different periods or phases are frequently found at almost the same level or with only slight differences between them. Thus the differentiation of the various periods and phases is very difficult, especially when only a thin layer marks the boundary between periods. But this results also from the fact that the inhabitants of the prehistoric settlement often built exactly on the site of earlier structures, which (especially in the case of clay structures) might be intentionally reused. This can be observed in many locations on the site: the most characteristic example is the existence of an oven abutting the northwest side of the Bronze Age apsidal building, which, according to the date from the charcoal sample DEM-2058, belongs to the years 4770‑4550 cal BC.

7. As far as the gap in occupation of the settlement between the end of the Neolithic period and the end of the Early-beginning of the Middle Bronze Age is concerned, the combination of the observations above (especially 3, 5, 6), and the precision and spread of the absolute dates, leads us to pose the following questions:

   1. Does the thickness of the deposits and the stratigraphic sequence at a site (in the context of a single excavated trench which constitutes only part of a site) provide a secure tool for interpreting the length of abandonment of a settlement at a particular period? The inhabitants of a community could simply have moved a little way away from the place which, according to the stratigraphy, they “abandoned” for some reason. Later, after a longer or shorter period, they would return to a place where conditions were appropriate to meet the basic needs of the inhabitants.

   2. Could the absence of a period or phase at an excavation be simply “accidental”, resulting from our limited knowledge of the extent of any one settlement?

28To conclude this presentation of the first results of the excavation at Magoula Agios Ioannis at Prodromos in Western Thessaly, we may observe that the fortunate circumstance of the existence of a number of secure ¹⁴C dates from one and the same site constitutes in itself a very significant contribution to research and to the study of the Late Neolithic and part of the Bronze Age in the wider region. There are two main reasons for that. First, the excavation at Magoula, even though rescue in nature, with all that entails, constitutes one of the largest recent excavations (in respect of its area) of a major prehistoric settlement of the western Thessalian plain, which, given the present results, was occupied for a very long period. Secondly, although the finds are in several respects not very impressive, their full study in combination with the absolute dates obtained will constitute a “reference point” for these periods in the region. It is to be hoped that this material will provide starting points for young researchers wishing to investigate the different branches of material culture.