Group identities in the Central Balkan Late Neolithic

Investigations of the Central Balkans Late Neolithic1 have been continuing for over a century. Although in the last few decades and particularly in recent years, more and more texts which discuss the social organisation of Vin≠a culture in the territory of Serbia have been published, it must be borne in mind that field investigations in the same period did not progress much further beyond investigations of individual houses. It is important to mention that in Serbia, no Late Neolithic settlements have been investigated completely, or at least to a considerable extent. The eponymous Belo Brdo site at Vin≠a near Belgrade has been intermittently investigated for over a century. The importance of this site, with vertical stratigraphy encompassing almost the entire life of the Vin≠a culture, is immense, although neither the area of the settlement, nor its shape, have been confirmed in the literature.2 Thus we are able to follow the chronological sequence of the entire Neolithic ‘culture’ in one location, but in fact we can not be sure how big the settlement was. Even when frequently cited Late Neolithic settlements in Serbia such as Gomolava and Divostin are concerned, we have the problem of the size of the entire settlement. At Gomolava, around 35% of the surviving settlement area has been investigated (Brukner 1988.20) but we can not even assume its original size, because the Sava River has been eroding the western section of the Neolithic settlement for millennia. The assumed settlement area at Divostin is ABSTRACT – The final period of Neolithic Vin≠a culture, which occupied wide areas in the Balkans, is characterised by large settlements, which were built, judging by the most recent investigations, according to premeditated plan. What was their purpose? Were they autonomous or part of some wider communities? How large was the territory within which people of that time defined themselves as ‘we’ and where did communities of ‘others’ begin? The objective of this work is to indicate the possibilities for studying the complexity of group identities in the Late Vin≠a societies. We take as a starting point the micro-region of Drenski Vis in north-western Serbia, where five Late Vin≠a settlements have been discovered.


Introduction
Investigations of the Central Balkans Late Neolithic 1 have been continuing for over a century.Although in the last few decades and particularly in recent years, more and more texts which discuss the social organisation of Vin≠a culture in the territory of Serbia have been published, it must be borne in mind that field investigations in the same period did not progress much further beyond investigations of individual houses.It is important to mention that in Serbia, no Late Neolithic settlements have been investigated completely, or at least to a considerable extent.The eponymous Belo Brdo site at Vin≠a near Belgrade has been intermittently investigated for over a century.The importance of this site, with vertical stratigraphy encompassing almost the entire life of the Vin≠a culture, is immense, although neither the area of the settlement, nor its shape, have been confirmed in the literature. 2 Thus we are able to follow the chronological sequence of the entire Neolithic 'culture' in one location, but in fact we can not be sure how big the settlement was.Even when frequently cited Late Neolithic settlements in Serbia such as Gomolava and Divostin are concerned, we have the problem of the size of the entire settlement.At Gomolava, around 35% of the surviving settlement area has been investigated (Brukner 1988.20)but we can not even assume its original size, because the Sava River has been eroding the western section of the Neolithic settlement for millennia.The assumed settlement area at Divostin is ABSTRACT -The final period of Neolithic Vin≠a culture, which occupied wide areas in the Balkans, is characterised by large settlements, which were built, judging by the most recent investigations, according to premeditated plan.What was their purpose?Were they autonomous or part of some wider communities?How large was the territory within which people of that time defined themselves as 'we' and where did communities of 'others' begin?The objective of this work is to indicate the possibilities for studying the complexity of group identities in the Late Vin≠a societies.We take as a starting point the micro-region of Drenski Vis in north-western Serbia, where five Late Vin≠a settlements have been discovered.
KEY WORDS -Vin≠a culture; Late Neolithic; settlements; group identity; Balkan 1 In this work, I discuss Late Vin≠a settlements in the territory of Serbia. 2 One of the reasons for this is modern construction works, which covered a large area of the Neolithic settlement.
DOI> 10.4312\dp.39.11 around 15ha according to the distribution of surface finds; however, only 1.17% of the entire settlement has been investigated (McPherron, Srejovi≤ 1988). 3  On the other hand, there is still another reason why Vin≠a and Late Vin≠a settlements have not been fully investigated.The limiting factor was often lack of funding, so the investigations were of a rescue and sondage character, and therefore relatively small areas have been excavated.Still, lack of funds is no excuse for the many decades when a general strategy for investigating Vin≠a culture was lacking.First of all, I have in mind the lack of organized large and well-planned site surveys (surface prospecting), aerial prospection and the use of non-destructive investigation methods.The data on the distribution, size and chronology of the Vin≠a sites in Serbia are very meagre and provisional.
Thus we are in a situation to study social relations within settlements on the basis of the contents of investigated houses, which to large extent are actually random samples, since we do not have the following data: G in which section of the settlement were investigated houses located4 , G the actual size of the settlement, G whether the settlement was surrounded by trenches and/or palisade, G how the size of a given settlement corresponds with the size of neighbouring contemporary settlements, G what level in size hierarchy within one settlement cluster has the settlement with investigated houses.
The method of geomagnetic mapping was used in prospection of Late Neolithic settlements in Serbia over four decades ago for the first time.This made it possible to grasp the distribution of structures over a larger area within one settlement (McPherron, Ralph 1970;Mu∫ijevi≤, Ralph 1988;Tringham, Brukner and Voytek 1985.427-428),but this activity was resumed only at the beginning of the 21 st century (Crnobrnja, Simi≤ and Jankovi≤ 2010;Ar-si≤, Mileti≤ and Mileti≤ 2011).Also, I must mention that a few authors have paid special attention to the distribution of the Vin≠a settlements within certain areas (Risti≤-Opa≠i≤ 2005;Chapman 1981;1990).Nevertheless, these works are also based on insufficient data that was collected in earlier decades without being re-checked in the field and at least partially revised.
I will try in this paper to point briefly to some guidelines which in future investigations could help in seeking answers to questions concerning the organisation of life in the final period of the Vin≠a culture and could make the recognition of the complexity of group identities in Late Vin≠a societies possible.5 I have found good guidelines in articles dealing with similar questions about the same period and in the immediate vicinity, on the Great Hungarian Plain (Parkinson 2002;2006) and eastern Balkans (Chapman 1989), and the work of Constantinos Doxiadis (Doxiadis 1968;Doksijadis 1982) had also influenced my reasoning.6 The most frequent category according to which it is attempted to grasp the Late Neolithic social organization is the individual house, i.e. the household, its internal organisation and the artefacts discovered within.In the organisation of interior and establishing the contents of one house many factors take place (Kuijt 2002.140),including standards imposed by customs and religion,7 already established and generally accepted organisational rules needed to satisfy economic needs and functionality of space of the inhabitants of that distinct house, and finally, personal taste, the affinities and needs of the house owner.Thus, the number of vessels discovered inside the house would actually reflect first of all the needs and possibilities of the occupants of that very house, but it will tell us little about the interaction of its occupants with the occupants of other houses in the settlement.
On the other hand, when the site at Crkvine-Stubline is concerned, a settlement plan is available, which makes it possible to understand relations within the community to a much greater extent.
A settlement lasts longer than any single house in it, and is permanently enlarged and reintegrated in time.Therefore, the 'imprints' of social processes taking place at a given time in the community creating the settlement could be comprehended only by an understanding of the matrix of the settlement and houses within it.The relations between individual houses and settlement must be observed as reciprocal results.At first, personal affinity in the selection of house location can be observed, but this was the main consideration in establishing of the settlement.On the contrary, providing for community needs was more important even at the very founding of a settlement than satisfying individual needs, so even the initial distribution of structures within a given space must be sufficiently functional for all members of the community.
Increase in house numbers in the settlement resulted, at least where Crkvine-Stubline is concerned, in greater uniformity in their arrangement; the reason for this was adaptation to the needs of community (from economic, social, security and, conditionally, political aspects) (Doksijadis 1982.48-50).
Taking into account the facts stated above, the influence of individual or family needs and affinities could be considered as of secondary importance in the creation of the settlement matrix in comparison with the dominant influence of social organisation of the community and relations within the community.The arrangement of the main functions, houses and communications in the settlement depends on the level and mode of organisation, as well as the needs of the community.
Therefore, I believe that with considerable reliability we could take up the examination and study of settlement structure as some kind of material imprint of social organisation of a community which built and inhabited a given settlement.
The settlement as the most discernible highest unity of hierarchy should be the starting point, and we should then continue by planning two subsequent directions of investigation: firstly, to tackle the lower organisational levels within the settlement (house, household, group of houses), and secondly, and much more difficult to understand, the possible higher organisational levels (groups of settlements, micro-regional and regional connections of the settlement) (Crnobrnja 2011.142).
If we want to consider the possibility that multi-layered group identities existed, we have to establish how many organisational levels existed in the given society within which people recognised themselves as parts of a wider community, thus sacrificing some individual rights for the benefit of being a member of a community or communities.In this case, we must first engage in a quest for relations existing above the level of the single house or household, i.e. the material imprints resulting from activities at those levels.
As a small case study, I will consider the micro-region which Belgrade City Museum has been investigating intensively in recent years -the Drenski Vis plateau.This is a relatively small deluvial-proluvial plain, with a few Late Vin≠a settlements, including the settlement at Crkvine-Stubline which has been intensively investigated using geophysical methods and, to a smaller extent, by archaeological excavations.

Location
Crkvine is situated near the village of Stubline, c. 40km southwest of Belgrade.It is located in terrain categorised as second river terrace on the very edge of a deluvial-proluvial plain (Drenski Vis 115-120m above sea level) extending in the background (Fig. 1).The settlement is situated on a hill whose axis is oriented in northwest-southeast.From north and east, the hill is surrounded by streams which meet under its southeast end and flow toward the Tamnava River.There are also many springs at the site and immediately next to it.Despite the fact that the settlement is situated at a relatively low elevation (112m above sea level), its position makes exceptional visual communication possible with areas to the south and east, but not with settlements in its background in the area of Drenski Vis.

History of investigations
The general size of the Crkvine-Stubline settlement was first determined by an intensive surface survey.Most of the pottery material gathered on that occasion dates from the Late Vin≠a period in its final phase (Vin≠a D), while quite a few fragments date from later periods (Late Eneolithic).An area of 85 000m 2 , c. 70% of the entire settlement area, was then surveyed by geomagnetic mapping (Fig. 2).
On the basis of geomagnetic mapping, positions for placing geoelectric profiles were determined and were measured over a total length of 1250m.It was concluded that the remains of all houses first located by geomagnetic mapping and then confirmed by geoelectric scanning are situated in the physically same horizon. 8The results of the geomagnetic mapping proved to be very reliable, and have been confirmed by excavations in six places, but only relatively shallow structures were precisely registered.Two hitherto investigated houses yielded pottery material from the Vin≠a D-2 phase (dated at 4650/ 4600 BC, cf.Bori≤ 2009) which corresponds to the material obtained from the surface survey, i.e. to material from houses damaged by cultivation.By combining the results obtained by above mentioned methods, we can assume with great probability that all the structures recorded by geomagnetic mapping date from the final horizon of occupation.

Settlement data
On the basis of above-mentioned investigations, I tried to determine the approximate positions of houses within the settlement and their dimensions, taking into account the following principles: I considered as assumed house locations geomagnetic anomalies over 6m long and over 10 nT intensity.Using this method, I reached the following conclusions about the Late Vin≠a settlement at Crkvine-Stubline just before life at the site ended: G The settlement covers a relatively large area (c.12.5ha including ditches i.e. around 10ha within the trenches).G The settlement was surrounded by ditches9 .G From all appearances, the houses detected by geomagnetic mapping date from the last horizon of occupation and were all destroyed by fire.G In the area covered by geomagnetic mapping, it could be assumed that 219 anomalies represent the remains of houses (Fig. 3).G The average surface of the anomalies assumed to be houses is 58.3m 2 .G Houses are almost identically oriented (NE-SW) except for few structures.G Houses in the settlement were built in tightlypacked rows.G The distances between houses in rows are small, c. 2m on average (varying from 1 to 3.5m).G At a few locations, houses are clustered around rather large open areas resembling small squares (500-1200m 2 ).G Houses arranged in rows and blocks around open areas is the basic module of settlement texture.

The relationship between areas with structures and open areas
If we take into account only the area within ditches the ratio between open areas and house occupied areas is 6.8 : 1.If we consider only the area occupied by houses, disregarding the area between the last houses and trenches (7.3ha), the ratio is 5.7 : 1.Nevertheless, this method of calculating occupied areas should be taken with some reservation.Firstly, it should be borne in mind that most of these calculations are based on the picture obtained from a small area investigated within the settlement that is then extrapolated to the entire settlement (Chapman 1989.35;Por≠i≤ 2010.203).The mistakes might result from such a method can be seen in Fig. 4. When the construction index within the settlement is examined, a few other things must be taken in consideration.First, the houses are separated by small spaces (1-3m, on the average 2m), so we can also assume that the spaces between them were occupied, as it was hardly sufficient for storing goods and passing between the houses.Secondly we do not know how many houses in the settlement had an upper storey or attic. 1 10 0 Even if the upper storeys of the houses were used only for storage, they contribute to relieving the house floor area.Thus plot ratios (area occupied by structures in relation to the settlement area) are no longer the only issue here; the total floor area ratio (floor area of ground floors and storeys of houses/structures in relation to settlement area) as well as the total occupation of space (besides the houses within the settlement, there were also passages, rubbish pits and heaps and ma-terial stored outside the houses) also need to be taken into account.This information should certainly be taken into account in cases when house floor areas are used in calculations (cf.Por≠i≤ 2011.323).When all this is taken into account, it is not difficult to understand that most of the space outside the houses was also occupied.Therefore, little space remains for to confirm John Chapman's assumption of a 'house and garden' concept, according to which the area around houses could have been used for gardening, and this was based on the ratio between the areas with structures and open areas within the settlement (Chapman 1989.38). 1 11 1

Spatial organisation and the organisation of life in the settlement
The space occupied by one settlement includes not only the area where houses were built, but also the area surrounding the settlement that its inhabitants used for farming and obtaining basic resources (Chapman 1989.34;Roberts 1996.24-25, 29, Fig. 2.5 and 2.6), so in considering the organisation of life, all of this space should be understood as a single entity.
When organizational units within the settlement at Crkvine in Stubline are concerned, we can distinguish four levels: individual houses (assumed 219 structures; Fig. 3), which are arranged in rows of houses (16 rows with 5-12 houses each could be identified; Fig. 5) and sections of two or more rows could make groups of houses clustered around open areas (10 such areas from 500 to 1200m 2 each could be assumed; Fig. 6); houses located on the settlement periphery and not arranged in rows or groups, but distinct from the mentioned organisational units.
Housing density, the identical orientation of the houses and small distances between them, the arran- gement of houses in rows12 and open spaces between houses at some locations in the settlement indicate almost planned building activity.
But we must also bear in mind that not houses but people lived in the settlement, so we must not disregard their individual and personal needs like interaudibility and inter-visibility in the settlement (Chapman 1989.35) or their non-material values, ideals and emotions (Whittle 2003.16).These categories could hardly be recognised by studying artefacts and individual structures, but understanding the area where these people lived offers a better possibility to envisage the above mentioned categories.
How could the people living in a settlement like the one at Crkvine-Stubline have felt?The spaces around the houses were so small that the occupants actually could not have spent a single moment outside the house without meeting their neighbours and they could probably hear their voices all the time they were awake.The activities which could have been carried out around the houses were also limited because of the small distance between them, so the open spaces between the houses, small 'squares', of between 500 to 1200m 2 , remained the only suitable areas (Fig. 6).The views within the settlement were also very limited and the view of the occupants of most houses was limited only to the 'square' in front and houses across.All activity outside the houses was carried out in a public, communal area.A similar tendency can be noticed at the level of the entire settlement.The grouping of houses in clusters around open spaces, so-called 'squares', could suggest a division into kinship groups within the settlement.Even if this assumption is correct13 , how strict was the division between such groups either of kinship or other character, and does this mean that the settlement was divided into distinct sub-entities occupied by certain clans or extended families?I sought an answer to this question in the relations between house groups and the assumed disposition of primary and auxiliary communications within the settlement (Fig. 7).I started from the assumption that each house, i.e. every group of houses, must have had free and rational access to the settlement entrance and exit.As all farming land was outside the settlement and springs were located near the streams along its fringes, these were routes which most of the inhabitants had to use many times every day.In order to cover these routes undisturbed on a daily basis the occupants of each house group must have passed through the communal areas of some other groups a few times a day.This suggests that no single group could have laid absolute claim even to 'their' communal area as they had to share it with people from other sections of the settlement.All this leads to the conclusion that all open spaces in the settlement, the so-called 'squares', were more or less public and communal, and hence belonged to the entire settlement.Such a situation is possible in settlements within which there is strong sense of belonging to the community at the settlement level, the feeling of importance of the community at this level, which requires the sacrifice of a portion of individual and family identities for the survival and prosperity of wider community.Such an organisation also assumes a certain degree of hierarchy, the existence of which in the Late Neolithic is usually open to question.

Fig. 4. Three possible examples of house density at Crkvine-Stubline on the basis of excavations without geophysical investigations (20 x 50m).
I will not enter into wider examination of the problem of the existence or non-existence of hierarchy in social organisations in Late Neolithic communities in the central Balkans here, but I would like to draw attention to a certain find from Crkvine-Stubline that indirectly suggest the high probability that vertical and horizontal social stratification existed at that time.The unique arrangement of a composition consisting of 43 figurines discovered in a house vestigated in 2008 indicates that in the mental maps of people of the time there was a sense of complex organisation, which must have existed also in real life in order for it to be materialised in this composition of figurines (Crnobrnja 2011).
However, many activities took place outside the settlement: farming, grazing, obtaining essential resources and additional provisions available in the immediate vicinity (wood, water, possibly hunting and fishing).
Cultivable land had a distinct value in farming communities and can not be considered as merely another area of land belonging to the settlement.Cul-tivable soil is the element which provides crops; special cults and rituals are related to it in all agricultural communities; it has mystical characteristics and is in some way a sacred object.
Therefore, it is very important to direct our considerations to how the inhabitants of large, densely populated settlements like the one at Crkvine-Stubline carried out internal redistribution of cultivable land which was obviously outside the settlement perimeter, and whether they divided the land at all or farmed it communally, dividing only the crops.
If the cultivable land belonged to individual households, there must have been many hundreds of parcels near the settlement.Such a mode of ownership would have been very irrational considering the technology of the time.As a result of an uneven redistribution of land between the families (considering quality and distance of the parcel) conspicuous social stratification would very soon have occurred.
Another possibility is that land was cultivated by the combined labour of a few families of the same lineage or from some organisational entity within the settlement (e.g., occupants of house groups around 'squares').This system of land exploitation and join forces should not be surprising, especially given that even building an average house in the settlement required the joint effort not only of one nuclear family, but at least dozen men in their prime (Snashall 2002.8;Hofmann et al. 2009.40).
On the other hand, such a pattern in the economy and association also assumes the existence of some kind of 'symbolic capital' acquired both by individu- als as well as the group to which he/she belongs. 14ooperative cultivation and division of crops should be understood not only as a joint physical effort resulting only in the division of material gain.This activity also entails joint rituals, the exchange of symbolic capital leading to the reinforcement of the identity of the group/subgroup which carries out work/ activities together.
Similar behaviour could be also assumed for the settlement as a whole.Making the ditches which surrounded the settlement required participation of a very large number of people, if not the entire population that was fit to work for a relatively long period.This volume of work assumes the existence of one or two mechanisms: forced labour or voluntary effort.If we assume some kind of forced labour, this would mean the existence of a distinct group capable of forcing the entire settlement community to work.If we assume voluntary activity, we must assume as well that there was strong sense of identification with the community at the settlement level.Only a strong group identity could have made it possible for people to accept the need to sacrifice and contribute the time and labour of the individual, the family and the group, as well as to invest in certain material goods15 for the well-being and progress of the entire community.It should be borne in mind that the ditches surrounding and bordering the settlement as well as the act of their construction could have had not only a practical (defensive) but also a symbolic dimension, as some authors have suggested (Raczky, Anders 2008.37;Chapman, Gaydarska and Hardy 2006.20).
Either of the two mentioned mechanisms which made investing such a large effort for the communal project of ditch construction possible means that certain at least an initial type of hierarchy was present (Raczky, Anders 2008.38).

Settlement cluster
Single permanent settlement regarding its size never makes one self-sufficient entity but belongs to a wider community.Large settlements distinguished by their size are not created only to satisfy the needs of their inhabitants.Large settlements in all periods are also established to satisfy the needs of many small settlements gravitating toward them and simultaneous complex interaction (communication, exchange and cooperation) with lower-ranking settlements (Doksijadis 1982.73-75).
The situation we encountered in the immediate vicinity of Crkvine in Stubline is very interesting.Four smaller Late Vin≠a settlements (covering 1-3ha) lie within a 10km radius, their last occupation horizon being contemporary with the last horizon of occupation at Crkvine-Stubline (Fig. 8).The data on the pedological characteristics of the soil indicate three clearly distinguished possible economic zones: land suitable for meadows and woodland at Drenski Vis plateau (covering 120km 2 ), fertile arable land along the waterways fertilised by flooding every spring (90km 2 ), watercourses and marshes abounding in fish in the immediate vicinity.
In the Late Neolithic of the Great Hungarian Plain, discrete settlement clusters of 2 to 7 settlements in which most clusters had one settlement of tell type or one very large horizontal 'super-site' have been recorded by John Chapman (Chapman, Gaydarska and Hardy 2006.29),Wiliam A. Parkinson (2002. 410;2006.43, 53), Pál Raczky (Raczky, Anders 2008.38).
At the present stage of investigation of sites near Crkvine-Stubline we could not make reliable assumptions about the inter-relations of all these settlements.It might be concluded with certainty that the settlement at Crkvine-Stubline is the largest of them and that settlements are at small distances from each other.The assumed maximum population of somewhat over 2000 inhabitants in the settlement at Crkvine-Stubline (Por≠i≤ 2010.342) and four more smaller settlements at relatively small distances16 indicate that a large number of inhabitants was concentrated within a relatively small area in comparison with the assumed population density in the Balkans in the Late Neolithic (Müller 2007;Chapman 1981.48) and in the Great Hungarian Plain (Kalicz 2001.157).Given the locations of contemporary settlements at Drenski Vis and distribution of cultivable land the question arises as to how the division of resources between these settlements was effected.
The co-existence of settlements within a relatively small area assumes some kind of agreement between their inhabitants concerning the occupation of farming land, pastures, woodland and hunting grounds.Chapman (1989. 39) emphasises that: "No matter how large the villages, no matter how complex the exchange and alliance networks integrating these communities, no matter how wide the disparities in communal or individual prestige goods between them, there is no evidence for inter-polity domination at village level in the Balkan Neolithic and Copper Age.Instead, there is evidence for close structural and functional links between parent communities and dispersed farmsteads and between settlements of similar sizes." If we are to accept this view as probable, the question remains as to how the settlement inhabitants themselves in each such settlement cluster perceived certain types of agreement on co-existence.Was such a relationship sufficient to develop among these people into some sense of belonging to the community created by a few settlements?How were the inhabitants of one such group of settlements perceived by other neighbouring groups of settlements?

Conclusion
I attempted in this work to indicate the possibility of recognising complexities in group identities through an understanding of the basic organisational levels in Late Vin≠a communities within one micro-region.I will briefly state some preliminary conclusions which perhaps should rather be understood as working hypotheses for future investigations.The levels of social organisation which I identified and which could correspond with certain group identities are as follows: Single house -belonging to a nuclear family.
Groups of houses surrounding open spaces ('square') -at this level group identity related to lineage or corporate group membership could be assumed.
Settlement -the organisational settlement pattern, assumed communication lines within settle-ment, division of resources and confirmed communal activities demanding the participation of the entire community bear witness to the existence of a strong group identity at the single settlement level.
Group of settlements -existence of some kind of agreement between settlement inhabitants is necessary for the division of territory and resources; it is not clear whether and to what extent group identity was developed at this level.
It is my view that the sense of belonging to the community, i.e. group identity, was strongly developed at the level of house groups within settlements and at the settlement level.On the other hand, whether relations existing within settlement clusters could have created a sense of belonging to the wider community is not known.Was the border within which a sense of collective identity and an exterior of 'others' at the level of single settlements or groups of settlements?What was the potential of large Late Neolithic settlements to establish domination over other settlements or cooperate with them?Can such phenomena in the material culture or landscape be identified, and if so, how?Should we continue to study Vin≠a culture in order to identify the varieties of material culture in given regions or should we first define the general territorial and social spheres of influence of Vin≠a culture?

Fig. 1 .
Fig. 1.Geographical position of the Drenski Vis region and Late Vin≠a sites at Drenski Vis and in the immediate vicinity which have been investigated.