Körös culture environment, settlement and subsistence

Preliminary report on the second season of an interdisciplinary project at Ecsegfalva, County Békés, Hungary

Alasdair Whittle, School of History and Archaeology, Cardiff University, PO Box 909, Cardiff CF10 3XU (whittle@cardiff.ac.uk),

Zalai-Gaál István, Institute of Archaeology, MTA, H-1250 Budapest, uri útca 49,

and Sümegi Pál, Department of Geology and Mineralogy, Kossuth Lajos University, Debrecen H-4010


The contexts, aims and objectives of an interdisciplinary project on the environment, settlement and subsistence of the early Neolithic Körös culture (c. 6000-5500 BC) in south-eastern Hungary are briefly rehearsed.

Continued work on the local environmental setting in the Berettyó valley at Ecsegfalva is described, including sediment and pollen analyses from the Kiri-tó meander, of Pleistocene origin but still water in the Holocene. Further geophysical survey on known Körös sites around the Kiri-tó is detailed. Test excavation at Ecsegfalva site 23 on the south side of the meander showed both a well preserved Körös culture occupation level and subsoil features. There were abundant finds of a very wide range, including pottery, daub, netsinkers, a figurine fragment, animal and fish bones, freshwater shells, cereal and other plant remains, and some obsidian, flint and hard stone. A crouched burial was found on the probable edge of the settlement. Test excavation at Ecsegfalva site 18A on the north side of the meander produced less abundant finds, of pottery, daub and animal bones; the pottery was exclusively AVK (Linear Pottery of the Great Plain).

Future work in 2000 and 2001 will concentrate on Ecsegfalva 23, to increase the size of samples with which questions of duration, seasonality and resource range can be approached, and to extend understanding of internal variation and the role of features and structures.

Research context

The unresolved questions about the beginning of the Neolithic period across Europe still abound (Whittle 1996) . How did the phenomenon spread ? What indeed was the phenomenon, and was it the same from region to region ? Who were the principal actors involved and where did they come from ? Were they permanently settled ? What impact did they have on their environments ? What was their perception and experience of their landscapes ? What use did they make of their various subsistence resources ? How quickly were changes introduced and why ?

One of the most interesting regions for approaching these questions lies in Hungary, in the northern part of the Carpathian basin, and thus between the main mass of the Balkans to the south, and the loesslands of central and west Europe to the north-west. On the southern part of the Great Hungarian Plain, along the Tisza and especially in its tributaries to the east, the Early Neolithic Körös culture (c.6000-5500 BC) has been extensively investigated by Hungarian researchers since at least the 1920s, and there was important work in the region by Andrew Sherratt in the early 1980s (Sherratt 1982). Abundant known sites cluster along waterways, with rich remains of cultural material and

food residues, leaving however generally thin stratigraphies and often enigmatic evidence for permanent structures. For various reasons there has been relatively little research on this culture in recent years, though new work by Kalicz and by Bánffy has started on northern Star?evo culture sites on the west side of the Danube between the Drava and Lake Balaton (Kalicz et al.1998). New research by Kertész in the 1990s (e.g. Kertész 1996) has shown a Mesolithic presence on the northern edge of the Plain, but that remains absent in the Plain itself, reinforcing the status of the Körös culture as a distinctively new adaptation. As debated possibilities widen also for the beginnings of the main Linear Pottery culture to the north-west, beyond the Danube, from c. 5500 BC, it is timely to return to the Körös area.

Aims and objectives

The aim of this interdisciplinary project is to broaden and deepen our understanding of the lifestyle of the early Neolithic Körös culture (c.6000-5500 BC) in south-eastern Hungary. Previous research, carried out since at least the 1920s by Hungarian and other archaeologists has shown a dense distribution of sites in certain parts of the river systems of the southern part of the Great Hungarian Plain. These comprise concentrations of pits, some insubstantial structures (perhaps houses), thin occupation levels where these have survived more recent ploughing, and a few graves. Sites are normally rich in pottery and contain a very wide range of residues, from the bones of animals (both wild and tame) and fish to shellfish to charred plant remains including cereals.

The objective of the project has been to locate a microregion where both environmental reconstruction and settlement archaeology could be carried out side by side, in order to examine key aspects and issues of the environment, settlement and subsistence of the Körös culture. Specifically, it has been the objective to investigate in detail the local setting in one Körös culture microregion, including the water regime and the vegetation, to attempt to recover new evidence for the character and duration of settlement, and to examine the range and balance of both domesticated and wild resources. What were these waterside settings like in detail ? What local and wider variation is there ? What was the vegetation like locally and did the Körös culture, with its new practices, have any measurable impact on this ? Did people settle permanently in one place, to achieve short-term or long-term sedentism, or did they move around ? What local and wider variation is there ? Did they use permanent, substantial houses, or do structures represent something flimsier and shorter-lived ? How were new resources such as sheep and goats and cereals integrated into a broad-spectrum economy which included hunting, fishing and shellfish collection ? Does some of the use of wild resources also represent a new adaptation ? Is this an altogether new population, ultimately of southern origin, or is it in part to do with indigenous people from around the fringes of the Carpathian basin (see Whittle 1998) ? What was the social context of change, and were social imperatives as important as economic ones ? There is no shortage of fundamental questions to address.

The first objective of the project has been achieved in the work undertaken in 1998 at Ecsegfalva, Co. Békés (see preliminary report for 1998). Here Holocene sediments in an old meander of the Berettyó river were located by coring and have been shown to contain pollen. Sites around the meander, recorded by the Hungarian topographical survey (Ecsedy et al. 1982), included Körös culture occupations and geophysical survey suggested the presence of abundant subsoil features. The Berettyó runs south-west into the Körös, which in turn runs into the Tisza.

Fig. 1. The north-western part of Co. Békés, with
Körös culture sites in red (after Ecsedy et al. 1982)


The local setting

Coring in the Kiri-tó meander at Ecsegfalva has shown the existence of Holocene and late Pleistocene deposits. The sequence suggests a change from a meandering Pleistocene river channel to a Holocene cut-off channel consituting a partially open oligotrophic lake, with localised backswamp areas. Limonite traces in the sediments suggest variations in water levels. Locally, Körös occupations are on levées along either side of what would have been largely still water, locations either periodically disturbed or surrounded by flooding. Sites 23, 21 and 20, on the south side of the Kiri-tó, were also at or near the junctions with other watercourses. In the case of site 23, there was another old, Pleistocene, channel a little to the east. There were also shallow backswamps in the locality, including immediately behind sites 23 and 18; a shallow cross-channel through site 23 connected such backswamps to the main meander or oxbow, and occupation is to be found on either side of this (see further below). A little to the west ran the Holocene course of the Berettyó, but this appears to have few Körös occupations along it (Ecsedy et al. 1982). To the north, there are the extensive backswamps of the Nagy Sárrét, which may have formed one of the northern limits of the Körös culture, while only a few kilometres to the south are the large alluvial islands and floodplains of the Dévaványa area, where topographical survey and preliminary investigations both suggest the presence of larger Körös occupations than at Ecsegfalva (Ecsedy et al. 1982; Sherratt 1983a; 1983b).

It is clear that this was not a uniform environment, and though a basic observation, this is a significant result in addressing questions of variation in the character and duration of settlement. One of the ways to examine possible flood regimes further will be GIS modelling, to be undertaken by Dr Mark Gillings (Leicester University). The study of historical maps, freshwater shells (Sümegi) and of fish bones (Dr László Bartosiewicz) will also contribute.

Fig. 2. View over Ecsegfalva site 23, with excavations in Trench 23B.
The cross-channel is visible as a light mark in the middle ground. The Kiri-tó is partially visible at top right.

Pollen analysis of the sediments from the Kiri-tó has passed to Dr Kathy Willis (Oxford). Samples are being counted at 16cm intervals initially, to be followed by close resolution counting at 2 cm intervals across the period of Körös occupation. Pollen is reasonably well preserved but due to the presence of large amounts of extremely fine organic material in the sediment is difficult to isolate for counting. New preparation techniques have therefore been implemented which include fine sieving (using 10 micron sieves) of the final residue as an additional stage in the pollen preparation process. Although this is time consuming, preliminary results indicate that the additional technique is vital for the successful extraction of pollen. Levels counted to date indicate that a full postglacial record is present in this sequence and this will provide an important paleoenvironmental reconstruction of the impact of the Körös occupation on the early postglacial forested environment. AMS radiocarbon dates will be obtained on the organic fraction of sediments.

While there is more analysis to be carried out and the profile has yet to be radiocarbon dated, significant differences in vegetation composition are already being detected between this site and others in north-east Hungary. Until recently, only partial pollen sequences had been obtained from the Great Plain, by Járai-Kómlodi (1966), including in the Tisza valley near Csongrad; most other pollen sequences came from north-east Hungary. Adam Gardner (1999) has recovered pollen from a channel deposit at Tarnabod, Co. Heves, on the Plain near its northern edge, and from Sirok on the edge of the Matra hills nearby. While the latter location suggests subtle changes in forest composition compatible with disturbance by coppicing, the environment at Tarnabod was probably rather open. Agricultural activity at Tarnabod from the AVK/Linear Pottery culture onwards produced little or no vegetational impact, presumably because there was little need to clear woodland in order to cultivate levées or other suitable soils (cf.Willis and Bennett 1994). At Ecsegfalva in the Kiri-tó, however, although there is a significant amount of grass pollen in the pollen sequence, there is also the presence of a number of tree types including oak, beech, hazel, hornbeam, lime and birch, suggesting an open-parkland environment rather than grassland. Also of interest is the almost total absence of pine, which is the predominant taxon of many other Hungarian postglacial sequences.

Fig. 3. Water flotation in operation beside Ecsegfalva site 23,
with a view over the southern end of the Kiri-tó.

Ecsegfalva site 23

The probable north-west edge of the site was surveyed by magnetometer in 1998. In 1999 geophysical survey was extended by Dr Michael Hamilton (University of Wales, Newport) along the meander levée to show the presence of magnetic anomalies over a total length of some 100 m and a width of some 50 m. The area is not, however, continuous. There are two smaller and one larger cluster of anomalies. A shallow cross-channel, probably originally connecting local backswamps to the meander, ran transversely through the site, dividing it roughly into two. To the north-west, a recent drainage bank and ditch slightly obscures the distribution of anomalies, though it is looks as though the north-west edge of the site forms a third cluster of features. The site has clearly suffered some damage along its eastern side, since it was the larger local drainage works which led to its discovery in the first place (Ecsedy et al. 1982). The extent and internal variation of the site pose classic questions. Was this a large settlement, occupied in one go ? Or was it a favoured place to which many repeat visits were made over a period of time ? On the basis of other local site sizes, as so far understood, it might be suggested that site 23 stands out by comparison. Dr Hamilton also surveyed sites 21 and 20 to the west along the Kiri-tó, which were considerably smaller. Site 16, on the north side of the meander, is probably also smaller. It is not clear if the south loop of the meander has been systematically surveyed. A little further south, along the edges of the larger alluvial islands towards Dévaványa, Körös occupations appear to stretch along greater lengths (Sherratt 1983a; 1983b).

The magnetometer survey showed dense concentrations of anomalies, which before excavation were taken to be subsoil features. It is clear from test excavation that there were subsoil features on the site, but the anomalies may in fact be produced by the dense concentrations of daub on the site.

Fig.4. Main areas of magnetometer survey at site 23 by Michael Hamilton.
The approximate locations of 23A and 23B are given.


The main sondage was Trench 23B, 10 by 5 m, laid out in the main part of the site as indicated by magnetometer survey. Topsoil was cleared by machine. The ploughsoil was only 25-40 cm deep, and directly below this there was a well preserved occupation level. There must of course have been some plough damage to the site (in the most recent years the main part of the site has been under lucerne and is now being returned to grassland under the management of the Körös-Maros National Park), but the lack of plough damage, at least recently, could be seen by the state of animal bones, sherds, daub concentrations and freshwater shells. Further to the south, plough damage was found to between 30 and 60 cm deep at the Körös site of Endr?d 119 (Makkay 1992). The Ecsegfalva discovery is important.

The occupation level was carefully investigated, with excavation in thin spits, recording of finds to metre squares or individual positions, and extensive dry and wet sieving. The level was rich in finds of many kinds from the outset, including large pottery sherds, daub fragments, netsinkers or perforated fired clay weights, a probable fired clay figurine fragment, animal bones including domesticated and wild species and small mammals, fish bones, freshwater shells, and some obsidian, flint and hard stone. This meant that excavation was slow. The whole of the trench was investigated in the uppermost part of the level, and then in three sub-areas, representing less than half of the whole. The occupation level is at least 15 cm deep. Below it in one sub-area, two small test sondages were made into what proved to be part of one large subsoil feature, presumably a large pit but whose full extent has not yet been established.

Fig. 5. View of the upper part of the main occupation level in Trench 23B of
Ecsegfalva site 23. White tags mark the position of individual finds. Scale: 2 m.

Within the main occupation level, there were abundant finds, but no discernible features apart from one small area of flat, fired clay paving. Charcoal was virtually absent, despite extensive wet sieving. There were many daub fragments, largely scattered across the areas exposed. In two areas, these formed tight concentrations, one of them linear, but neither could be seen as in any way in an original position. The daub fragments frequently contained the impressions of close-set narrow members, probably the stems of reeds. It seems therefore that occupation was accompanied by the presence of structures, but whose character remains to be established. The abundance of fragments and the apparent lack of postholes might be compatible with repeated buildings and destructions or abandonments of flimsy structures. The daub fragments were rarely more than a few centimetres thick, and may largely come from quite low walls without any great load on them.

Fig. 6. Detail of excavation of the main occupation level in Trench 23B,
showing something of the density of finds of pottery, daub and bone.

In some areas of the lower part of the main occupation level and extending also into the probable subsoil pit, there were large concentrations of whiteish material, which was taken during excavation to be also the remains of daub, but less fired, if modified by fire at all. This is to be tested by soil micromorphology (Dr Richard Macphail, UCL). If this proves to be the case, the material seems to represent a further episode or episodes of structure abandonment or destruction. The pit fill had two bands of greyer material which could represent periods of relative stability.

The incompletely investigated stratigraphy of the site therefore already suggests the possibility of a complex cycle of events, involving pit-digging, construction of light structures, their repeated abandonment or deliberate destruction, and renewed occupation that may or may not have involved further pit-digging. The timescale over which this took place is not yet known. Pottery has not yet been analysed, and radiocarbon determinations have yet to be made, though abundant samples are available. One of the key aims of the varied analyses to be carried out will be to examine timescales.

Pottery is to be examined in detail by Elisabetta Starnini (Genoa University). It includes fine and coarse wares, and a range of vessel shapes, principally bowls and jars. There were many sherds of finely burnished bowls; painted decoration was not seen during excavation, but many sherds have yet to have a calcite incrustation removed. Lugs and small handles, and fingernail and fingertip impressions were common. There was some light linear incision. No sharply angled profiles were seen. It is premature to assign the assemblage to any putative phase of Körös culture ceramic development, and one of the aims in analysis will be to look for internal development through the development of the site. It is hoped that in due course sherds will be examined for possible food residues, in the form of lipids and even proteins (Dr Carl Heron, Bradford). The pottery from Trench 23B was exclusively of Körös culture style.

Animal bones are being examined by Dr László Bartosiewicz (Eötvös Loránd University, Budapest). They appear to represent most parts of the body, including jaws, teeth and feet, though no whole skulls were found. A wide range of species (all initial identifications subject to correction) was noted during excavation, including sheep and goats, domesticated cattle, pigs and dogs, and red and roe deer, horse, ass, aurochs, and many small mammals and birds. A wide range of ages is probably represented, including subadult specimens. There is enormous scope here for analysis. Animal diversity will contribute to understanding the mosaic environment. The balance between domesticated and tame will give clues to the main patterns of subsistence. Age and sex data, taken together with studies of teeth wear and growth, will contribute to understanding exploitation strategies, as well as issues of seasonality. After the three planned seasons of excavation, the animal bone assemblage is likely to match the size of that from Endr?d 119 (Bökönyi 1992).


Trench 23A, 5 by 5 m, was laid out on the probable north-west edge of the site, as indicated by 1998 geophysical survey. The modern surface is grassland, established after the abandonment of cultivation following artifical build-up of the surface in nineteenth-century and perhaps later drainage operations. The Körös culture level was here at about 90 cm below the modern surface, reflecting also in part the original landform and slope of the levée as well as more recent landuse history. The trench was initially dug by hand and then by machine, and an area of over 2.5 by 2.5 m was examined by hand through the Körös level.

Fig. 7. View of an early stage of the excavation of Trench 23A, with the Kiri-tó behind.

Körös culture sherds, with only one or two of later styles, and some animal bones, were found throughout the profile, right up to the surface. Some may have been brought in by drainage operations and surface alterations, while others may have been taken up through the profile by natural soil processes. The lower part of the profile seems to represent a soil horizon, modified by later processes, with a scatter of Körös culture sherds and animal bones. There was virtually no daub, and the density and abundance of finds in general were much less than in the main occupation level described in Trench 23A. This is compatible with a position on the edge of an occupation.

The further distinguishing feature of the trench was the discovery of a tightly crouched adult human skeleton, probably female, above what appears to be a pit, which contained cereal and other plant remains. The skeleton (to be examined by Dr Pap Ildikó, Natural History Museum, Budapest) was very tightly crouched, on its left side, with arms also mostly under the left side. There was no sign of a grave pit as such, and the body had presumably been lain on the soil or in a depression in the soil. This may contribute to understanding the continuity of occupation of the site. The teeth of the skeleton were well ground, and had bad abscesses, as well as a pronounced narrow groove across the front upper incisors. There are possibilities here also for isotopic analysis of diet and DNA analysis of genetic descent (Oxford Research Laboratory, and Prof. Bryan Sykes, Institute of Molecular Medicine, Oxford).

Fig. 8. The human burial in Ecsegfalva 23A.

Cereal and other plant remains were recovered by water flotation, carried out by Amy Bogaard and Jo Denning on behalf of Dr Glynis Jones (Sheffield University). Trench 23B produced very few remains, though there were several sherds with plant impressions including cereal spikelets. The top of the pit below the burial produced several hundred remains, including emmer, einkorn, barley and weeds. This assemblage should further extend our understanding of subsistence diversity and perhaps also of seasonality of activities, and may provide opportunities for further DNA investigations (Dr Terry Brown, Manchester).

Ecsegfalva site 18A

A series of discrete clusters of magnetic anomalies were found during 1998 geophysical survey along the north side of the Kiri-tó, within the general area of Ecsegfalva site 18 (Ecsedy et al. 1982). One of these, 18A, on the ridge immediately overlooking the Kiri-tó, was selected for test excavation in 1999, and a trench 5 by 5 m was investigated by hand. Initial excavation produced sherds, animal bone, daub fragments and a little obsidian and flint, but the sherds were exclusively of the AVK culture, with typical curvilinear decoration. Subsequent investigation was confined to a smaller area, 2.5 by 2.5 m. There was one marked concentration of daub fragments, very similar to those from Ecsegfalva 23, and below it a possible remant of a dark occupation level; there was nothing comparable to the occupation level seen in Ecsegfalva 23B. Cultivation has here presumably been more intense (at present for lucerne, maize and cereals). There were no subsoil features, though of course the area opened was very small indeed.

The AVK culture is beyond the principal remit of the project, but it is of considerable interest nonetheless. Radiocarbon determinations from site 18A would help to establish the general duration locally of the Körös culture. AVK sites appear locally often to be smaller and more dispersed than Körös ones (Sherratt 1983a), and often discontinuous with Körös occupations. Their distribution serves to accentuate the distinctive nature of Körös sites, including Ecsegfalva 23.

Future strategy

Future objectives are to complete the programme of coring and pollen analysis, and to complement these with GIS modelling of possible water regimes. Excavation in 2000 will concentrate on Ecsegfalva 23, finishing and extending Trench 23B, and opening another 10 by 5 m sondage beyond the cross-channel. It is hoped to open a larger area in 2001, but the feasibility of this will depend in part on the further experience of 2000. In this way, the number of samples can be increased with which to approach questions of duration and seasonality of occupation and of the range of subsistence resources used, as well as opening possibilities for understanding the range of variation in structures, features and activities within the site. A full programme of post-excavation analyses will be carried out, including extensive radiocarbon dating.


The 1999 fieldwork was supported by the British Academy, The Society of Antiquaries and Cardiff University. The help of the Ecsegfalva authorities including especially the polgármeister and of the Körös-Maros National Park is gratefully acknowledged.

The research is a cooperation between Cardiff University, The Institute of Archaeology, MTA, Budapest, and the County Békés Museum, Békéscsaba, with permission granted from 1998-2001. Grateful thanks are due to Professor Csanád Bálint and Dr Igor Grin for help with these arrangements, and to Dr Imre Szatmári for endless practical help.

Thanks are due also to the colleagues in Britain, Hungary and elsewhere who will provide specialist studies. Obsidian and flint will be studied by Inna Mateiciucová (Brno University). We are grateful to Dr Paul Pettitt, Oxford, for advice on AMS C14 dating.

Dr Rick Schulting provided invaluable help in running and recording the excavations, and Cardiff and Leipzig University students did the hard work of digging.


Bökönyi, S. 1992. The Early Neolithic vertebrate fauna of Endrod 119. In S. Bökönyi (ed.), Cultural and landscape changes in south-east Hungary, 195-299. Budapest: Archaeolingua.

Ecsedy, I., Kovács, L., Maráz, B. and Torma, I. 1982. Magyarország Régészeti Topográfiája: a Szeghalmi járás IV/1. Budapest: Akadémiai Kiadó.

Gardner, A. 1999. The impact of Neolithic agriculture on the environments of south-east Europe. Unpublished PhD thesis, Cambridge University.

Járai-Kómlodi, M. 1996. The Late Glacial and Holocene flora of the Great Hungarian Plain. Annales Universitatis Scientiarum Budapestensis de Lórando Eötvös Nominatae 9-10, 199-225.

Kalicz, N., Virág, Z.M. and Biró, K.T. 1998. The northern periphery of the Early Neolithic Starcevo culture in south-western Hungary: a case study of an excavation at Lake Balaton. Documenta Praehistorica 25, 151-88.

Kertész, R. 1996. The Mesolithic in the Great Hungarian Plain. In L. Tálas (ed.), At
the fringes of three worlds: hunter-gatherers and farmers in the middle Tisza valley, 5-34. Szolnok: Damjanich Museum.

Makkay, J. 1992. Excavations at the Körös culture settlement of Endrod-Öregszolok 119 in 1986-1989. In S. Bökönyi (ed.), Cultural and landscape changes in south-east Hungary, 121-93. Budapest: Archaeolingua.

Sherratt, A.G. 1982. The development of Neolithic and Copper Age settlement in the Great Hungarian Plain. Part 1: the regional setting. Oxford Journal of Archaeology 1, 287-316.

Sherratt, A.G. 1983a. The development of Neolithic and Copper Age settlement in the Great Hungarian Plain. Part II: site survey and settlement dynamics. Oxford Journal of Archaeology 2, 13-41

Sherratt, A.G. 1983b. Early agrarian settlement in the Körös region of the Great Hungarian Plain. Acta Archaeologica Academiae Scientiarum Hungaricae 35, 155-69.

Whittle, A, 1996. Europe in the Neolithic: the creation of new worlds. Cambridge: Cambridge University Press.

Whittle, A. 1998. Fish, faces and fingers: presences and symbolic identities in the Mesolithic-Neolithic transition in the Carpathian basin. Documenta Praehistorica 25, 133-50.

Willis, K.J. and Bennett, K.D. 1994. The Neolithic transition fact or fiction ? Palaeoecological evidence from the Balkans. The Holocene 4, 326-30.

Return to Research Reports