by Chris Fowler
(after Riedhammer 2003, fig. 2). Not to scale.
The orthodox functional interpretation of the three parts of the LBK house also relies on circumstantial evidence. The ever-present central part is seen as the main living/working room, burnt material in post-holes sometimes suggesting a hearth at the rear (e.g. Modderman 1988, 96; Kirleis and Willerding 2008, 143). The north/west part, often enclosed with a wall-trench in later phases, is considered a secluded living/sleeping room: Coudart (1998, 105) talks about the ‘banality’ of its spatial arrangements. The south/east part, frequently with rows of doubled posts, is interpreted as a raised area for grain storage: the Merzbachtal excavations revealed more chaff near houses with this section (Lüning 2000, 158) and its disappearance in later phases at Štúrovo coincides with the presence of large storage pits (Pavúk 1994, 245–247). However, whilst the early LBK houses demonstrate the fundamental importance of the central part, their shorter southern parts might suggest different interpretations (Stäuble 2005a, 191–194). And although geochemical analyses remain rare, they have generally failed to corroborate clear functional differences between the three parts (Lienemann 1998; Stäuble and Lüning 1999). This suggests we should move away from static, unifunctional interpretations: the ‘granaries’, for instance, could perhaps be rethought in terms of a ‘versatile structure’, like the Karo rice barns of south-east Asia (Waterson 1990, 59), which serve multiple functions without losing their sacred aspect.
MN houses often have fewer internal posts and lack a clear modular arrangement; direct clues to the functions of different parts remain sparse, though one exception is the consistent placement of hearth-pits within the Großgartach houses at Jülich-Welldorf in the Rhineland (Dohrn-Ihmig 1983). Most likely, entrances were still at the south/east end (Hampel 1989, 71), which was often widened as structures became more trapezoidal in plan, or elaborated with a porch (Fig. 14.1c); there may therefore be a greater symbolic focus on the threshold. The Lengyel houses of Slovakia diverge most from the Danubian tradition, with a bipartite division and evidence for an upper storey, again suggesting south-eastern influences (Pavúk 2003); in contrast, late Lengyel houses in the Kujavia region of Poland are trapezoidal structures in the Danubian tradition, lacking internal partitions (e.g. Grygiel and Bogucki 1997).
Understanding the domestic domain requires ‘a focus on all venues of domestic life’ (Robin 2002, 261) many of which would have lain outside the house (Pavúk 1994, 254; Whittle 2003, 141). External household space was structured by various features, most notably the flanking pits, which are usually interpreted as constructional features supplying daub for the walls and then re-used for refuse disposal (see below). By mirroring the internal structure of the buildings, they served to make the segmentation of external space comparable to that of house interiors (Bradley 2001, 52; see Fig. 14.1). Apart from these, certain features, such as pit-ovens, are more commonly found outside than inside houses (e.g. Lenneis 1995, 18), though there is much regional variation in their occurrence (Lüning 2004). Ancillary structures are rare (but see Wüstehube 1993), so activities requiring shelter must generally have taken place inside the longhouse (Lüning 2000, 157). Fences may indicate garden areas or stock enclosures, though their relationships to individual houses are often unclear (Pavúk 1994, 253f).
The most influential model for understanding the organization of household space is the concept of the Hofplatz, devised for the Merzbachtal settlements. This denotes a ‘farmyard’ area with a radius of 25m around the house, marked by pits in specific locations (Boelicke 1982) (Fig. 14.2a). However, its applicability elsewhere is doubtful, even for nearby sites (e.g. Bernhardt 1986; van de Velde 2007a), let alone those further afield (Pavúk 1994). Many settlements, such as Ulm-Eggingen in Baden-Württemberg (Kind 1989) or Cuiry-lès-Chaudardes in the Paris Basin (Ilett and Hachem 2001), have far fewer pits away from the longhouses than the Merzbachtal sites. Elsewhere, features were arranged differently, for example the semi-circle of pits around some houses at Brunn-Wolfholz (Lenneis et al. 1996, 102). Though it has generated plausible settlement models for the Merzbachtal, the Hofplatz concept can also be criticized for simplistic assumptions about the succession of houses and the fill mechanisms of pits (Claßen 2005, 118; Frirdich 2005, 94f).
In the MN, the model of extended household space—however it was organized—generally breaks down, with flanking pits less common and large communal pit complexes more in evidence. Exceptions include the early Stichbandkeramik (SBK) in Bavaria, where continuity of settlement and even individual Hofplätze is suggested (Herren 2003), and Villeneuve-Saint-Germain (VSG) settlements in northern France, which retain a layout akin to local late LBK sites (Bostyn 2003). A similar spatial structure reappears in the ‘household clusters’ of late Lengyel Poland (Grygiel 1986). Meanwhile, fenced enclosures directly attached to longhouses seem more common in the MN, especially the SBK (Riedhammer 2003), indicating that formal division of external household space was sometimes appropriate.
THE SOCIAL HOUSEHOLD
Whatever spatial models are developed, it is hard to make sense of how houses were occupied without some understanding of social structures. Unfortunately, inferences are difficult and even estimates of household size vary widely (Pavúk 1994, 256–258; Sommer 2001, 259). Usually households are seen as fairly small: the mass grave at Talheim has been taken to suggest the presence of nuclear families (Bentley et al. 2008), although such a group may not represent a complete household, whilst Lüning (1982) has suggested that it was only with the large Rössen longhouses that multi-family dwellings appeared.
The number of inhabitants should to some extent be reflected in the quantities of material associated with a house, along with the duration of occupation. The latter is generally not thought to exceed 25–30 years, although in principle houses could have stood much longer (Sommer 2001, 259) and there are arguments for longer lifespans in the early LBK (Lenneis and Stadler 2002, 200; Stäuble 2005a, 204f). The size of household ceramic inventories seems to vary widely, however (cf. Pavúk 1994, 174–180; Lanchon 2003; van de Velde 2007b, 120f), suggesting large differences in numbers of occupants, the use of pottery, or site formation processes (see below). Perhaps the intensity of occupation varied over the lifespan of a house: Whittle (1996, 162f; 2003, 141) has suggested that LBK longhouses were compatible with mobility in people’s lifestyles and fluidity in the composition of social groups, an idea supported by recent evidence for transhumance (Bentley and Knipper 2005). In any case, we must confront the assumption that houses which looked the same were inhabited in the same way.
Social models of the LBK household have principally been derived from the associated ceramic assemblages, with the distribution of different motifs seen as evidence for exogamous virilocal residence patterns (e.g. van de Velde 1979; Krahn 2003). Support for virilocality has also come from isotope studies of sites in south-west Germany, with suggestions of non-local (possibly forager) women marrying into the resident group (Bentley 2007; Bentley et al. 2008; but see Bickle and Hofmann 2007). A moiety system has been suggested for settlements showing structured distributions of specific decorative motifs on pottery (van de Velde 1979; 2007b). Although affiliations of this type were rarely marked in domestic architecture or spatial organization, buildings in different parts of the site at Vaihingen-Enz in Baden-Württemberg were differently organized (Krause 1998, 15), and houses with particular features (wall trenches) may correlate with specific ceramic traits (Strien 2005, 195).
Of course, residential groups and descent groups were probably not the same, but whilst lineages might have been distributed across contemporary houses at different sites, the location of new buildings seems to reflect hereditary principles. The arrangement of successive houses at Schwanfeld and Langweiler 8, for example, shows how genealogical connections were made symbolically visible through the construction of longhouses in particular spatial relationships to their predecessors (Lüning 2005).
Danubian societies are usually seen as egalita
rian (but see Jeunesse 1997), though specialized roles may have existed. At the VSG site of Poses, material culture distributions suggest houses had different functions, but no true specializations (Bostyn 2003, 212) whilst at late LBK Cuiry-lès-Chaudardes, Hachem (2000, 310f) argues for a relationship between house size and the relative importance of hunting and herding (on family ‘occupations’, see also Bentley et al. 2008, 301). The distinctive type 1a LBK buildings with a continuous wall trench (Modderman 1970, fig. 12) are sometimes considered to be chiefly or elite dwellings because there are usually only one or two per phase (van de Velde 2007c, 237–238), and their associated finds do not suggest a communal function (van de Velde 1979, 140f; but see Milisauskas 1986). However, the idea that bigger houses imply higher status can be criticized (Sommer 2001, 258f). It may be more productive to consider house size and form as related to ebbs and flows in household composition, or the social network a household could draw on during construction. On the other hand, type 1a houses were sometimes treated differently on abandonment: although evidence for deliberate burning is rare in the LBK compared to south-east Europe, burnt houses of this type are known from several sites (e.g. van de Velde 2007d), suggesting a special significance.
LIFE IN THE LONGHOUSE
For Whittle (2003, 138f), longhouses encouraged certain ways of moving. In part, we can get at these routines through phenomenological approaches (Hofmann 2006, 88–91), but fuller understanding depends on the finds associated with each building. The surface of a Danubian settlement no doubt resembled the Maya farmsteads analysed by Robin (2002, fig. 2 and 257), with their palimpsests of pathways, work and refuse areas marking ‘people’s diverse and often entwined lifeways’. Unfortunately, very few LBK sites have preserved surfaces, and it has even been suggested that the houses had raised floors (Rück 2007). However, at the unusual site of Hanau-Klein-Auheim hearths were found inside and outside houses, with finds distributions suggesting activity in front of house entrances (Sommer 2006). This contrasts with the MN site of Jablines in northern France, where there is evidence for activities within and behind two buildings, but few finds from the presumed entrance areas (Hachem 2000, 308f). These scanty data at least confirm that external areas were used as intensively as the houses themselves. Similarly, in Stäuble and Lüning’s (1999) phosphate study of early LBK houses the highest values came from areas behind the houses.
Otherwise, we are dependent on pit assemblages for evidence of the cumulative patterns of domestic life—though as secondary refuse deposits these are hard to interpret (Last 1995, 1998). The Hofplatz model of household space requires various assumptions about the locations of activities and associated discard practices—but if households were small and work collaborative, refuse patterning could be independent of the Hofplätze (Frirdich 2005, 94f). The contents of the flanking pits therefore most likely reflect activities associated with specific houses and are usually assumed to have filled gradually during the lifetime of the adjacent longhouse (Coudart 1998, 73). However, based on stratigraphic relationships between the flanking pits and the outer trenches of early LBK houses, Stäuble (1997, 2005a) argues that pits may have been filled during house construction, their contents presumably deriving from middens already present in the settlement. Similarly for Birkenhagen (2003, 148), the low quantities of finds and homogeneous fills in many flanking pits (notably on the Merzbachtal sites) contradict assertions that they were open for the entire lifespan of a house. On the other hand, persistent patterning in the distribution of finds within these pits at other sites, such as Cuiry-lès-Chaudardes (Hachem 2000, 310) and Poses (Bostyn 2003) (Fig. 14.2b), would be unlikely to occur if pits had been backfilled. At Cuiry there was also a tendency for each household to discard faunal material on one side rather than the other, suggesting that social or cosmological aspects of longhouse architecture could have influenced discard. At Olszanica in Poland, certain types of lithic artefact were mainly on one or the other side of the house, which is tentatively interpreted as indicating male and female areas (Milisauskas 1986).
Pit contents—like other aspects of external space—therefore provide a far more varied scenario than the architecture. Even for the early LBK, Stäuble’s (1997) interpretation may not apply at Neckenmarkt and Strögen in Austria, which showed more varied patterning in the distributions of different materials, suggesting activity zones in front (south) of the houses (Lenneis and Lüning 2001). Examples of spatial patterning of materials from later LBK sites include pottery (Boelicke 1988), lithics (de Grooth 2003), animal bone (e.g. Marciniak 2004), and charcoal (Lüning 2000, 158), though in other cases it is harder to identify clear trends (e.g. Kvetina 2007). Untangling patterning in longhouse routines becomes even harder for the MN, when—with the exceptions of the early SBK and VSG cultures—flanking pits are frequently absent (Coudart 1998, 52).
We also need to consider other modes of deposition beyond generalized discard. For instance, Bostyn (2003, 208) suggests the homogeneity of some groups of lithics at Poses reflects specific events; similarly, Allard (2005) has studied concentrations of knapping waste in pits at Verlaine in Belgium. Some finds from flanking pits may represent deliberately placed deposits, a comparatively under-researched area (Hofmann 2006, 84–86). Examples include a complete inverted early LBK pot from Enkingen (Reuter 1991) and a deposit of grindstones from the late LBK site of Irchonwelz (Constantin et al. 1978); human burials from settlements should also be considered here (e.g. Veit 1993; Schmotz 2002). In the MN, especially the SBK and Lengyel, placed deposits also occur within buildings, such as at Postoloprty in Bohemia (Soudsky 1969).
We should not, however, draw too rigid a distinction between structured deposition and ‘normal’ refuse. The latter would have had its own symbolic qualities and connotations, not necessarily negative (Douny 2007). For Marciniak (2004, 137), LBK pit-digging and filling can be seen as an intervention into the ancestral past, whilst Hodder (1990, 127) points out the conceptual significance of changes in the MN, when the use of communal pit complexes for refuse disposal meant that discard practice no longer marked out each house as an independent unit.
NO HOUSE IS AN ISLAND
Individual longhouses were linked, physically and socially, within wider settlements and landscapes. Settlements varied in size from ‘farmsteads’ with a single house (e.g. Pavlů 1998) to ‘villages’ with 20 or more contemporary houses (e.g. van de Velde 2007c, 233) (Fig. 14.3). Site structure could influence architecture, with the standardization of house form greater at more isolated sites and in less densely populated areas (Coudart 1998, 96), but in general the longhouse principles took precedence: building orientation, for instance, is the same regardless of settlement layout or topography, hindering the creation of communal space (Zimmermann et al. 2005, 31). Even where open areas (Pavúk 1994) or enclosures (e.g. Krause 1998) created a concentrically ordered space, the orientation of individual houses conformed to the standard model. Only in the Lengyel culture was house orientation sometimes subordinated to settlement structure, notably at the rondels of Polgár-Csőszhalom and Svodín (Pasztor et al. 2008), again emphasizing the non-Danubian aspects of Lengyel houses.
FIG. 14.3. Examples of settlement layouts: (a) rows of successive early LBK houses at Schwanfeld (after Gronenborn 1999, fig. 11a); (b) houses and enclosures at Erkelenz-Kückhoven (after Lehmann 2004); (c) related geometry of selected houses and the northern enclosure at Köln-Lindenthal (after van Berg 1989, fig. 4); (d) Rössen longhouses with suggested Hofplätze (shaded) and communal structures at Inden (after Lüning 1982, fig. 12).
Although rows of houses on sites like Schwanfeld (Lüning 2005) and Cuiry-lès-Chaudardes (Ilett and Hachem 2001) seem to represent sequences of building rather than planned settlements (but see Rück 2007), there is some evidence for higher-level organization. At Geleen-Janskamperveld, two ‘wards’ divided by a central space may represent different lineage groups (van de Velde 2007c, 237–238) whilst at Bylany in Bohemia small (type 3) houses were concentrated in one par
t of the site (Coolen 2004, 79), as were the houses associated with high numbers of wild animal remains at Cuiry-lès-Chaudardes (Ilett and Hachem 2001, 182). The distribution of finds and features at Landshut-Sallmannsberg in Bavaria suggests an ordered village (Brink-Kloke 1992) but the different layouts of other Bavarian sites show that the community-household dialectic played out differently even within one region.
Most sites contain pits that cannot be assigned to individual houses, whilst other communal features include wells, as at Erkelenz-Kückhoven (Weiner 1995), and groups of ovens, as at Olszanica (Milisauskas 1986). Also significant are the enclosures (see Petrasch, this volume), which could reflect or establish relationships of inequality between ostensibly similar households, because they were either laid out with specific reference to particular houses (van Berg 1989) or marked out certain houses as ‘central’ or ‘peripheral’ (Lehmann 2004, 295–298).
Finally, we need to acknowledge links between houses on different sites. For Whittle (2003, 143), LBK society was constituted by multiple identities, alliances, and exchange networks (as well as more violent interactions), whilst Bogucki (2003) envisages a web of kinship ties and exchange relationships between households. A key result of the Merzbachtal excavations is the evidence for differential participation of settlements in lithic exchange networks, with the relative quantities of debitage and tools at different sites suggesting only a few had direct access to flint sources and then supplied their neighbours (Zimmermann et al. 2005, 30f). No doubt the uniformity of the longhouse played a role in broader cultural integration, although the nature of household social networks changed over time. For the Merzbachtal sites, Frirdich (1994) noted a change in settlement patterns related to the transition from Flomborn pottery to more variable late LBK styles. This is interpreted as a shift from a homogeneous material culture, aiding the integration of people from other regions, to an emphasis on local identities, also reflected in features like enclosures. But notably, compared to the transition between early LBK and Flomborn, house form itself changed relatively little.
