Thanks to 64 years of sustained and intensive work, a small area in the western Konya Plain in the south-central Anatolian Plateau of Turkey now provides uniquely rich evidence for the study of the Neolithic of south-west Asia and the earliest transition from groups of mobile foragers to settled agricultural societies. Much of the work has been sponsored by the British Institute at Ankara (BIAA). This research was initiated by the BIAA’s survey in the area by Mellaart, French, and Hall, which identified the now-famous World Heritage Site of Çatalhöyük as Neolithic, and which Mellaart went on to excavate in the early 1960s. BIAA-sponsored projects have continued to identify, survey, and excavate new Epipalaeolithic (that is, the period between the Upper Palaeolithic and Neolithic) and Neolithic sites over the decades since this pioneering work. The results are shedding fresh light on the transition to behaviour associated with the Neolithic: settled lifestyles linked to farming and residential sites that were occupied over the long-term.
Rather than using a single site to explore the transition from forager to farmer and the emergence of larger-scale societies, these surveys and excavations have provided a detailed understanding of multiple communities in a restricted area. The sites cover the period c.15,000 BC to 5000 BC, and lie outside the classic area of the ‘Fertile Crescent’ in the Middle East, where work on the appearance of the earliest settled and farming communities had previously focused. This has provided distinctive understandings of the local ways that mobile Epipalaeolithic communities developed into settled agricultural communities. Here we will consider three sites in particular: Pınarbaşı, a rock shelter on the edge of the Konya Plain first used by Epipalaeolithic groups; Boncuklu, the earliest Neolithic settlement known on the Plain; and Çatalhöyük, which lay 9km away and developed into an impressive Neolithic village. Evidence for lifestyles becoming more settled emerges in the 10th millennium BC at Pınarbaşı and Boncuklu, followed by small-scale cultivation, herding, and pottery appearing in the mid 9th millennium BC at Boncuklu. Very large settlements become apparent during the 7th millennium BC, as at Çatalhöyük, which were more committed to substantial mixed farming. We have also been able to document the differing responses of contemporary communities to such developments, suggesting a complex mosaic of adoption, rejection, and spread of sedentism and farming.
David French’s investigations of Neolithic and Chalcolithic settlements on the Konya Plain at Canhasan more than 50 years ago were already innovative for developing techniques like flotation and extensive surface-scraping. A very large range of new approaches have also been adopted at Pınarbaşı, Boncuklu, and Çatalhöyük. Alongside now-standard approaches, these include ancient DNA (henceforth aDNA), isotope analyses of human and animal bone, fingerprint studies, chemical sourcing, biomarker studies, microarchaeology, and soil micromorphology, which we have combined with the well-preserved, fine-scale nature of stratigraphy and microstratigraphic investigation. This has allowed unique insights into the stories of individuals and households caught up in the process of the development of the Neolithic. Such an approach contrasts with many of the classic broad-scale narratives of the appearance of sedentism and farming communities, which seem unconnected to the experiences and decisions of those whose collective practices created what we see as the Neolithic Transition. Some of these insights are shared here as an indication of the exciting possibilities of such methods, when integrated with detailed stratigraphic and contextual records.
The regional context
The Konya Plain is an enclosed basin in the high-elevation central Anatolian Plateau (c.1,000m above sea level). It is one of the driest areas of the peninsula, with modern rainfall at c.320mm/year. At the height of the last Ice Age, though, an extensive lake covered much of the basin. Following its gradual drying, deltas of the Çarşamba and May rivers developed over the former lake soils, with a swathe of semi-arid steppe around them, and the Boz and Taurus Mountains beyond. Numerous analyses carried out in the area indicate that, during the Early Holocene (c.9500-6000 BC), this steppe would have been wetter and hosted a species-rich savannah with large mammals like aurochs, wild asses, and wild boars, making it a prime target for forager communities. During the later stages of the Ice Age and Early Holocene, the surrounding hills and mountains were covered by woodland, which was rich in nuts and woodland species. By contrast, the Çarşamba/May delta, with its fertile alluvial soils and abundant surface water, would have created a distinctive environment for experimentation with early cereal-farming techniques.
Within the river delta, some of the archaeological evidence from these early communities is either buried by later deposits, or damaged by later human activity on the same spot, and thus often limited to a few scattered artefacts. Conversely, archaeological sites in the highlands are often eroded away from the slopes or more difficult to detect due to their poor visibility. These limitations notwithstanding, numerous surveys have identified several sites broadly contemporary with Pınarbaşı, Boncuklu, and Çatalhöyük. In the highlands, these include rock shelters, temporary camps, and single artefacts (mostly points, which are indicative of hunting activities), broadly highlighting the presence of mobile groups. In the delta, archaeological sites are essentially represented by mounded settlements, called höyük in Turkish. These human-made hills were created by the superimposition of settlement layers and the slow decay of building materials (mud bricks, straw, and wood) over long periods of time, often centuries or millennia. They are generally interpreted as evidence for larger communities, who occupied the site on a more permanent basis.
Even if we cannot currently provide a precise chronological framework for these surveyed settlements, we can suggest that the Boncuklu community (late 10th-early 8th millennia BC) in the fertile delta would have been surrounded by other small sedentary groups, possibly no larger than 100-150 people in size. These represent some of the earliest farming and sedentary forager communities of central Anatolia and the Near East. For instance, the excavated Pınarbaşı rock shelter shows that its inhabitants did not consume domesticated plants and animals, even though the Boncuklu people – a mere 30km away – were actively experimenting with both. At the same time, groups in the surrounding savannah and highlands may have continued to lead a more mobile forager lifestyle. While it is difficult to reconstruct archaeologically small-scale interaction between these different communities, isotope analysis at Boncuklu (see below) suggests that several individuals there may have come from areas that were within the plain, but outside the immediate environs of the settlement in the Çarşamba delta. Evidence of larger-scale exchange comes from chipped-stone assemblages, which are dominated by obsidian extracted at Göllü Dağ and Nenezi Dağ in Cappadocia. These two volcanoes are more than 200km away from the Konya Plain, a likely 15- to 20-day return trip for prehistoric travellers. Complex patterns of procurement and distribution are apparent and vary considerably across sites and time. This suggests that, while it is possible that some obsidian was aquired at source by communities in the delta, it is more plausible that at least some exchange brought different communities into contact with each other, allowing further transfer of techniques, ideas, and other products.
During the 7th millennium BC, at the time when Çatalhöyük grows to become one of the largest known villages in central Anatolia, there is surprisingly little evidence for other contemporary farming sites in the delta, even as mobile groups exploit its periphery. In fact, both archaeological and textual evidence indicate that these mobile communities were still present until well into the Chalcolithic and Bronze Ages, when they gradually transitioned from foraging to specialised pastoralism.
Survey projects document a rapid expansion, between the late 7th and early 6th millennia BC, of settled farming communities (marked by höyük-sites) into mountain valleys and other sub-optimal ecological niches outside the Çarşamba delta. While the cause is currently unknown, it is possible that innovations in agricultural practices (for example, the introduction of the plough) may have contributed to an ability to turn previously marginal areas into prime farmlands.
Long-term population history
Recent work on the aDNA from human remains at Pınarbaşı, Boncuklu, and Çatalhöyük – with teams led by Somel and Fernandez-Dominguez – has provided insights into the long-term population history of the area, as well as knowledge of genetic relationships between individuals from these sites (discussed further below). Combined with isotope work by Pearson’s team, we can also understand mobility at the scale of the individual. These isotopes allow us to see whether individuals mostly spent their earlier childhood in the Konya Basin, or beyond in surrounding terrace or mountainous areas, and whether they operated in more southerly or northerly parts of the region. The aDNA work on an Epipalaeolithic individual from Pınarbaşı, who lived c.13,500 BC, indicates that the Konya Plain population was distinct from others in south-west Asia, which are documented in the broadly contemporary Levant and Zagros. Perhaps unsurprisingly, the Pınarbaşı population shared some genetic ancestry with Levantine individuals and with Upper Palaeolithic groups in south-east Europe. There also appear to be high degrees of genetic continuity with the later Boncuklu population (dated from c.8500 BC), indicating ongoing occupation or at least use of the Anatolian Plateau from the end of the Ice Age through into the first two millennia of the Holocene.
Intriguingly, some gene-flow from areas to the east – specifically the northern Zagros and north Mesopotamia area – is suggested during this time-period, probably as a result of modest movements of small numbers of individuals to and from central Anatolia, something which is also indicated by the exchange of items and ideas. We can set this beside the strontium- and oxygen-isotope work, which suggests the mobile Epipalaeolithic group were fairly localised in their movements, keeping relatively close to their settlement sites in the Konya Basin, as – by and large – did the Boncuklu community. This reveals, too, that it was indigenous communities on the Konya Plain who adopted sedentary behaviours, cultivation, and herding, rather than them being introduced by migrating communities of farmers. More specifically, we can suggest local communities of sedentary foragers probably adopted cultivation from areas to the south and east, with the idea spreading along well-established exchange routes. The development of herding is more likely to have involved manipulation of local species, as suggested by recent aDNA work on sheep and goats. Intriguingly, communities of western Anatolia involved in the earliest transmission of farming into Europe were also genetically relatively closely related to the Konya Plain populations.
The aDNA suggests that the Çatalhöyük community was genetically related to their Boncuklu predecessors, a conclusion supported by very specific continuities in the use of domestic space and ritual between the two sites. At the same time, however, the aDNA also shows higher degrees of interregional gene-flow than in earlier periods, in this case with a notable link to the Levantine area. This suggests increased degrees of interaction, and probably movement of people, between central Anatolia and the north Levantine region in the period between 7500 and 6000 BC. This matches with increased levels of mobility represented in the isotope evidence from Çatalhöyük relative to the earlier sites, although this mobility is more localised. At Çatalhöyük, c.9% of sampled individuals seem to have spent significant periods of their childhoods in areas outside the Konya Basin, either because some members of the community frequently operated at a greater distance from the settlement than most of its members, or they arrived at Çatalhöyük in their teens or adulthood from other communities. This pattern appears, at least in the sample, in the later part of the occupation, from 6700 to 6300 BC. Such general population histories provide important context for our understanding of broad-scale developments in the Konya Basin and interactions with other areas. Combining such data with detailed information from specific houses allows us to tease out how individual biographies contributed to or contrasted with these broader processes.
At Boncuklu, we have excavated a series of oval mud-brick buildings. The majority of these seem to be residential structures. While their groundplan is not the same as the later Çatalhöyük buildings, there are indications of continuities between these communities in the way that the floorplans reflect a division into a ‘dirty’ kitchen area and a ‘clean’ area for socialising, sleeping, and also ritual practices including burial. When these Boncuklu buildings are oriented north- west to south-east or west to east, the kitchen area is in the north-west/west. Both areas were regularly plastered, but the kitchen area in a more frequent and patchy way, using less good plaster. The clean area saw more extensive replastering with whiter plaster. Bayesian analysis of radiocarbon dates suggests this occurred every one or perhaps occasionally two years, a pattern that matches more recent ethnohistorically attested mud-brick houses in the region. These floor sequences thus provide rough floor-clocks for the dating of features within the structures, which in the clean area include apparently ritually related painting of floors and walls, insertion of animal bones into floors and walls, construction of clay reliefs, and human burials.
We can use this to build up biographies for the houses, which are presumably related to the histories of the households that used them. For example, the floor- clock approach shows how the use of individual buildings spans from five floors to 30, amounting to roughly five to 30 years. Buildings were also rebuilt in the same location – up to six times at Boncuklu – so we can document multi-generational household histories. This process seems to represent a symbolic statement of household continuity and the transcendence of the household. That the identity of households was expressed through idiosyncratic features is clear. For instance, while wild cattle skulls are often present in houses, they were placed in different ways in different buildings: in one case, we have a pair of skulls; in another, a single horn by an entrance; on a third occasion, a skull had been placed high on a wall; and, in a fourth, the skull was subsequently retrieved, leaving the lower part of its nasal area in the wall. Although in many buildings the animal remains were on the inside of buildings, one exception involves a skull and a separate boar jaw set into the face of an external wall.
By integrating isotopes and aDNA, we can investigate the intertwined life of houses and individuals as they formed households with distinctive identities. To take the example of one sequence, in the 83rd century BC an earlier house, known as Building 12, (henceforth B12) was replaced by B14, which in turn was replaced by B5. B12 had a minimum of 28 floors in the clean area and B14 had a minimum of 27, while B5 had only five. Two individuals were buried in B12: the first was a child and the second an adult male. The latter was the brother of the first person interred in the succeeding B14, in this case an adult female. She was buried in an early floor of B14, and was over 30, meaning she had been alive during the whole of the life of B12. Like her brother, she remained closely associated with this sequence of houses. Intriguingly, this woman was buried with a perinatal child who was not genetically related to her or her brother. Late in the life of B14, another adult female was buried, before being followed a few floors/years later by her son. Neither of them were related to the earlier brother/sister or perinatal child, indicating some of the potential dynamics of these households. The way a perinatal child was buried with a biologically unrelated adult female shows how burial practice may express complex relationships across lineages, with the house providing a focus.
Isotope study shows that all of these individuals spent their childhoods in the Boncuklu area. Interestingly, though, isotopes also suggest that the diet of the adult son became distinct from others in the community, meaning he probably spent large parts of his adult years – in the decade before he died – at distant locations, either on an extended regular seasonal basis or for the best part of several years. Even so, he came back to be buried at Boncuklu under the same house as his mother. Intriguingly, he is one of the earliest individuals known to have been infected with Hepatitis B, so his journeys give a hint of how such diseases may have spread. All of the individuals in this building sequence show the presence of osteoarthritis, so they shared a life of carrying heavy loads, causing significant wear and tear on their bodies. It is noticeable that the burials mostly lack grave goods, except for the adolescent buried first in B12. This is a low proportion relative to graves elsewhere, but seems to be a pattern among these house burials, suggesting local household traditions.
Some centuries later, an adult woman was buried in the same area as this sequence of houses, with her grave cutting into the wall of B5. Isotope evidence shows she spent her childhood in the southern areas of the Konya Basin, either as part of a community located there or as a consequence of distant trips by small elements of the Boncuklu community. In this vignette, we see how this community constructed tightly woven social networks that transcend individual lives through associations between related living and dead individuals and buildings. This was achieved in flexible ways, accommodating biologically unrelated children and seeing the coming and going of group members. These behaviours probably contributed to the longevity of such households, and were symbolically played out in house reconstruction and elaborate house rituals.
The presence of biologically unrelated children in house burial groups becomes more common at Çatalhöyük (where the relevant part of the site is known as the ‘East Mound’ and dated to 7100-5950 BC). Houses are again built on top of each other, especially in the earlier levels, with much more frequent replastering than at Boncuklu. Breaks between a building and its successor are also more common. Some buildings are preferentially used as burial places for those living in other buildings. We have to be careful, then, when using the term ‘house’ at Çatalhöyük, since each individual building was a house, but groups of buildings also made up what we can think of as a corporate House.
Several lines of evidence point to an increased complexity in the relationships between those buried in the houses. Analysis of certain tooth traits, for example, indicated that individuals interred together in the houses at Çatalhöyük were probably not close biological relatives. This initial work by Pilloud and Larsen indicated that the social structure was not based on biological kin, but was practically organised. Thus, individuals may have been adopted or fostered out after birth, so that they had both biological and non-biological parents. In a second study, Chyle´nski and associates analysed mtDNA profiles from ten individuals buried in four neighbouring buildings. All ten individuals were found to carry distinct mtDNA lineages, suggesting that individuals buried in the same buildings were either non-kin groups or those descended down the male rather than female line. A final study undertaken by Mehmet Somel and associates managed to obtain genomes from ten individuals at the site. These were mainly young individuals, since the bones of adults are poorly preserved. The proportion of people in co-burying groups who are not related is much higher at Çatalhöyük than at earlier sites such as Aşıklı and Boncuklu. Among individuals from Çatalhöyük who were co-buried in three buildings during the middle phases of occupation, only two out of ten genomes had genetic kin identified. Biological and practical kin thus seem to have been cross-cutting.
The situation becomes more intriguing (and more complex) when it is realised that the co-burying group in the Çatalhöyük Houses consumed similar food. Isotopic evidence and detailed archaeobotanical data suggest that those buried together ate together and/or obtained their food from the same parts of the environment or using similar food-procurement methods. Individuals growing up at Çatalhöyük had complex overlapping networks, both biological and practical or fictive, on which they could call in times of hardship. At least in the middle phases of occupation, this was both an important safety net and also made demands as individuals and Houses responded to the social needs of those around them.
We know from a wide range of data-sets that the landscape around Çatalhöyük was a complex mosaic of river delta and steppe, and that each House used it in diverse ways. There was much variability between houses in the ways that resources were extracted from the environment. On the other hand, there were strong social codes, as seen particularly in the internal layout of buildings. As at Boncuklu, there were clear differences between ‘clean’ and ‘dirty’ floors, although here the former was in the north of the building and the latter in the south. The northern and eastern area in the middle phases was for adult burial, symbolic reliefs, and paintings, whereas the southern and western area was for food preparation and cooking and infant burials.
This conformity is associated, as noted above, with a mostly local population. There have been suggestions that a largely conservative and local site was slow to adopt innovations such as pig- and cattle-domestication. As also noted, this focus on conformity misses the evidence for radical variability and complexity within the site. One further way to explore this issue is provided by the isotopes already mentioned, which allow us to drill down again into the specifics of people’s lives.
Pearson and associates identified a small number of individuals – both female and male – at Çatalhöyük who returned results consistent with a childhood spent either in, or regularly travelling to, the Taurus Mountains or the limestone terrace away from the alluvial plain on which the site rests. Can we say anything further about these individuals and how they played a role in Çatalhöyük society? Pearson had already noted that these burials of non-locals are not clustered in one house, but spread through a diversity of residences. None occurs in the more elaborate houses, where more burials occurred. What about the detailed burial locations of these individuals? The data are summarised in the table opposite.
The elaboration index measures the number of symbolic fixtures and fittings, such as bull horns and wall paintings, in a building, and it varies from 2 to 42 for the site as a whole. The two elaboration numbers shown in the table are low, but the measure can only be calculated for buildings that have been mostly excavated – many of the non-locals are in very eroded or partially excavated buildings. The number of burials per building vary from 0 to 60 across the site as a whole, so the numbers in the table are again low.
The sample is small, but there is a suggestion that the non-locals are not in the more elaborate buildings. Instead, they are found in modest houses, and often buried in marginal places within those buildings. All the non-local individuals sampled are adults, and Scott Haddow shows that in the Middle and Late time-periods, when these non-locals are buried, adults are less frequently buried in side-rooms or near hearths and ovens in the southern portion of main rooms. Given the large number of individuals studied by Pearson in the more elaborate buildings with more burials, the pattern seems significant.
So there is some evidence of a marginalisation of non-locals, in that they tend to be buried in less elaborate buildings that contain fewer burials, and in the southern and western – that is, the cooking and food-preparation areas – of those buildings. Strong social codes may have led to this marginalisation. Perhaps a related piece of evidence is that, in several cases, we have found skeletons showing evidence of malformity or congenital disease that were not buried in houses, but in marginal locations such as open areas or midden.
The sequence of sites studied over many decades on the Konya Plain allows us to compare rich data-sets, and to push the boundaries of conventional archaeological explanation. Put together with the use of modern techniques such as isotope and aDNA research, it allows us to observe not only long-term trends but also the specifics of individual lives. It is perhaps not surprising that a series of projects focused on one area should detect a certain insularity, but there seem to be good reasons for identifying continuities both at the individual, house, and site levels, and at the regional level. Nevertheless, for all this emphasis on tradition, there is also change, as in the realigning of the Boncuklu clean/dirty spatial arrangement from south/north to north/south at Çatalhöyük. And there is transfer in and out in terms of exchange and some increasing movement of people. Rather than insularity, there is a strong local tradition that follows its own pathways.
Perhaps most striking is that the detailed analysis of individual lives has uncovered patterns we did not expect. Throughout the sequence, individuals seem to be embedded within highly complex relationships. The old idea of social and economic life being built on the nuclear or the nuclear extended family does not seem to work. Rather, increasingly through time it is the house itself and the activities associated within it that form the main social unit. Already at Boncuklu, biologically unrelated children are somehow accommodated, and the group sees the coming and going of its members. By the time of Çatalhöyük, the individual house was cross cut by affiliations to a larger House, but also to biological links beyond the House and to a range of associations that linked houses together.