Institute of Archaeology


Activity structures of early farming settlements of Northwestern Romania in a comparative perspective


In 1972, Kent Flannery postulated a general development from small round houses and communal storage to the development of rectangular houses with internal storage and the restriction of sharing to a core family in the course of Neolithisation (see also Flannery 2002). Ian Hodder (1990) has famously claimed that the domestication of the human mind, the adaptation to different forms of co-habitation precedes the domestication of plants or animals (cf. Wilson 1988). These social changes have mainly been traced by changes in settlement structure and house architecture (Banning 2003).

Near East

Byrd (1993) set out some of the expected architectural traits indicative of increased intra-settlement co-operation and integration for the early Neolithic of the Levant, coinciding with an shrinking social network for sharing (household level instead of band/tribe level) and the development of "more institutionalized mechanisms for integrating the community as a whole" (ibd., 641). These would be the existence of communal buildings like shrines and communal areas to congregate and changed house-forms. Privacy was expected to increase (restricted physical and visual access to houses) and features for storing, preparing and consuming food are expected to shift to inside the houses.

In addition, Kuijt and Goodale (2009) have looked at changing patterns of artefact distribution between the Natufien and the PPNA, concluding that there was a change from a fairly unorganised use of space in the Natufien to a "more delineated use of space" during the PPNA, maybe as a result of decreased mobility. Karen Wright (2000) has interpreted the distribution of ground stone and cooking facilities in Levantine settlements to indicate a change from public food-preparation and consumption in the early to food- and storage related activity zones inside the houses the late PPNB, which probably correspond to a changed position of women in society. The existence of communal features and structures like shrines, cemeteries and rubbish dumps (Hardy-Smith and Edwards 2004) has also been interpreted as evidence for the genesis of true village communities. This would be accompanied by changes in the social structure of the settlement and the formation of supra-family solidarity.

In contrast, the ceramic Neolithic society at Çatal Höyük has been described as entirely house-based, without any obvious communal structures. Here, memory and bodily practices served to bind individuals into tight domestic groups that made living together on a restricted space and delayed consumption feasible (Hodder and Cessford 2004).

South-East Europe

The early Neolithic settlements of the Balkan area have generally been interpreted as villages, even if they generally lack communal features. The tightly built-up nature of most tell settlements necessitated some form of planning, and for Stella Souvatzi (2013), communal efforts like the building of ditches and boundaries created a tightly knit community. John Chapman (1997) has also emphasized how these long-term settlements established a very tangible relationship with the past. The term house-societies, derived from the work of Levi-Strauss, has also been applied to early Neolithic cultures (Boric 2007). 

There have been some studies of house-related activity areas, but they were mainly concerned with the problem of incipient craft-specialisation (Nanoglu 2008). The houses and settlement plans of the early Neolithic Starčevo/Körös/Criș culture are not very well known outside the distribution area of tell-settlements in the South-Eastern part of the distribution area of this culture. The number of house-plans excavated is limited, and, with the exception of Ecsegfalva 23 in Hungary (Whittle 2007), no detailed analysis of finds distribution has yet been undertaken.

Beyond the Danube, there are almost no tell settlements. In open settlements, the Neolithic surface is rarely preserved, thus the finds from pits offer the only clue to activities conducted both in the houses and in the areas between individual buildings. Lüning (1982) claims that there are no village-communities proper in the LBK settlement-area, even if several, sometimes even up to a hundred houses can be found in one agglomeration. Other authors have pointed, however, to the existence of communal rubbish dumps (Fritsch 1998; Hoppe and Kuhlmann 2012) and, of course, ditches. The lack of preserved occupation layers as well as the unresolved relation between houses and pits in general (cf. Stäuble and Wolfram 2012) makes the question of settlement organisation in the LBK difficult to solve.

Excavations in Tășnad-Sere

The site of Tăşnad-Sere (47°46'38''/22°58'20''), located near the town of Tășnad, Satu Mare county, North-Western Romania (Transylvania), was chosen for long-term research into intra-site distributions because it offers excellent preservation.

The settlement remains are buried almost two meters deep under sediments deposited by the Cehal River, a tributary of the Ier that flows into the Tisa/Tisza. Under this riverine deposit, the postholes of early Neolithic houses and the pits accompanying the houses are covered by an additional 30-50 cm of occupation-layer that has been preserved in-situ.

In 2012, we discovered early Iron Age (Gava) sherds in ca. 1m depth, which shed an important light on the sedimentation-history of the Cehal-river. More extensive test-trenches are planned for the future.

The Satu Mare County Museum has been conducting rescue-excavation here since 2001, uncovering an occupation layer, the postholes of houses, pits, ovens and numerous finds from the early to middle Neolithic period and the Copper Age. This means that the fine-grained results of our research-excavations can be linked-in to wider-scale explorations of the site in its entirety.

Excavation methods

Our project started in 2012 as a rescue excavation. The aim was to elucidate the settlement structure and intra-settlement activity areas of this early farming village. The excavation is a joint project UCL Institute of Archaeology and the Satu Mare County Museum/Muzeul Judetean Satu Mare (Astaloș et al. 2013). It is scheduled to continue for at least three more years.

Where possible, we dig in natural layers, otherwise in 5 cm spits. For every 1x1 m square, 5 cm spit and context, environmental samples are taken and processed on site. Wet-sieving is targeted at discovering small items like beads and fishbones. Every square is documented with plans and photogrammetry.

The precise provenance, orientation and dip (in both axes) of each find >1cm is recorded. This is an unusual way to excavate Neolithic settlements and more similar to techniques commonly used on Palaeolithic sites (cf. Hahn 1988). This allows a comprehensive assessment of the depositional processes at work (Sommer 1991, 1992) and a detailed analysis of the finds-distribution in the occupation layers. Once pits have been uncovered, the relationship between finds in the occupation layer, the finds in the pits and any possible house-plans can be investigated.

Finds consist of pottery, chipped stone artefacts made from obsidian, local and imported flint and quarzite, groundstone axes, hearth-fragments, daub and animal bones. Refitting sherds are often found close together, indicating distinct dumping episodes.

There is no indication of any preferred orientation of the finds, making riverine deposition highly unlikely.